diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21273 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.5258439824527943,
+  "eval_steps": 100,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000508614660817598,
+      "grad_norm": 10.098897560307945,
+      "learning_rate": 6.779661016949153e-08,
+      "loss": 0.8632,
+      "step": 1
+    },
+    {
+      "epoch": 0.001017229321635196,
+      "grad_norm": 10.041067862694794,
+      "learning_rate": 1.3559322033898305e-07,
+      "loss": 0.8846,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015258439824527943,
+      "grad_norm": 7.927388636908495,
+      "learning_rate": 2.0338983050847458e-07,
+      "loss": 0.8551,
+      "step": 3
+    },
+    {
+      "epoch": 0.002034458643270392,
+      "grad_norm": 8.10881143423477,
+      "learning_rate": 2.711864406779661e-07,
+      "loss": 0.8714,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025430733040879903,
+      "grad_norm": 9.38203064823776,
+      "learning_rate": 3.3898305084745766e-07,
+      "loss": 0.8486,
+      "step": 5
+    },
+    {
+      "epoch": 0.0030516879649055886,
+      "grad_norm": 9.870435987928643,
+      "learning_rate": 4.0677966101694916e-07,
+      "loss": 0.8984,
+      "step": 6
+    },
+    {
+      "epoch": 0.0035603026257231864,
+      "grad_norm": 8.96993574750974,
+      "learning_rate": 4.745762711864407e-07,
+      "loss": 0.8254,
+      "step": 7
+    },
+    {
+      "epoch": 0.004068917286540784,
+      "grad_norm": 8.305190005669553,
+      "learning_rate": 5.423728813559322e-07,
+      "loss": 0.9104,
+      "step": 8
+    },
+    {
+      "epoch": 0.004577531947358383,
+      "grad_norm": 8.557301077641153,
+      "learning_rate": 6.101694915254238e-07,
+      "loss": 0.844,
+      "step": 9
+    },
+    {
+      "epoch": 0.005086146608175981,
+      "grad_norm": 8.971079666670127,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 0.7577,
+      "step": 10
+    },
+    {
+      "epoch": 0.0055947612689935785,
+      "grad_norm": 7.760483081371288,
+      "learning_rate": 7.457627118644069e-07,
+      "loss": 0.7961,
+      "step": 11
+    },
+    {
+      "epoch": 0.006103375929811177,
+      "grad_norm": 7.264571181136625,
+      "learning_rate": 8.135593220338983e-07,
+      "loss": 0.7641,
+      "step": 12
+    },
+    {
+      "epoch": 0.006611990590628775,
+      "grad_norm": 6.087876526391279,
+      "learning_rate": 8.813559322033899e-07,
+      "loss": 0.7725,
+      "step": 13
+    },
+    {
+      "epoch": 0.007120605251446373,
+      "grad_norm": 7.705173924071754,
+      "learning_rate": 9.491525423728814e-07,
+      "loss": 0.8689,
+      "step": 14
+    },
+    {
+      "epoch": 0.007629219912263971,
+      "grad_norm": 6.344430370207566,
+      "learning_rate": 1.016949152542373e-06,
+      "loss": 0.7655,
+      "step": 15
+    },
+    {
+      "epoch": 0.008137834573081568,
+      "grad_norm": 7.294300875090611,
+      "learning_rate": 1.0847457627118644e-06,
+      "loss": 0.7863,
+      "step": 16
+    },
+    {
+      "epoch": 0.008646449233899168,
+      "grad_norm": 6.277409380490764,
+      "learning_rate": 1.152542372881356e-06,
+      "loss": 0.8177,
+      "step": 17
+    },
+    {
+      "epoch": 0.009155063894716766,
+      "grad_norm": 6.801980953020517,
+      "learning_rate": 1.2203389830508477e-06,
+      "loss": 0.7464,
+      "step": 18
+    },
+    {
+      "epoch": 0.009663678555534364,
+      "grad_norm": 6.371323067416573,
+      "learning_rate": 1.288135593220339e-06,
+      "loss": 0.817,
+      "step": 19
+    },
+    {
+      "epoch": 0.010172293216351961,
+      "grad_norm": 7.206478372678896,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 0.8275,
+      "step": 20
+    },
+    {
+      "epoch": 0.010680907877169559,
+      "grad_norm": 7.893454590115214,
+      "learning_rate": 1.4237288135593222e-06,
+      "loss": 0.7829,
+      "step": 21
+    },
+    {
+      "epoch": 0.011189522537987157,
+      "grad_norm": 6.661426232931436,
+      "learning_rate": 1.4915254237288139e-06,
+      "loss": 0.837,
+      "step": 22
+    },
+    {
+      "epoch": 0.011698137198804755,
+      "grad_norm": 6.9831913365363105,
+      "learning_rate": 1.5593220338983054e-06,
+      "loss": 0.8334,
+      "step": 23
+    },
+    {
+      "epoch": 0.012206751859622354,
+      "grad_norm": 6.518187972479781,
+      "learning_rate": 1.6271186440677967e-06,
+      "loss": 0.837,
+      "step": 24
+    },
+    {
+      "epoch": 0.012715366520439952,
+      "grad_norm": 5.530233171179341,
+      "learning_rate": 1.6949152542372882e-06,
+      "loss": 0.8338,
+      "step": 25
+    },
+    {
+      "epoch": 0.01322398118125755,
+      "grad_norm": 5.280904002395758,
+      "learning_rate": 1.7627118644067799e-06,
+      "loss": 0.7897,
+      "step": 26
+    },
+    {
+      "epoch": 0.013732595842075148,
+      "grad_norm": 5.888389102352453,
+      "learning_rate": 1.8305084745762714e-06,
+      "loss": 0.696,
+      "step": 27
+    },
+    {
+      "epoch": 0.014241210502892745,
+      "grad_norm": 6.597728523323801,
+      "learning_rate": 1.8983050847457629e-06,
+      "loss": 0.8288,
+      "step": 28
+    },
+    {
+      "epoch": 0.014749825163710343,
+      "grad_norm": 5.4401261575880415,
+      "learning_rate": 1.9661016949152544e-06,
+      "loss": 0.7062,
+      "step": 29
+    },
+    {
+      "epoch": 0.015258439824527943,
+      "grad_norm": 5.8683069941456765,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 0.7573,
+      "step": 30
+    },
+    {
+      "epoch": 0.01576705448534554,
+      "grad_norm": 5.935777273724286,
+      "learning_rate": 2.1016949152542374e-06,
+      "loss": 0.7532,
+      "step": 31
+    },
+    {
+      "epoch": 0.016275669146163137,
+      "grad_norm": 5.384052903276787,
+      "learning_rate": 2.169491525423729e-06,
+      "loss": 0.648,
+      "step": 32
+    },
+    {
+      "epoch": 0.016784283806980736,
+      "grad_norm": 5.9683230581925715,
+      "learning_rate": 2.2372881355932204e-06,
+      "loss": 0.7951,
+      "step": 33
+    },
+    {
+      "epoch": 0.017292898467798336,
+      "grad_norm": 5.926354799822477,
+      "learning_rate": 2.305084745762712e-06,
+      "loss": 0.7453,
+      "step": 34
+    },
+    {
+      "epoch": 0.017801513128615932,
+      "grad_norm": 5.026552839094986,
+      "learning_rate": 2.372881355932204e-06,
+      "loss": 0.7182,
+      "step": 35
+    },
+    {
+      "epoch": 0.01831012778943353,
+      "grad_norm": 5.057968749094965,
+      "learning_rate": 2.4406779661016953e-06,
+      "loss": 0.6881,
+      "step": 36
+    },
+    {
+      "epoch": 0.018818742450251127,
+      "grad_norm": 5.179037522260524,
+      "learning_rate": 2.5084745762711864e-06,
+      "loss": 0.775,
+      "step": 37
+    },
+    {
+      "epoch": 0.019327357111068727,
+      "grad_norm": 5.692479898302508,
+      "learning_rate": 2.576271186440678e-06,
+      "loss": 0.6644,
+      "step": 38
+    },
+    {
+      "epoch": 0.019835971771886323,
+      "grad_norm": 4.939393658202731,
+      "learning_rate": 2.64406779661017e-06,
+      "loss": 0.7446,
+      "step": 39
+    },
+    {
+      "epoch": 0.020344586432703923,
+      "grad_norm": 5.0805789593564885,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 0.8183,
+      "step": 40
+    },
+    {
+      "epoch": 0.020853201093521522,
+      "grad_norm": 5.496455003785856,
+      "learning_rate": 2.779661016949153e-06,
+      "loss": 0.6419,
+      "step": 41
+    },
+    {
+      "epoch": 0.021361815754339118,
+      "grad_norm": 4.840728013953906,
+      "learning_rate": 2.8474576271186443e-06,
+      "loss": 0.7752,
+      "step": 42
+    },
+    {
+      "epoch": 0.021870430415156718,
+      "grad_norm": 5.620415239028736,
+      "learning_rate": 2.915254237288136e-06,
+      "loss": 0.6877,
+      "step": 43
+    },
+    {
+      "epoch": 0.022379045075974314,
+      "grad_norm": 5.345077082164693,
+      "learning_rate": 2.9830508474576277e-06,
+      "loss": 0.6368,
+      "step": 44
+    },
+    {
+      "epoch": 0.022887659736791913,
+      "grad_norm": 5.05195008197204,
+      "learning_rate": 3.0508474576271192e-06,
+      "loss": 0.7016,
+      "step": 45
+    },
+    {
+      "epoch": 0.02339627439760951,
+      "grad_norm": 7.128357748590802,
+      "learning_rate": 3.1186440677966107e-06,
+      "loss": 0.8184,
+      "step": 46
+    },
+    {
+      "epoch": 0.02390488905842711,
+      "grad_norm": 4.476986210927716,
+      "learning_rate": 3.186440677966102e-06,
+      "loss": 0.7784,
+      "step": 47
+    },
+    {
+      "epoch": 0.02441350371924471,
+      "grad_norm": 5.063138191593033,
+      "learning_rate": 3.2542372881355933e-06,
+      "loss": 0.6435,
+      "step": 48
+    },
+    {
+      "epoch": 0.024922118380062305,
+      "grad_norm": 5.206345852888816,
+      "learning_rate": 3.322033898305085e-06,
+      "loss": 0.7068,
+      "step": 49
+    },
+    {
+      "epoch": 0.025430733040879904,
+      "grad_norm": 5.82482718673877,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 0.7353,
+      "step": 50
+    },
+    {
+      "epoch": 0.0259393477016975,
+      "grad_norm": 4.651888654535835,
+      "learning_rate": 3.457627118644068e-06,
+      "loss": 0.7217,
+      "step": 51
+    },
+    {
+      "epoch": 0.0264479623625151,
+      "grad_norm": 5.427592931485119,
+      "learning_rate": 3.5254237288135597e-06,
+      "loss": 0.7316,
+      "step": 52
+    },
+    {
+      "epoch": 0.0269565770233327,
+      "grad_norm": 5.062943238999723,
+      "learning_rate": 3.5932203389830512e-06,
+      "loss": 0.7521,
+      "step": 53
+    },
+    {
+      "epoch": 0.027465191684150295,
+      "grad_norm": 5.81533973810054,
+      "learning_rate": 3.6610169491525427e-06,
+      "loss": 0.8159,
+      "step": 54
+    },
+    {
+      "epoch": 0.027973806344967895,
+      "grad_norm": 4.515638727609432,
+      "learning_rate": 3.7288135593220342e-06,
+      "loss": 0.7087,
+      "step": 55
+    },
+    {
+      "epoch": 0.02848242100578549,
+      "grad_norm": 4.922995981596415,
+      "learning_rate": 3.7966101694915257e-06,
+      "loss": 0.7314,
+      "step": 56
+    },
+    {
+      "epoch": 0.02899103566660309,
+      "grad_norm": 4.692768239081016,
+      "learning_rate": 3.864406779661018e-06,
+      "loss": 0.6935,
+      "step": 57
+    },
+    {
+      "epoch": 0.029499650327420687,
+      "grad_norm": 4.68382526045054,
+      "learning_rate": 3.932203389830509e-06,
+      "loss": 0.7496,
+      "step": 58
+    },
+    {
+      "epoch": 0.030008264988238286,
+      "grad_norm": 4.538058723120196,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7663,
+      "step": 59
+    },
+    {
+      "epoch": 0.030516879649055886,
+      "grad_norm": 5.57577986800722,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 0.8538,
+      "step": 60
+    },
+    {
+      "epoch": 0.03102549430987348,
+      "grad_norm": 5.705087151760233,
+      "learning_rate": 4.135593220338983e-06,
+      "loss": 0.7456,
+      "step": 61
+    },
+    {
+      "epoch": 0.03153410897069108,
+      "grad_norm": 5.308335410955593,
+      "learning_rate": 4.203389830508475e-06,
+      "loss": 0.759,
+      "step": 62
+    },
+    {
+      "epoch": 0.03204272363150868,
+      "grad_norm": 4.886851064230893,
+      "learning_rate": 4.271186440677967e-06,
+      "loss": 0.7143,
+      "step": 63
+    },
+    {
+      "epoch": 0.03255133829232627,
+      "grad_norm": 5.343380385437803,
+      "learning_rate": 4.338983050847458e-06,
+      "loss": 0.6437,
+      "step": 64
+    },
+    {
+      "epoch": 0.03305995295314387,
+      "grad_norm": 4.8074810063804705,
+      "learning_rate": 4.40677966101695e-06,
+      "loss": 0.8053,
+      "step": 65
+    },
+    {
+      "epoch": 0.03356856761396147,
+      "grad_norm": 4.9017842372442395,
+      "learning_rate": 4.474576271186441e-06,
+      "loss": 0.6844,
+      "step": 66
+    },
+    {
+      "epoch": 0.03407718227477907,
+      "grad_norm": 4.947794937510279,
+      "learning_rate": 4.542372881355933e-06,
+      "loss": 0.7611,
+      "step": 67
+    },
+    {
+      "epoch": 0.03458579693559667,
+      "grad_norm": 4.7456711569191405,
+      "learning_rate": 4.610169491525424e-06,
+      "loss": 0.689,
+      "step": 68
+    },
+    {
+      "epoch": 0.035094411596414264,
+      "grad_norm": 4.917551424410196,
+      "learning_rate": 4.677966101694916e-06,
+      "loss": 0.7564,
+      "step": 69
+    },
+    {
+      "epoch": 0.035603026257231864,
+      "grad_norm": 4.4293946938058735,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 0.712,
+      "step": 70
+    },
+    {
+      "epoch": 0.03611164091804946,
+      "grad_norm": 4.614807444943532,
+      "learning_rate": 4.813559322033899e-06,
+      "loss": 0.7683,
+      "step": 71
+    },
+    {
+      "epoch": 0.03662025557886706,
+      "grad_norm": 4.256122218607177,
+      "learning_rate": 4.881355932203391e-06,
+      "loss": 0.797,
+      "step": 72
+    },
+    {
+      "epoch": 0.03712887023968466,
+      "grad_norm": 4.59974103281584,
+      "learning_rate": 4.949152542372882e-06,
+      "loss": 0.7015,
+      "step": 73
+    },
+    {
+      "epoch": 0.037637484900502255,
+      "grad_norm": 4.966299736413361,
+      "learning_rate": 5.016949152542373e-06,
+      "loss": 0.7429,
+      "step": 74
+    },
+    {
+      "epoch": 0.038146099561319854,
+      "grad_norm": 4.98842998811629,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.8923,
+      "step": 75
+    },
+    {
+      "epoch": 0.038654714222137454,
+      "grad_norm": 4.7642830381959875,
+      "learning_rate": 5.152542372881356e-06,
+      "loss": 0.7778,
+      "step": 76
+    },
+    {
+      "epoch": 0.039163328882955054,
+      "grad_norm": 5.681311961424697,
+      "learning_rate": 5.220338983050848e-06,
+      "loss": 0.774,
+      "step": 77
+    },
+    {
+      "epoch": 0.039671943543772646,
+      "grad_norm": 4.728121716283249,
+      "learning_rate": 5.28813559322034e-06,
+      "loss": 0.7414,
+      "step": 78
+    },
+    {
+      "epoch": 0.040180558204590246,
+      "grad_norm": 4.562973415948566,
+      "learning_rate": 5.355932203389831e-06,
+      "loss": 0.7609,
+      "step": 79
+    },
+    {
+      "epoch": 0.040689172865407845,
+      "grad_norm": 4.60637937283868,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 0.6995,
+      "step": 80
+    },
+    {
+      "epoch": 0.041197787526225445,
+      "grad_norm": 4.563755646458949,
+      "learning_rate": 5.491525423728814e-06,
+      "loss": 0.6325,
+      "step": 81
+    },
+    {
+      "epoch": 0.041706402187043044,
+      "grad_norm": 5.035528223992881,
+      "learning_rate": 5.559322033898306e-06,
+      "loss": 0.73,
+      "step": 82
+    },
+    {
+      "epoch": 0.04221501684786064,
+      "grad_norm": 5.485333046551667,
+      "learning_rate": 5.6271186440677975e-06,
+      "loss": 0.7851,
+      "step": 83
+    },
+    {
+      "epoch": 0.042723631508678236,
+      "grad_norm": 4.564865962021306,
+      "learning_rate": 5.694915254237289e-06,
+      "loss": 0.7482,
+      "step": 84
+    },
+    {
+      "epoch": 0.043232246169495836,
+      "grad_norm": 4.435280043094116,
+      "learning_rate": 5.7627118644067805e-06,
+      "loss": 0.7082,
+      "step": 85
+    },
+    {
+      "epoch": 0.043740860830313436,
+      "grad_norm": 5.928279101309869,
+      "learning_rate": 5.830508474576272e-06,
+      "loss": 0.7722,
+      "step": 86
+    },
+    {
+      "epoch": 0.044249475491131035,
+      "grad_norm": 4.5438001795205665,
+      "learning_rate": 5.8983050847457635e-06,
+      "loss": 0.8396,
+      "step": 87
+    },
+    {
+      "epoch": 0.04475809015194863,
+      "grad_norm": 6.417916444552532,
+      "learning_rate": 5.9661016949152555e-06,
+      "loss": 0.7807,
+      "step": 88
+    },
+    {
+      "epoch": 0.04526670481276623,
+      "grad_norm": 3.9890248411547415,
+      "learning_rate": 6.0338983050847465e-06,
+      "loss": 0.6505,
+      "step": 89
+    },
+    {
+      "epoch": 0.04577531947358383,
+      "grad_norm": 4.442241231222374,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 0.7866,
+      "step": 90
+    },
+    {
+      "epoch": 0.046283934134401426,
+      "grad_norm": 4.6914214960592195,
+      "learning_rate": 6.1694915254237295e-06,
+      "loss": 0.6912,
+      "step": 91
+    },
+    {
+      "epoch": 0.04679254879521902,
+      "grad_norm": 5.457454853369151,
+      "learning_rate": 6.2372881355932215e-06,
+      "loss": 0.7311,
+      "step": 92
+    },
+    {
+      "epoch": 0.04730116345603662,
+      "grad_norm": 4.7835960865404425,
+      "learning_rate": 6.3050847457627125e-06,
+      "loss": 0.7829,
+      "step": 93
+    },
+    {
+      "epoch": 0.04780977811685422,
+      "grad_norm": 4.839835474039535,
+      "learning_rate": 6.372881355932204e-06,
+      "loss": 0.7853,
+      "step": 94
+    },
+    {
+      "epoch": 0.04831839277767182,
+      "grad_norm": 5.164692758024435,
+      "learning_rate": 6.440677966101695e-06,
+      "loss": 0.72,
+      "step": 95
+    },
+    {
+      "epoch": 0.04882700743848942,
+      "grad_norm": 3.962436008839197,
+      "learning_rate": 6.508474576271187e-06,
+      "loss": 0.769,
+      "step": 96
+    },
+    {
+      "epoch": 0.04933562209930701,
+      "grad_norm": 4.715612720147406,
+      "learning_rate": 6.576271186440678e-06,
+      "loss": 0.7497,
+      "step": 97
+    },
+    {
+      "epoch": 0.04984423676012461,
+      "grad_norm": 4.812403888485783,
+      "learning_rate": 6.64406779661017e-06,
+      "loss": 0.7435,
+      "step": 98
+    },
+    {
+      "epoch": 0.05035285142094221,
+      "grad_norm": 5.1908604411745625,
+      "learning_rate": 6.7118644067796615e-06,
+      "loss": 0.8667,
+      "step": 99
+    },
+    {
+      "epoch": 0.05086146608175981,
+      "grad_norm": 4.560037653918152,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 0.7572,
+      "step": 100
+    },
+    {
+      "epoch": 0.05086146608175981,
+      "eval_loss": 0.7078994512557983,
+      "eval_runtime": 171.0591,
+      "eval_samples_per_second": 19.362,
+      "eval_steps_per_second": 9.681,
+      "step": 100
+    },
+    {
+      "epoch": 0.05137008074257741,
+      "grad_norm": 4.366891595031732,
+      "learning_rate": 6.8474576271186445e-06,
+      "loss": 0.6863,
+      "step": 101
+    },
+    {
+      "epoch": 0.051878695403395,
+      "grad_norm": 4.320529483913089,
+      "learning_rate": 6.915254237288136e-06,
+      "loss": 0.8277,
+      "step": 102
+    },
+    {
+      "epoch": 0.0523873100642126,
+      "grad_norm": 4.507722374396927,
+      "learning_rate": 6.9830508474576275e-06,
+      "loss": 0.6911,
+      "step": 103
+    },
+    {
+      "epoch": 0.0528959247250302,
+      "grad_norm": 4.384991598784556,
+      "learning_rate": 7.0508474576271195e-06,
+      "loss": 0.7145,
+      "step": 104
+    },
+    {
+      "epoch": 0.0534045393858478,
+      "grad_norm": 4.804040384521479,
+      "learning_rate": 7.1186440677966106e-06,
+      "loss": 0.8164,
+      "step": 105
+    },
+    {
+      "epoch": 0.0539131540466654,
+      "grad_norm": 4.820900854870408,
+      "learning_rate": 7.1864406779661025e-06,
+      "loss": 0.8034,
+      "step": 106
+    },
+    {
+      "epoch": 0.05442176870748299,
+      "grad_norm": 4.4498523130448,
+      "learning_rate": 7.2542372881355936e-06,
+      "loss": 0.8495,
+      "step": 107
+    },
+    {
+      "epoch": 0.05493038336830059,
+      "grad_norm": 3.912498786121704,
+      "learning_rate": 7.3220338983050855e-06,
+      "loss": 0.7381,
+      "step": 108
+    },
+    {
+      "epoch": 0.05543899802911819,
+      "grad_norm": 4.793123548972308,
+      "learning_rate": 7.3898305084745766e-06,
+      "loss": 0.8694,
+      "step": 109
+    },
+    {
+      "epoch": 0.05594761268993579,
+      "grad_norm": 4.046815378346397,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 0.8067,
+      "step": 110
+    },
+    {
+      "epoch": 0.05645622735075338,
+      "grad_norm": 5.161610169652202,
+      "learning_rate": 7.52542372881356e-06,
+      "loss": 0.7785,
+      "step": 111
+    },
+    {
+      "epoch": 0.05696484201157098,
+      "grad_norm": 4.343080976820007,
+      "learning_rate": 7.5932203389830515e-06,
+      "loss": 0.7902,
+      "step": 112
+    },
+    {
+      "epoch": 0.05747345667238858,
+      "grad_norm": 5.272682237582275,
+      "learning_rate": 7.661016949152543e-06,
+      "loss": 0.7834,
+      "step": 113
+    },
+    {
+      "epoch": 0.05798207133320618,
+      "grad_norm": 4.846341528885435,
+      "learning_rate": 7.728813559322035e-06,
+      "loss": 0.7032,
+      "step": 114
+    },
+    {
+      "epoch": 0.05849068599402378,
+      "grad_norm": 4.68446801356017,
+      "learning_rate": 7.796610169491526e-06,
+      "loss": 0.7598,
+      "step": 115
+    },
+    {
+      "epoch": 0.05899930065484137,
+      "grad_norm": 4.635617826861782,
+      "learning_rate": 7.864406779661017e-06,
+      "loss": 0.8313,
+      "step": 116
+    },
+    {
+      "epoch": 0.05950791531565897,
+      "grad_norm": 4.725330539020835,
+      "learning_rate": 7.93220338983051e-06,
+      "loss": 0.8386,
+      "step": 117
+    },
+    {
+      "epoch": 0.06001652997647657,
+      "grad_norm": 3.6853732425481205,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6231,
+      "step": 118
+    },
+    {
+      "epoch": 0.06052514463729417,
+      "grad_norm": 4.676375552900727,
+      "learning_rate": 8.067796610169492e-06,
+      "loss": 0.7657,
+      "step": 119
+    },
+    {
+      "epoch": 0.06103375929811177,
+      "grad_norm": 4.0987071534314135,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 0.6752,
+      "step": 120
+    },
+    {
+      "epoch": 0.061542373958929364,
+      "grad_norm": 3.8270380462198643,
+      "learning_rate": 8.203389830508475e-06,
+      "loss": 0.7783,
+      "step": 121
+    },
+    {
+      "epoch": 0.06205098861974696,
+      "grad_norm": 3.919935377522522,
+      "learning_rate": 8.271186440677966e-06,
+      "loss": 0.7493,
+      "step": 122
+    },
+    {
+      "epoch": 0.06255960328056456,
+      "grad_norm": 4.105697552868438,
+      "learning_rate": 8.338983050847458e-06,
+      "loss": 0.6531,
+      "step": 123
+    },
+    {
+      "epoch": 0.06306821794138216,
+      "grad_norm": 4.69018721802624,
+      "learning_rate": 8.40677966101695e-06,
+      "loss": 0.8303,
+      "step": 124
+    },
+    {
+      "epoch": 0.06357683260219976,
+      "grad_norm": 4.3664340493557265,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.7988,
+      "step": 125
+    },
+    {
+      "epoch": 0.06408544726301736,
+      "grad_norm": 4.130506827819449,
+      "learning_rate": 8.542372881355933e-06,
+      "loss": 0.8247,
+      "step": 126
+    },
+    {
+      "epoch": 0.06459406192383496,
+      "grad_norm": 4.510534671123722,
+      "learning_rate": 8.610169491525424e-06,
+      "loss": 0.8166,
+      "step": 127
+    },
+    {
+      "epoch": 0.06510267658465255,
+      "grad_norm": 3.688839555956898,
+      "learning_rate": 8.677966101694915e-06,
+      "loss": 0.6947,
+      "step": 128
+    },
+    {
+      "epoch": 0.06561129124547015,
+      "grad_norm": 4.439157955294087,
+      "learning_rate": 8.745762711864407e-06,
+      "loss": 0.7284,
+      "step": 129
+    },
+    {
+      "epoch": 0.06611990590628775,
+      "grad_norm": 4.452435142220346,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 0.9075,
+      "step": 130
+    },
+    {
+      "epoch": 0.06662852056710535,
+      "grad_norm": 3.9543095800355266,
+      "learning_rate": 8.88135593220339e-06,
+      "loss": 0.8084,
+      "step": 131
+    },
+    {
+      "epoch": 0.06713713522792294,
+      "grad_norm": 4.052750141159092,
+      "learning_rate": 8.949152542372881e-06,
+      "loss": 0.7161,
+      "step": 132
+    },
+    {
+      "epoch": 0.06764574988874054,
+      "grad_norm": 4.048450767056609,
+      "learning_rate": 9.016949152542374e-06,
+      "loss": 0.7093,
+      "step": 133
+    },
+    {
+      "epoch": 0.06815436454955814,
+      "grad_norm": 4.399168343408497,
+      "learning_rate": 9.084745762711865e-06,
+      "loss": 0.7191,
+      "step": 134
+    },
+    {
+      "epoch": 0.06866297921037574,
+      "grad_norm": 4.06998563834241,
+      "learning_rate": 9.152542372881356e-06,
+      "loss": 0.7436,
+      "step": 135
+    },
+    {
+      "epoch": 0.06917159387119334,
+      "grad_norm": 3.6909015290284923,
+      "learning_rate": 9.220338983050847e-06,
+      "loss": 0.6975,
+      "step": 136
+    },
+    {
+      "epoch": 0.06968020853201093,
+      "grad_norm": 4.264366998420375,
+      "learning_rate": 9.28813559322034e-06,
+      "loss": 0.7714,
+      "step": 137
+    },
+    {
+      "epoch": 0.07018882319282853,
+      "grad_norm": 3.928436504870956,
+      "learning_rate": 9.355932203389831e-06,
+      "loss": 0.7723,
+      "step": 138
+    },
+    {
+      "epoch": 0.07069743785364613,
+      "grad_norm": 5.698881059542351,
+      "learning_rate": 9.423728813559322e-06,
+      "loss": 0.8494,
+      "step": 139
+    },
+    {
+      "epoch": 0.07120605251446373,
+      "grad_norm": 3.838089411200466,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 0.7534,
+      "step": 140
+    },
+    {
+      "epoch": 0.07171466717528133,
+      "grad_norm": 4.2536887821766225,
+      "learning_rate": 9.559322033898306e-06,
+      "loss": 0.814,
+      "step": 141
+    },
+    {
+      "epoch": 0.07222328183609893,
+      "grad_norm": 5.532842374568545,
+      "learning_rate": 9.627118644067797e-06,
+      "loss": 0.7301,
+      "step": 142
+    },
+    {
+      "epoch": 0.07273189649691653,
+      "grad_norm": 4.033824719427634,
+      "learning_rate": 9.69491525423729e-06,
+      "loss": 0.6852,
+      "step": 143
+    },
+    {
+      "epoch": 0.07324051115773413,
+      "grad_norm": 3.9829625636818893,
+      "learning_rate": 9.762711864406781e-06,
+      "loss": 0.7516,
+      "step": 144
+    },
+    {
+      "epoch": 0.07374912581855173,
+      "grad_norm": 4.171651884249137,
+      "learning_rate": 9.830508474576272e-06,
+      "loss": 0.7842,
+      "step": 145
+    },
+    {
+      "epoch": 0.07425774047936932,
+      "grad_norm": 4.686778712409695,
+      "learning_rate": 9.898305084745763e-06,
+      "loss": 0.8184,
+      "step": 146
+    },
+    {
+      "epoch": 0.07476635514018691,
+      "grad_norm": 3.9068189282955985,
+      "learning_rate": 9.966101694915256e-06,
+      "loss": 0.8062,
+      "step": 147
+    },
+    {
+      "epoch": 0.07527496980100451,
+      "grad_norm": 4.01618543936886,
+      "learning_rate": 1.0033898305084746e-05,
+      "loss": 0.8141,
+      "step": 148
+    },
+    {
+      "epoch": 0.07578358446182211,
+      "grad_norm": 5.070193249632902,
+      "learning_rate": 1.0101694915254238e-05,
+      "loss": 0.7128,
+      "step": 149
+    },
+    {
+      "epoch": 0.07629219912263971,
+      "grad_norm": 4.172665867511384,
+      "learning_rate": 1.016949152542373e-05,
+      "loss": 0.8366,
+      "step": 150
+    },
+    {
+      "epoch": 0.07680081378345731,
+      "grad_norm": 4.197405358815827,
+      "learning_rate": 1.0237288135593222e-05,
+      "loss": 0.8366,
+      "step": 151
+    },
+    {
+      "epoch": 0.07730942844427491,
+      "grad_norm": 4.068600812717093,
+      "learning_rate": 1.0305084745762712e-05,
+      "loss": 0.8288,
+      "step": 152
+    },
+    {
+      "epoch": 0.07781804310509251,
+      "grad_norm": 4.012004719526989,
+      "learning_rate": 1.0372881355932204e-05,
+      "loss": 0.7192,
+      "step": 153
+    },
+    {
+      "epoch": 0.07832665776591011,
+      "grad_norm": 4.731187408315437,
+      "learning_rate": 1.0440677966101695e-05,
+      "loss": 0.7642,
+      "step": 154
+    },
+    {
+      "epoch": 0.0788352724267277,
+      "grad_norm": 4.185459835626729,
+      "learning_rate": 1.0508474576271188e-05,
+      "loss": 0.8959,
+      "step": 155
+    },
+    {
+      "epoch": 0.07934388708754529,
+      "grad_norm": 3.9225598227495926,
+      "learning_rate": 1.057627118644068e-05,
+      "loss": 0.8711,
+      "step": 156
+    },
+    {
+      "epoch": 0.07985250174836289,
+      "grad_norm": 3.8808071553403574,
+      "learning_rate": 1.0644067796610172e-05,
+      "loss": 0.7376,
+      "step": 157
+    },
+    {
+      "epoch": 0.08036111640918049,
+      "grad_norm": 4.060278417126708,
+      "learning_rate": 1.0711864406779661e-05,
+      "loss": 0.8525,
+      "step": 158
+    },
+    {
+      "epoch": 0.08086973106999809,
+      "grad_norm": 4.108580769998749,
+      "learning_rate": 1.0779661016949154e-05,
+      "loss": 0.8342,
+      "step": 159
+    },
+    {
+      "epoch": 0.08137834573081569,
+      "grad_norm": 4.516160345847918,
+      "learning_rate": 1.0847457627118645e-05,
+      "loss": 0.7887,
+      "step": 160
+    },
+    {
+      "epoch": 0.08188696039163329,
+      "grad_norm": 4.062955155917798,
+      "learning_rate": 1.0915254237288135e-05,
+      "loss": 0.7143,
+      "step": 161
+    },
+    {
+      "epoch": 0.08239557505245089,
+      "grad_norm": 4.4475484631596,
+      "learning_rate": 1.0983050847457627e-05,
+      "loss": 0.8392,
+      "step": 162
+    },
+    {
+      "epoch": 0.08290418971326849,
+      "grad_norm": 3.505605568372029,
+      "learning_rate": 1.1050847457627118e-05,
+      "loss": 0.8694,
+      "step": 163
+    },
+    {
+      "epoch": 0.08341280437408609,
+      "grad_norm": 4.0757638227116875,
+      "learning_rate": 1.1118644067796611e-05,
+      "loss": 0.6307,
+      "step": 164
+    },
+    {
+      "epoch": 0.08392141903490367,
+      "grad_norm": 3.5713343253643157,
+      "learning_rate": 1.1186440677966102e-05,
+      "loss": 0.7734,
+      "step": 165
+    },
+    {
+      "epoch": 0.08443003369572127,
+      "grad_norm": 3.63172814377711,
+      "learning_rate": 1.1254237288135595e-05,
+      "loss": 0.7519,
+      "step": 166
+    },
+    {
+      "epoch": 0.08493864835653887,
+      "grad_norm": 4.3366313757158075,
+      "learning_rate": 1.1322033898305084e-05,
+      "loss": 0.8086,
+      "step": 167
+    },
+    {
+      "epoch": 0.08544726301735647,
+      "grad_norm": 4.095775307095219,
+      "learning_rate": 1.1389830508474577e-05,
+      "loss": 0.8051,
+      "step": 168
+    },
+    {
+      "epoch": 0.08595587767817407,
+      "grad_norm": 4.168687610073566,
+      "learning_rate": 1.1457627118644068e-05,
+      "loss": 0.7298,
+      "step": 169
+    },
+    {
+      "epoch": 0.08646449233899167,
+      "grad_norm": 4.491021205596363,
+      "learning_rate": 1.1525423728813561e-05,
+      "loss": 0.9365,
+      "step": 170
+    },
+    {
+      "epoch": 0.08697310699980927,
+      "grad_norm": 4.178028611731175,
+      "learning_rate": 1.159322033898305e-05,
+      "loss": 0.8835,
+      "step": 171
+    },
+    {
+      "epoch": 0.08748172166062687,
+      "grad_norm": 3.933825276748244,
+      "learning_rate": 1.1661016949152543e-05,
+      "loss": 0.8416,
+      "step": 172
+    },
+    {
+      "epoch": 0.08799033632144447,
+      "grad_norm": 4.0755031532497235,
+      "learning_rate": 1.1728813559322034e-05,
+      "loss": 0.6932,
+      "step": 173
+    },
+    {
+      "epoch": 0.08849895098226207,
+      "grad_norm": 7.8653435009191535,
+      "learning_rate": 1.1796610169491527e-05,
+      "loss": 0.7699,
+      "step": 174
+    },
+    {
+      "epoch": 0.08900756564307966,
+      "grad_norm": 4.447413800987256,
+      "learning_rate": 1.1864406779661018e-05,
+      "loss": 0.9491,
+      "step": 175
+    },
+    {
+      "epoch": 0.08951618030389726,
+      "grad_norm": 4.991718300035977,
+      "learning_rate": 1.1932203389830511e-05,
+      "loss": 0.7746,
+      "step": 176
+    },
+    {
+      "epoch": 0.09002479496471485,
+      "grad_norm": 3.991183578363838,
+      "learning_rate": 1.2e-05,
+      "loss": 0.8822,
+      "step": 177
+    },
+    {
+      "epoch": 0.09053340962553245,
+      "grad_norm": 3.8797017767432647,
+      "learning_rate": 1.2067796610169493e-05,
+      "loss": 0.8105,
+      "step": 178
+    },
+    {
+      "epoch": 0.09104202428635005,
+      "grad_norm": 3.9280790218554564,
+      "learning_rate": 1.2135593220338984e-05,
+      "loss": 0.8262,
+      "step": 179
+    },
+    {
+      "epoch": 0.09155063894716765,
+      "grad_norm": 3.686446734508338,
+      "learning_rate": 1.2203389830508477e-05,
+      "loss": 0.8326,
+      "step": 180
+    },
+    {
+      "epoch": 0.09205925360798525,
+      "grad_norm": 3.9583266944160336,
+      "learning_rate": 1.2271186440677966e-05,
+      "loss": 0.9054,
+      "step": 181
+    },
+    {
+      "epoch": 0.09256786826880285,
+      "grad_norm": 3.4753229224466615,
+      "learning_rate": 1.2338983050847459e-05,
+      "loss": 0.8309,
+      "step": 182
+    },
+    {
+      "epoch": 0.09307648292962045,
+      "grad_norm": 4.337069856713931,
+      "learning_rate": 1.240677966101695e-05,
+      "loss": 0.8226,
+      "step": 183
+    },
+    {
+      "epoch": 0.09358509759043804,
+      "grad_norm": 4.316046983776023,
+      "learning_rate": 1.2474576271186443e-05,
+      "loss": 0.7765,
+      "step": 184
+    },
+    {
+      "epoch": 0.09409371225125564,
+      "grad_norm": 4.397013188015921,
+      "learning_rate": 1.2542372881355932e-05,
+      "loss": 0.9153,
+      "step": 185
+    },
+    {
+      "epoch": 0.09460232691207324,
+      "grad_norm": 4.21434819911789,
+      "learning_rate": 1.2610169491525425e-05,
+      "loss": 0.9158,
+      "step": 186
+    },
+    {
+      "epoch": 0.09511094157289084,
+      "grad_norm": 3.9445771089072696,
+      "learning_rate": 1.2677966101694916e-05,
+      "loss": 0.7612,
+      "step": 187
+    },
+    {
+      "epoch": 0.09561955623370844,
+      "grad_norm": 3.6921729527855613,
+      "learning_rate": 1.2745762711864407e-05,
+      "loss": 0.7356,
+      "step": 188
+    },
+    {
+      "epoch": 0.09612817089452604,
+      "grad_norm": 4.210506721501183,
+      "learning_rate": 1.28135593220339e-05,
+      "loss": 0.8663,
+      "step": 189
+    },
+    {
+      "epoch": 0.09663678555534364,
+      "grad_norm": 3.6756877177841747,
+      "learning_rate": 1.288135593220339e-05,
+      "loss": 0.8152,
+      "step": 190
+    },
+    {
+      "epoch": 0.09714540021616123,
+      "grad_norm": 4.36534360881226,
+      "learning_rate": 1.2949152542372882e-05,
+      "loss": 0.8514,
+      "step": 191
+    },
+    {
+      "epoch": 0.09765401487697883,
+      "grad_norm": 4.0558245466607525,
+      "learning_rate": 1.3016949152542373e-05,
+      "loss": 0.794,
+      "step": 192
+    },
+    {
+      "epoch": 0.09816262953779643,
+      "grad_norm": 3.8317292353989005,
+      "learning_rate": 1.3084745762711866e-05,
+      "loss": 0.7648,
+      "step": 193
+    },
+    {
+      "epoch": 0.09867124419861402,
+      "grad_norm": 3.727204785288113,
+      "learning_rate": 1.3152542372881355e-05,
+      "loss": 0.7363,
+      "step": 194
+    },
+    {
+      "epoch": 0.09917985885943162,
+      "grad_norm": 4.279456215298531,
+      "learning_rate": 1.3220338983050848e-05,
+      "loss": 0.813,
+      "step": 195
+    },
+    {
+      "epoch": 0.09968847352024922,
+      "grad_norm": 3.4494909436148062,
+      "learning_rate": 1.328813559322034e-05,
+      "loss": 0.7542,
+      "step": 196
+    },
+    {
+      "epoch": 0.10019708818106682,
+      "grad_norm": 4.128331082070981,
+      "learning_rate": 1.3355932203389832e-05,
+      "loss": 0.842,
+      "step": 197
+    },
+    {
+      "epoch": 0.10070570284188442,
+      "grad_norm": 3.7166749951279274,
+      "learning_rate": 1.3423728813559323e-05,
+      "loss": 0.7208,
+      "step": 198
+    },
+    {
+      "epoch": 0.10121431750270202,
+      "grad_norm": 3.6644428763364707,
+      "learning_rate": 1.3491525423728816e-05,
+      "loss": 0.7632,
+      "step": 199
+    },
+    {
+      "epoch": 0.10172293216351962,
+      "grad_norm": 3.7263834848331965,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 0.7494,
+      "step": 200
+    },
+    {
+      "epoch": 0.10172293216351962,
+      "eval_loss": 0.7702630162239075,
+      "eval_runtime": 169.7855,
+      "eval_samples_per_second": 19.507,
+      "eval_steps_per_second": 9.753,
+      "step": 200
+    },
+    {
+      "epoch": 0.10223154682433722,
+      "grad_norm": 4.446765521398101,
+      "learning_rate": 1.3627118644067798e-05,
+      "loss": 0.8193,
+      "step": 201
+    },
+    {
+      "epoch": 0.10274016148515482,
+      "grad_norm": 3.4646532765430016,
+      "learning_rate": 1.3694915254237289e-05,
+      "loss": 0.6545,
+      "step": 202
+    },
+    {
+      "epoch": 0.1032487761459724,
+      "grad_norm": 3.536227473877957,
+      "learning_rate": 1.3762711864406782e-05,
+      "loss": 0.8559,
+      "step": 203
+    },
+    {
+      "epoch": 0.10375739080679,
+      "grad_norm": 4.147071282945361,
+      "learning_rate": 1.3830508474576271e-05,
+      "loss": 0.8173,
+      "step": 204
+    },
+    {
+      "epoch": 0.1042660054676076,
+      "grad_norm": 3.9860014344713814,
+      "learning_rate": 1.3898305084745764e-05,
+      "loss": 0.8205,
+      "step": 205
+    },
+    {
+      "epoch": 0.1047746201284252,
+      "grad_norm": 3.5944469273706185,
+      "learning_rate": 1.3966101694915255e-05,
+      "loss": 0.7651,
+      "step": 206
+    },
+    {
+      "epoch": 0.1052832347892428,
+      "grad_norm": 3.362175771971308,
+      "learning_rate": 1.4033898305084748e-05,
+      "loss": 0.846,
+      "step": 207
+    },
+    {
+      "epoch": 0.1057918494500604,
+      "grad_norm": 3.7059727273957606,
+      "learning_rate": 1.4101694915254239e-05,
+      "loss": 0.7987,
+      "step": 208
+    },
+    {
+      "epoch": 0.106300464110878,
+      "grad_norm": 3.9645867325485393,
+      "learning_rate": 1.416949152542373e-05,
+      "loss": 0.8246,
+      "step": 209
+    },
+    {
+      "epoch": 0.1068090787716956,
+      "grad_norm": 4.1210023408849334,
+      "learning_rate": 1.4237288135593221e-05,
+      "loss": 0.7855,
+      "step": 210
+    },
+    {
+      "epoch": 0.1073176934325132,
+      "grad_norm": 4.068543853459664,
+      "learning_rate": 1.4305084745762714e-05,
+      "loss": 0.8027,
+      "step": 211
+    },
+    {
+      "epoch": 0.1078263080933308,
+      "grad_norm": 3.852967785476594,
+      "learning_rate": 1.4372881355932205e-05,
+      "loss": 0.9203,
+      "step": 212
+    },
+    {
+      "epoch": 0.10833492275414838,
+      "grad_norm": 3.8125325185921057,
+      "learning_rate": 1.4440677966101698e-05,
+      "loss": 0.8968,
+      "step": 213
+    },
+    {
+      "epoch": 0.10884353741496598,
+      "grad_norm": 3.4044034659965834,
+      "learning_rate": 1.4508474576271187e-05,
+      "loss": 0.8244,
+      "step": 214
+    },
+    {
+      "epoch": 0.10935215207578358,
+      "grad_norm": 3.7167943091491407,
+      "learning_rate": 1.4576271186440678e-05,
+      "loss": 1.0188,
+      "step": 215
+    },
+    {
+      "epoch": 0.10986076673660118,
+      "grad_norm": 3.8517761790213574,
+      "learning_rate": 1.4644067796610171e-05,
+      "loss": 0.8534,
+      "step": 216
+    },
+    {
+      "epoch": 0.11036938139741878,
+      "grad_norm": 3.677648282980653,
+      "learning_rate": 1.4711864406779662e-05,
+      "loss": 0.9203,
+      "step": 217
+    },
+    {
+      "epoch": 0.11087799605823638,
+      "grad_norm": 3.553932193536461,
+      "learning_rate": 1.4779661016949153e-05,
+      "loss": 0.7549,
+      "step": 218
+    },
+    {
+      "epoch": 0.11138661071905398,
+      "grad_norm": 3.8007220335186602,
+      "learning_rate": 1.4847457627118644e-05,
+      "loss": 0.7409,
+      "step": 219
+    },
+    {
+      "epoch": 0.11189522537987158,
+      "grad_norm": 3.7937958716144067,
+      "learning_rate": 1.4915254237288137e-05,
+      "loss": 0.7806,
+      "step": 220
+    },
+    {
+      "epoch": 0.11240384004068918,
+      "grad_norm": 3.3836925322180043,
+      "learning_rate": 1.4983050847457628e-05,
+      "loss": 0.7843,
+      "step": 221
+    },
+    {
+      "epoch": 0.11291245470150676,
+      "grad_norm": 3.6142403974013018,
+      "learning_rate": 1.505084745762712e-05,
+      "loss": 0.7697,
+      "step": 222
+    },
+    {
+      "epoch": 0.11342106936232436,
+      "grad_norm": 4.044886745003774,
+      "learning_rate": 1.511864406779661e-05,
+      "loss": 0.6669,
+      "step": 223
+    },
+    {
+      "epoch": 0.11392968402314196,
+      "grad_norm": 3.6483958076529084,
+      "learning_rate": 1.5186440677966103e-05,
+      "loss": 0.8932,
+      "step": 224
+    },
+    {
+      "epoch": 0.11443829868395956,
+      "grad_norm": 3.5844666367035853,
+      "learning_rate": 1.5254237288135594e-05,
+      "loss": 0.9159,
+      "step": 225
+    },
+    {
+      "epoch": 0.11494691334477716,
+      "grad_norm": 3.369304796724752,
+      "learning_rate": 1.5322033898305085e-05,
+      "loss": 0.7841,
+      "step": 226
+    },
+    {
+      "epoch": 0.11545552800559476,
+      "grad_norm": 3.466644610432616,
+      "learning_rate": 1.5389830508474578e-05,
+      "loss": 0.8359,
+      "step": 227
+    },
+    {
+      "epoch": 0.11596414266641236,
+      "grad_norm": 3.181618743690104,
+      "learning_rate": 1.545762711864407e-05,
+      "loss": 0.8516,
+      "step": 228
+    },
+    {
+      "epoch": 0.11647275732722996,
+      "grad_norm": 3.8912370192730656,
+      "learning_rate": 1.552542372881356e-05,
+      "loss": 0.7626,
+      "step": 229
+    },
+    {
+      "epoch": 0.11698137198804756,
+      "grad_norm": 3.7616514395993086,
+      "learning_rate": 1.5593220338983053e-05,
+      "loss": 0.8045,
+      "step": 230
+    },
+    {
+      "epoch": 0.11748998664886515,
+      "grad_norm": 3.2746372983011116,
+      "learning_rate": 1.5661016949152542e-05,
+      "loss": 0.7691,
+      "step": 231
+    },
+    {
+      "epoch": 0.11799860130968275,
+      "grad_norm": 3.7413320180746075,
+      "learning_rate": 1.5728813559322035e-05,
+      "loss": 0.7936,
+      "step": 232
+    },
+    {
+      "epoch": 0.11850721597050035,
+      "grad_norm": 3.5285812691087446,
+      "learning_rate": 1.5796610169491528e-05,
+      "loss": 0.8233,
+      "step": 233
+    },
+    {
+      "epoch": 0.11901583063131795,
+      "grad_norm": 4.278102636731633,
+      "learning_rate": 1.586440677966102e-05,
+      "loss": 0.9416,
+      "step": 234
+    },
+    {
+      "epoch": 0.11952444529213554,
+      "grad_norm": 3.8269721338809064,
+      "learning_rate": 1.593220338983051e-05,
+      "loss": 0.9021,
+      "step": 235
+    },
+    {
+      "epoch": 0.12003305995295314,
+      "grad_norm": 3.8652663265717653,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7652,
+      "step": 236
+    },
+    {
+      "epoch": 0.12054167461377074,
+      "grad_norm": 3.5056972455547606,
+      "learning_rate": 1.6067796610169492e-05,
+      "loss": 0.812,
+      "step": 237
+    },
+    {
+      "epoch": 0.12105028927458834,
+      "grad_norm": 3.628887591195976,
+      "learning_rate": 1.6135593220338985e-05,
+      "loss": 0.8018,
+      "step": 238
+    },
+    {
+      "epoch": 0.12155890393540594,
+      "grad_norm": 3.625394076269772,
+      "learning_rate": 1.6203389830508474e-05,
+      "loss": 0.9188,
+      "step": 239
+    },
+    {
+      "epoch": 0.12206751859622354,
+      "grad_norm": 3.522747640466959,
+      "learning_rate": 1.6271186440677967e-05,
+      "loss": 0.8247,
+      "step": 240
+    },
+    {
+      "epoch": 0.12257613325704113,
+      "grad_norm": 3.825138967146286,
+      "learning_rate": 1.633898305084746e-05,
+      "loss": 0.7834,
+      "step": 241
+    },
+    {
+      "epoch": 0.12308474791785873,
+      "grad_norm": 4.198348156482736,
+      "learning_rate": 1.640677966101695e-05,
+      "loss": 0.9066,
+      "step": 242
+    },
+    {
+      "epoch": 0.12359336257867633,
+      "grad_norm": 3.3410805001812522,
+      "learning_rate": 1.6474576271186442e-05,
+      "loss": 0.9057,
+      "step": 243
+    },
+    {
+      "epoch": 0.12410197723949393,
+      "grad_norm": 2.839526383035744,
+      "learning_rate": 1.654237288135593e-05,
+      "loss": 0.7769,
+      "step": 244
+    },
+    {
+      "epoch": 0.12461059190031153,
+      "grad_norm": 3.329081653055383,
+      "learning_rate": 1.6610169491525424e-05,
+      "loss": 0.7415,
+      "step": 245
+    },
+    {
+      "epoch": 0.12511920656112913,
+      "grad_norm": 3.325746214460088,
+      "learning_rate": 1.6677966101694917e-05,
+      "loss": 0.9038,
+      "step": 246
+    },
+    {
+      "epoch": 0.1256278212219467,
+      "grad_norm": 3.7563603186632575,
+      "learning_rate": 1.674576271186441e-05,
+      "loss": 0.7783,
+      "step": 247
+    },
+    {
+      "epoch": 0.12613643588276433,
+      "grad_norm": 3.642612684152413,
+      "learning_rate": 1.68135593220339e-05,
+      "loss": 0.7479,
+      "step": 248
+    },
+    {
+      "epoch": 0.1266450505435819,
+      "grad_norm": 3.707419565039516,
+      "learning_rate": 1.6881355932203392e-05,
+      "loss": 0.8751,
+      "step": 249
+    },
+    {
+      "epoch": 0.12715366520439952,
+      "grad_norm": 4.2549060105590355,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 0.7859,
+      "step": 250
+    },
+    {
+      "epoch": 0.1276622798652171,
+      "grad_norm": 3.222436515946311,
+      "learning_rate": 1.7016949152542374e-05,
+      "loss": 0.8437,
+      "step": 251
+    },
+    {
+      "epoch": 0.12817089452603472,
+      "grad_norm": 3.6939051974719326,
+      "learning_rate": 1.7084745762711867e-05,
+      "loss": 0.6987,
+      "step": 252
+    },
+    {
+      "epoch": 0.1286795091868523,
+      "grad_norm": 3.962830702402428,
+      "learning_rate": 1.715254237288136e-05,
+      "loss": 0.8113,
+      "step": 253
+    },
+    {
+      "epoch": 0.12918812384766992,
+      "grad_norm": 3.511260175625705,
+      "learning_rate": 1.722033898305085e-05,
+      "loss": 0.7916,
+      "step": 254
+    },
+    {
+      "epoch": 0.1296967385084875,
+      "grad_norm": 4.246721686261816,
+      "learning_rate": 1.728813559322034e-05,
+      "loss": 0.8698,
+      "step": 255
+    },
+    {
+      "epoch": 0.1302053531693051,
+      "grad_norm": 3.274442168040355,
+      "learning_rate": 1.735593220338983e-05,
+      "loss": 0.8325,
+      "step": 256
+    },
+    {
+      "epoch": 0.1307139678301227,
+      "grad_norm": 3.1238295842871264,
+      "learning_rate": 1.7423728813559324e-05,
+      "loss": 0.7506,
+      "step": 257
+    },
+    {
+      "epoch": 0.1312225824909403,
+      "grad_norm": 3.687681419952911,
+      "learning_rate": 1.7491525423728813e-05,
+      "loss": 0.793,
+      "step": 258
+    },
+    {
+      "epoch": 0.1317311971517579,
+      "grad_norm": 3.843974285250106,
+      "learning_rate": 1.7559322033898306e-05,
+      "loss": 0.8038,
+      "step": 259
+    },
+    {
+      "epoch": 0.1322398118125755,
+      "grad_norm": 3.3388798979265966,
+      "learning_rate": 1.76271186440678e-05,
+      "loss": 0.8593,
+      "step": 260
+    },
+    {
+      "epoch": 0.1327484264733931,
+      "grad_norm": 3.1693822527159083,
+      "learning_rate": 1.769491525423729e-05,
+      "loss": 0.8214,
+      "step": 261
+    },
+    {
+      "epoch": 0.1332570411342107,
+      "grad_norm": 3.452821606042203,
+      "learning_rate": 1.776271186440678e-05,
+      "loss": 0.7813,
+      "step": 262
+    },
+    {
+      "epoch": 0.1337656557950283,
+      "grad_norm": 3.6484930454902504,
+      "learning_rate": 1.7830508474576274e-05,
+      "loss": 0.8756,
+      "step": 263
+    },
+    {
+      "epoch": 0.1342742704558459,
+      "grad_norm": 4.230624293799108,
+      "learning_rate": 1.7898305084745763e-05,
+      "loss": 0.8753,
+      "step": 264
+    },
+    {
+      "epoch": 0.13478288511666348,
+      "grad_norm": 3.7298231914433573,
+      "learning_rate": 1.7966101694915256e-05,
+      "loss": 0.8376,
+      "step": 265
+    },
+    {
+      "epoch": 0.1352914997774811,
+      "grad_norm": 4.989354244486709,
+      "learning_rate": 1.803389830508475e-05,
+      "loss": 0.8082,
+      "step": 266
+    },
+    {
+      "epoch": 0.13580011443829867,
+      "grad_norm": 3.91898996692258,
+      "learning_rate": 1.810169491525424e-05,
+      "loss": 0.8325,
+      "step": 267
+    },
+    {
+      "epoch": 0.1363087290991163,
+      "grad_norm": 3.956635857397599,
+      "learning_rate": 1.816949152542373e-05,
+      "loss": 0.8223,
+      "step": 268
+    },
+    {
+      "epoch": 0.13681734375993387,
+      "grad_norm": 3.7418280251649283,
+      "learning_rate": 1.823728813559322e-05,
+      "loss": 0.8913,
+      "step": 269
+    },
+    {
+      "epoch": 0.1373259584207515,
+      "grad_norm": 3.773723836266669,
+      "learning_rate": 1.8305084745762713e-05,
+      "loss": 0.8257,
+      "step": 270
+    },
+    {
+      "epoch": 0.13783457308156907,
+      "grad_norm": 3.515817594550988,
+      "learning_rate": 1.8372881355932202e-05,
+      "loss": 0.9599,
+      "step": 271
+    },
+    {
+      "epoch": 0.1383431877423867,
+      "grad_norm": 3.366040381793967,
+      "learning_rate": 1.8440677966101695e-05,
+      "loss": 0.8584,
+      "step": 272
+    },
+    {
+      "epoch": 0.13885180240320427,
+      "grad_norm": 3.449718192933553,
+      "learning_rate": 1.8508474576271188e-05,
+      "loss": 0.9568,
+      "step": 273
+    },
+    {
+      "epoch": 0.13936041706402186,
+      "grad_norm": 3.712985017992323,
+      "learning_rate": 1.857627118644068e-05,
+      "loss": 0.9225,
+      "step": 274
+    },
+    {
+      "epoch": 0.13986903172483947,
+      "grad_norm": 3.9278609958689543,
+      "learning_rate": 1.864406779661017e-05,
+      "loss": 0.9264,
+      "step": 275
+    },
+    {
+      "epoch": 0.14037764638565706,
+      "grad_norm": 3.745007815995865,
+      "learning_rate": 1.8711864406779663e-05,
+      "loss": 0.9228,
+      "step": 276
+    },
+    {
+      "epoch": 0.14088626104647467,
+      "grad_norm": 3.5605034838620435,
+      "learning_rate": 1.8779661016949152e-05,
+      "loss": 0.8132,
+      "step": 277
+    },
+    {
+      "epoch": 0.14139487570729226,
+      "grad_norm": 3.902048766143788,
+      "learning_rate": 1.8847457627118645e-05,
+      "loss": 0.8576,
+      "step": 278
+    },
+    {
+      "epoch": 0.14190349036810987,
+      "grad_norm": 3.8906698569044385,
+      "learning_rate": 1.8915254237288138e-05,
+      "loss": 0.9478,
+      "step": 279
+    },
+    {
+      "epoch": 0.14241210502892745,
+      "grad_norm": 3.8890188800944956,
+      "learning_rate": 1.898305084745763e-05,
+      "loss": 0.8798,
+      "step": 280
+    },
+    {
+      "epoch": 0.14292071968974507,
+      "grad_norm": 3.956918938898505,
+      "learning_rate": 1.905084745762712e-05,
+      "loss": 0.8214,
+      "step": 281
+    },
+    {
+      "epoch": 0.14342933435056265,
+      "grad_norm": 4.984207581007365,
+      "learning_rate": 1.9118644067796613e-05,
+      "loss": 0.9645,
+      "step": 282
+    },
+    {
+      "epoch": 0.14393794901138024,
+      "grad_norm": 3.768609221756561,
+      "learning_rate": 1.9186440677966102e-05,
+      "loss": 0.839,
+      "step": 283
+    },
+    {
+      "epoch": 0.14444656367219785,
+      "grad_norm": 3.1572558622596536,
+      "learning_rate": 1.9254237288135595e-05,
+      "loss": 0.7925,
+      "step": 284
+    },
+    {
+      "epoch": 0.14495517833301544,
+      "grad_norm": 4.585782569984627,
+      "learning_rate": 1.9322033898305087e-05,
+      "loss": 0.8718,
+      "step": 285
+    },
+    {
+      "epoch": 0.14546379299383305,
+      "grad_norm": 3.768149008839728,
+      "learning_rate": 1.938983050847458e-05,
+      "loss": 0.8975,
+      "step": 286
+    },
+    {
+      "epoch": 0.14597240765465064,
+      "grad_norm": 3.5773943692333474,
+      "learning_rate": 1.945762711864407e-05,
+      "loss": 0.9564,
+      "step": 287
+    },
+    {
+      "epoch": 0.14648102231546825,
+      "grad_norm": 4.118858157692901,
+      "learning_rate": 1.9525423728813562e-05,
+      "loss": 0.7775,
+      "step": 288
+    },
+    {
+      "epoch": 0.14698963697628584,
+      "grad_norm": 3.727901194298626,
+      "learning_rate": 1.9593220338983052e-05,
+      "loss": 0.7863,
+      "step": 289
+    },
+    {
+      "epoch": 0.14749825163710345,
+      "grad_norm": 3.31130456046418,
+      "learning_rate": 1.9661016949152545e-05,
+      "loss": 0.9085,
+      "step": 290
+    },
+    {
+      "epoch": 0.14800686629792104,
+      "grad_norm": 3.201508500261415,
+      "learning_rate": 1.9728813559322034e-05,
+      "loss": 0.9365,
+      "step": 291
+    },
+    {
+      "epoch": 0.14851548095873865,
+      "grad_norm": 3.866069223572578,
+      "learning_rate": 1.9796610169491527e-05,
+      "loss": 0.7752,
+      "step": 292
+    },
+    {
+      "epoch": 0.14902409561955623,
+      "grad_norm": 3.3704583021096353,
+      "learning_rate": 1.986440677966102e-05,
+      "loss": 0.7931,
+      "step": 293
+    },
+    {
+      "epoch": 0.14953271028037382,
+      "grad_norm": 3.7679193242339393,
+      "learning_rate": 1.9932203389830512e-05,
+      "loss": 0.9468,
+      "step": 294
+    },
+    {
+      "epoch": 0.15004132494119143,
+      "grad_norm": 3.163405060459075,
+      "learning_rate": 2e-05,
+      "loss": 0.8179,
+      "step": 295
+    },
+    {
+      "epoch": 0.15054993960200902,
+      "grad_norm": 3.254455007233009,
+      "learning_rate": 1.9999998428087046e-05,
+      "loss": 0.7803,
+      "step": 296
+    },
+    {
+      "epoch": 0.15105855426282663,
+      "grad_norm": 3.540924592907374,
+      "learning_rate": 1.9999993712348673e-05,
+      "loss": 0.7734,
+      "step": 297
+    },
+    {
+      "epoch": 0.15156716892364422,
+      "grad_norm": 3.1545387387922528,
+      "learning_rate": 1.9999985852786365e-05,
+      "loss": 0.7589,
+      "step": 298
+    },
+    {
+      "epoch": 0.15207578358446183,
+      "grad_norm": 3.8594087298073756,
+      "learning_rate": 1.9999974849402594e-05,
+      "loss": 0.9934,
+      "step": 299
+    },
+    {
+      "epoch": 0.15258439824527942,
+      "grad_norm": 3.4423815658242423,
+      "learning_rate": 1.999996070220082e-05,
+      "loss": 0.9351,
+      "step": 300
+    },
+    {
+      "epoch": 0.15258439824527942,
+      "eval_loss": 0.8128167986869812,
+      "eval_runtime": 169.9492,
+      "eval_samples_per_second": 19.488,
+      "eval_steps_per_second": 9.744,
+      "step": 300
+    },
+    {
+      "epoch": 0.15309301290609703,
+      "grad_norm": 3.1842182693619603,
+      "learning_rate": 1.999994341118549e-05,
+      "loss": 0.825,
+      "step": 301
+    },
+    {
+      "epoch": 0.15360162756691462,
+      "grad_norm": 3.259141243939762,
+      "learning_rate": 1.9999922976362032e-05,
+      "loss": 0.8681,
+      "step": 302
+    },
+    {
+      "epoch": 0.1541102422277322,
+      "grad_norm": 3.8088955960115154,
+      "learning_rate": 1.9999899397736882e-05,
+      "loss": 0.807,
+      "step": 303
+    },
+    {
+      "epoch": 0.15461885688854982,
+      "grad_norm": 3.6296971899421346,
+      "learning_rate": 1.999987267531745e-05,
+      "loss": 0.8412,
+      "step": 304
+    },
+    {
+      "epoch": 0.1551274715493674,
+      "grad_norm": 3.65082676028737,
+      "learning_rate": 1.9999842809112132e-05,
+      "loss": 0.829,
+      "step": 305
+    },
+    {
+      "epoch": 0.15563608621018502,
+      "grad_norm": 3.5782449360698836,
+      "learning_rate": 1.9999809799130326e-05,
+      "loss": 0.7962,
+      "step": 306
+    },
+    {
+      "epoch": 0.1561447008710026,
+      "grad_norm": 3.4243974141659557,
+      "learning_rate": 1.99997736453824e-05,
+      "loss": 0.9739,
+      "step": 307
+    },
+    {
+      "epoch": 0.15665331553182021,
+      "grad_norm": 3.0072756597826045,
+      "learning_rate": 1.9999734347879725e-05,
+      "loss": 0.7831,
+      "step": 308
+    },
+    {
+      "epoch": 0.1571619301926378,
+      "grad_norm": 3.8019538474133703,
+      "learning_rate": 1.9999691906634654e-05,
+      "loss": 0.8633,
+      "step": 309
+    },
+    {
+      "epoch": 0.1576705448534554,
+      "grad_norm": 3.2785979679765735,
+      "learning_rate": 1.9999646321660535e-05,
+      "loss": 0.8482,
+      "step": 310
+    },
+    {
+      "epoch": 0.158179159514273,
+      "grad_norm": 3.4823128794987186,
+      "learning_rate": 1.9999597592971692e-05,
+      "loss": 0.8754,
+      "step": 311
+    },
+    {
+      "epoch": 0.15868777417509058,
+      "grad_norm": 3.4150989626484645,
+      "learning_rate": 1.9999545720583448e-05,
+      "loss": 0.7806,
+      "step": 312
+    },
+    {
+      "epoch": 0.1591963888359082,
+      "grad_norm": 3.3914373453081628,
+      "learning_rate": 1.999949070451211e-05,
+      "loss": 0.7581,
+      "step": 313
+    },
+    {
+      "epoch": 0.15970500349672578,
+      "grad_norm": 3.5304391655861025,
+      "learning_rate": 1.9999432544774974e-05,
+      "loss": 0.8095,
+      "step": 314
+    },
+    {
+      "epoch": 0.1602136181575434,
+      "grad_norm": 3.1059302796781645,
+      "learning_rate": 1.9999371241390323e-05,
+      "loss": 0.8442,
+      "step": 315
+    },
+    {
+      "epoch": 0.16072223281836098,
+      "grad_norm": 3.0550338665822276,
+      "learning_rate": 1.9999306794377435e-05,
+      "loss": 0.8283,
+      "step": 316
+    },
+    {
+      "epoch": 0.1612308474791786,
+      "grad_norm": 3.4495433339279105,
+      "learning_rate": 1.9999239203756567e-05,
+      "loss": 0.7824,
+      "step": 317
+    },
+    {
+      "epoch": 0.16173946213999618,
+      "grad_norm": 3.3102387681569847,
+      "learning_rate": 1.9999168469548966e-05,
+      "loss": 0.8283,
+      "step": 318
+    },
+    {
+      "epoch": 0.1622480768008138,
+      "grad_norm": 3.746350101613877,
+      "learning_rate": 1.999909459177687e-05,
+      "loss": 0.7995,
+      "step": 319
+    },
+    {
+      "epoch": 0.16275669146163138,
+      "grad_norm": 3.316368579088905,
+      "learning_rate": 1.9999017570463516e-05,
+      "loss": 0.8463,
+      "step": 320
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 3.264539620391437,
+      "learning_rate": 1.9998937405633105e-05,
+      "loss": 0.8273,
+      "step": 321
+    },
+    {
+      "epoch": 0.16377392078326658,
+      "grad_norm": 3.560322413899894,
+      "learning_rate": 1.999885409731084e-05,
+      "loss": 0.7508,
+      "step": 322
+    },
+    {
+      "epoch": 0.16428253544408417,
+      "grad_norm": 3.664330576380636,
+      "learning_rate": 1.999876764552292e-05,
+      "loss": 0.8927,
+      "step": 323
+    },
+    {
+      "epoch": 0.16479115010490178,
+      "grad_norm": 4.140288256501451,
+      "learning_rate": 1.999867805029652e-05,
+      "loss": 0.8715,
+      "step": 324
+    },
+    {
+      "epoch": 0.16529976476571936,
+      "grad_norm": 3.894583163207164,
+      "learning_rate": 1.9998585311659804e-05,
+      "loss": 0.9358,
+      "step": 325
+    },
+    {
+      "epoch": 0.16580837942653698,
+      "grad_norm": 3.017618735820984,
+      "learning_rate": 1.9998489429641933e-05,
+      "loss": 0.8037,
+      "step": 326
+    },
+    {
+      "epoch": 0.16631699408735456,
+      "grad_norm": 3.27144560378145,
+      "learning_rate": 1.9998390404273045e-05,
+      "loss": 1.0372,
+      "step": 327
+    },
+    {
+      "epoch": 0.16682560874817218,
+      "grad_norm": 3.478199365580011,
+      "learning_rate": 1.9998288235584273e-05,
+      "loss": 0.8516,
+      "step": 328
+    },
+    {
+      "epoch": 0.16733422340898976,
+      "grad_norm": 3.258328696505344,
+      "learning_rate": 1.9998182923607743e-05,
+      "loss": 0.8988,
+      "step": 329
+    },
+    {
+      "epoch": 0.16784283806980735,
+      "grad_norm": 3.108902171620638,
+      "learning_rate": 1.9998074468376555e-05,
+      "loss": 0.7214,
+      "step": 330
+    },
+    {
+      "epoch": 0.16835145273062496,
+      "grad_norm": 3.623433399403741,
+      "learning_rate": 1.999796286992481e-05,
+      "loss": 0.998,
+      "step": 331
+    },
+    {
+      "epoch": 0.16886006739144255,
+      "grad_norm": 3.2496997988042886,
+      "learning_rate": 1.9997848128287593e-05,
+      "loss": 0.8672,
+      "step": 332
+    },
+    {
+      "epoch": 0.16936868205226016,
+      "grad_norm": 3.9455522823173945,
+      "learning_rate": 1.9997730243500975e-05,
+      "loss": 0.7821,
+      "step": 333
+    },
+    {
+      "epoch": 0.16987729671307775,
+      "grad_norm": 2.8712199695895,
+      "learning_rate": 1.999760921560202e-05,
+      "loss": 0.8387,
+      "step": 334
+    },
+    {
+      "epoch": 0.17038591137389536,
+      "grad_norm": 3.590527092750029,
+      "learning_rate": 1.9997485044628772e-05,
+      "loss": 0.8867,
+      "step": 335
+    },
+    {
+      "epoch": 0.17089452603471295,
+      "grad_norm": 2.875447445756167,
+      "learning_rate": 1.999735773062027e-05,
+      "loss": 0.8626,
+      "step": 336
+    },
+    {
+      "epoch": 0.17140314069553056,
+      "grad_norm": 2.8366483107562024,
+      "learning_rate": 1.999722727361654e-05,
+      "loss": 0.8427,
+      "step": 337
+    },
+    {
+      "epoch": 0.17191175535634814,
+      "grad_norm": 3.2568523355096928,
+      "learning_rate": 1.9997093673658598e-05,
+      "loss": 0.8405,
+      "step": 338
+    },
+    {
+      "epoch": 0.17242037001716576,
+      "grad_norm": 3.407474962507749,
+      "learning_rate": 1.9996956930788444e-05,
+      "loss": 0.7317,
+      "step": 339
+    },
+    {
+      "epoch": 0.17292898467798334,
+      "grad_norm": 3.5507024204413176,
+      "learning_rate": 1.9996817045049062e-05,
+      "loss": 0.8707,
+      "step": 340
+    },
+    {
+      "epoch": 0.17343759933880093,
+      "grad_norm": 3.0099308629507844,
+      "learning_rate": 1.999667401648444e-05,
+      "loss": 0.7976,
+      "step": 341
+    },
+    {
+      "epoch": 0.17394621399961854,
+      "grad_norm": 3.2968859920951235,
+      "learning_rate": 1.9996527845139532e-05,
+      "loss": 0.9595,
+      "step": 342
+    },
+    {
+      "epoch": 0.17445482866043613,
+      "grad_norm": 3.673800089605948,
+      "learning_rate": 1.9996378531060303e-05,
+      "loss": 0.873,
+      "step": 343
+    },
+    {
+      "epoch": 0.17496344332125374,
+      "grad_norm": 3.663406931583565,
+      "learning_rate": 1.9996226074293688e-05,
+      "loss": 0.837,
+      "step": 344
+    },
+    {
+      "epoch": 0.17547205798207133,
+      "grad_norm": 2.946214775486658,
+      "learning_rate": 1.9996070474887617e-05,
+      "loss": 0.7396,
+      "step": 345
+    },
+    {
+      "epoch": 0.17598067264288894,
+      "grad_norm": 3.0045028907532165,
+      "learning_rate": 1.999591173289101e-05,
+      "loss": 0.8474,
+      "step": 346
+    },
+    {
+      "epoch": 0.17648928730370653,
+      "grad_norm": 3.0611435357892156,
+      "learning_rate": 1.999574984835377e-05,
+      "loss": 0.9333,
+      "step": 347
+    },
+    {
+      "epoch": 0.17699790196452414,
+      "grad_norm": 3.0305146574047646,
+      "learning_rate": 1.9995584821326793e-05,
+      "loss": 0.8564,
+      "step": 348
+    },
+    {
+      "epoch": 0.17750651662534173,
+      "grad_norm": 3.2940651730156625,
+      "learning_rate": 1.999541665186196e-05,
+      "loss": 0.9178,
+      "step": 349
+    },
+    {
+      "epoch": 0.1780151312861593,
+      "grad_norm": 3.9243612121932774,
+      "learning_rate": 1.9995245340012138e-05,
+      "loss": 0.8038,
+      "step": 350
+    },
+    {
+      "epoch": 0.17852374594697693,
+      "grad_norm": 3.4246229869197373,
+      "learning_rate": 1.9995070885831192e-05,
+      "loss": 0.8713,
+      "step": 351
+    },
+    {
+      "epoch": 0.1790323606077945,
+      "grad_norm": 4.054818976297462,
+      "learning_rate": 1.9994893289373958e-05,
+      "loss": 0.7795,
+      "step": 352
+    },
+    {
+      "epoch": 0.17954097526861212,
+      "grad_norm": 2.83866864308291,
+      "learning_rate": 1.9994712550696275e-05,
+      "loss": 0.86,
+      "step": 353
+    },
+    {
+      "epoch": 0.1800495899294297,
+      "grad_norm": 3.3435583327076484,
+      "learning_rate": 1.9994528669854966e-05,
+      "loss": 0.8624,
+      "step": 354
+    },
+    {
+      "epoch": 0.18055820459024732,
+      "grad_norm": 2.894251677361805,
+      "learning_rate": 1.9994341646907832e-05,
+      "loss": 0.641,
+      "step": 355
+    },
+    {
+      "epoch": 0.1810668192510649,
+      "grad_norm": 3.852376298645072,
+      "learning_rate": 1.9994151481913678e-05,
+      "loss": 0.7826,
+      "step": 356
+    },
+    {
+      "epoch": 0.18157543391188252,
+      "grad_norm": 3.339409979598315,
+      "learning_rate": 1.999395817493228e-05,
+      "loss": 0.8908,
+      "step": 357
+    },
+    {
+      "epoch": 0.1820840485727001,
+      "grad_norm": 3.323308333669876,
+      "learning_rate": 1.999376172602442e-05,
+      "loss": 0.7612,
+      "step": 358
+    },
+    {
+      "epoch": 0.1825926632335177,
+      "grad_norm": 3.2567857180817064,
+      "learning_rate": 1.999356213525185e-05,
+      "loss": 0.8865,
+      "step": 359
+    },
+    {
+      "epoch": 0.1831012778943353,
+      "grad_norm": 3.03557286293727,
+      "learning_rate": 1.9993359402677324e-05,
+      "loss": 0.8189,
+      "step": 360
+    },
+    {
+      "epoch": 0.1836098925551529,
+      "grad_norm": 4.329055190402721,
+      "learning_rate": 1.9993153528364572e-05,
+      "loss": 0.8999,
+      "step": 361
+    },
+    {
+      "epoch": 0.1841185072159705,
+      "grad_norm": 2.9918352920769204,
+      "learning_rate": 1.9992944512378324e-05,
+      "loss": 0.8777,
+      "step": 362
+    },
+    {
+      "epoch": 0.1846271218767881,
+      "grad_norm": 3.34510547579984,
+      "learning_rate": 1.9992732354784287e-05,
+      "loss": 0.9237,
+      "step": 363
+    },
+    {
+      "epoch": 0.1851357365376057,
+      "grad_norm": 3.9079871625582046,
+      "learning_rate": 1.9992517055649155e-05,
+      "loss": 0.9144,
+      "step": 364
+    },
+    {
+      "epoch": 0.1856443511984233,
+      "grad_norm": 3.3013404146846734,
+      "learning_rate": 1.999229861504062e-05,
+      "loss": 0.9084,
+      "step": 365
+    },
+    {
+      "epoch": 0.1861529658592409,
+      "grad_norm": 3.4736717979081617,
+      "learning_rate": 1.9992077033027358e-05,
+      "loss": 0.8435,
+      "step": 366
+    },
+    {
+      "epoch": 0.1866615805200585,
+      "grad_norm": 3.461796569982609,
+      "learning_rate": 1.9991852309679026e-05,
+      "loss": 0.8355,
+      "step": 367
+    },
+    {
+      "epoch": 0.18717019518087608,
+      "grad_norm": 3.1010134424768885,
+      "learning_rate": 1.9991624445066275e-05,
+      "loss": 0.923,
+      "step": 368
+    },
+    {
+      "epoch": 0.1876788098416937,
+      "grad_norm": 3.411170974548555,
+      "learning_rate": 1.9991393439260743e-05,
+      "loss": 0.8342,
+      "step": 369
+    },
+    {
+      "epoch": 0.18818742450251127,
+      "grad_norm": 3.368956771478135,
+      "learning_rate": 1.999115929233505e-05,
+      "loss": 0.8169,
+      "step": 370
+    },
+    {
+      "epoch": 0.1886960391633289,
+      "grad_norm": 3.026649050836372,
+      "learning_rate": 1.999092200436281e-05,
+      "loss": 0.8914,
+      "step": 371
+    },
+    {
+      "epoch": 0.18920465382414647,
+      "grad_norm": 3.442051873498352,
+      "learning_rate": 1.9990681575418625e-05,
+      "loss": 0.8266,
+      "step": 372
+    },
+    {
+      "epoch": 0.1897132684849641,
+      "grad_norm": 3.014659351444044,
+      "learning_rate": 1.9990438005578076e-05,
+      "loss": 0.928,
+      "step": 373
+    },
+    {
+      "epoch": 0.19022188314578167,
+      "grad_norm": 3.453899866222047,
+      "learning_rate": 1.9990191294917742e-05,
+      "loss": 0.8491,
+      "step": 374
+    },
+    {
+      "epoch": 0.1907304978065993,
+      "grad_norm": 2.8108291749294843,
+      "learning_rate": 1.998994144351518e-05,
+      "loss": 0.863,
+      "step": 375
+    },
+    {
+      "epoch": 0.19123911246741687,
+      "grad_norm": 3.1418393951881516,
+      "learning_rate": 1.9989688451448947e-05,
+      "loss": 0.8477,
+      "step": 376
+    },
+    {
+      "epoch": 0.19174772712823446,
+      "grad_norm": 2.9764256883954014,
+      "learning_rate": 1.9989432318798573e-05,
+      "loss": 0.9176,
+      "step": 377
+    },
+    {
+      "epoch": 0.19225634178905207,
+      "grad_norm": 3.185278922762636,
+      "learning_rate": 1.9989173045644577e-05,
+      "loss": 0.9403,
+      "step": 378
+    },
+    {
+      "epoch": 0.19276495644986966,
+      "grad_norm": 2.87416827568908,
+      "learning_rate": 1.9988910632068482e-05,
+      "loss": 0.8676,
+      "step": 379
+    },
+    {
+      "epoch": 0.19327357111068727,
+      "grad_norm": 3.4833754401574755,
+      "learning_rate": 1.998864507815278e-05,
+      "loss": 0.8449,
+      "step": 380
+    },
+    {
+      "epoch": 0.19378218577150486,
+      "grad_norm": 3.910302219467288,
+      "learning_rate": 1.998837638398095e-05,
+      "loss": 0.7726,
+      "step": 381
+    },
+    {
+      "epoch": 0.19429080043232247,
+      "grad_norm": 2.8091223886403296,
+      "learning_rate": 1.9988104549637474e-05,
+      "loss": 0.8147,
+      "step": 382
+    },
+    {
+      "epoch": 0.19479941509314005,
+      "grad_norm": 3.0870581801121695,
+      "learning_rate": 1.9987829575207814e-05,
+      "loss": 0.7978,
+      "step": 383
+    },
+    {
+      "epoch": 0.19530802975395767,
+      "grad_norm": 2.990270413830018,
+      "learning_rate": 1.9987551460778406e-05,
+      "loss": 0.836,
+      "step": 384
+    },
+    {
+      "epoch": 0.19581664441477525,
+      "grad_norm": 4.179347742835237,
+      "learning_rate": 1.9987270206436694e-05,
+      "loss": 0.8294,
+      "step": 385
+    },
+    {
+      "epoch": 0.19632525907559287,
+      "grad_norm": 2.691133393960953,
+      "learning_rate": 1.9986985812271094e-05,
+      "loss": 0.7861,
+      "step": 386
+    },
+    {
+      "epoch": 0.19683387373641045,
+      "grad_norm": 3.3207589601032406,
+      "learning_rate": 1.9986698278371016e-05,
+      "loss": 0.9578,
+      "step": 387
+    },
+    {
+      "epoch": 0.19734248839722804,
+      "grad_norm": 2.9815113648820626,
+      "learning_rate": 1.9986407604826862e-05,
+      "loss": 0.8172,
+      "step": 388
+    },
+    {
+      "epoch": 0.19785110305804565,
+      "grad_norm": 3.6987492586672066,
+      "learning_rate": 1.9986113791730004e-05,
+      "loss": 0.6913,
+      "step": 389
+    },
+    {
+      "epoch": 0.19835971771886324,
+      "grad_norm": 3.2455952946776185,
+      "learning_rate": 1.998581683917282e-05,
+      "loss": 0.8363,
+      "step": 390
+    },
+    {
+      "epoch": 0.19886833237968085,
+      "grad_norm": 2.8473668834460573,
+      "learning_rate": 1.998551674724866e-05,
+      "loss": 0.9246,
+      "step": 391
+    },
+    {
+      "epoch": 0.19937694704049844,
+      "grad_norm": 3.5352363703800638,
+      "learning_rate": 1.998521351605187e-05,
+      "loss": 0.7643,
+      "step": 392
+    },
+    {
+      "epoch": 0.19988556170131605,
+      "grad_norm": 3.8425680684877808,
+      "learning_rate": 1.9984907145677786e-05,
+      "loss": 0.8566,
+      "step": 393
+    },
+    {
+      "epoch": 0.20039417636213364,
+      "grad_norm": 3.2679344662430396,
+      "learning_rate": 1.9984597636222723e-05,
+      "loss": 0.8759,
+      "step": 394
+    },
+    {
+      "epoch": 0.20090279102295125,
+      "grad_norm": 3.253911965171677,
+      "learning_rate": 1.998428498778398e-05,
+      "loss": 0.7766,
+      "step": 395
+    },
+    {
+      "epoch": 0.20141140568376883,
+      "grad_norm": 3.558627784809115,
+      "learning_rate": 1.998396920045985e-05,
+      "loss": 0.7998,
+      "step": 396
+    },
+    {
+      "epoch": 0.20192002034458642,
+      "grad_norm": 3.400090799802178,
+      "learning_rate": 1.9983650274349618e-05,
+      "loss": 0.8725,
+      "step": 397
+    },
+    {
+      "epoch": 0.20242863500540403,
+      "grad_norm": 3.022981673386697,
+      "learning_rate": 1.998332820955354e-05,
+      "loss": 0.829,
+      "step": 398
+    },
+    {
+      "epoch": 0.20293724966622162,
+      "grad_norm": 3.312379510955405,
+      "learning_rate": 1.998300300617287e-05,
+      "loss": 0.8599,
+      "step": 399
+    },
+    {
+      "epoch": 0.20344586432703923,
+      "grad_norm": 3.363160028167116,
+      "learning_rate": 1.998267466430985e-05,
+      "loss": 0.8561,
+      "step": 400
+    },
+    {
+      "epoch": 0.20344586432703923,
+      "eval_loss": 0.8010753393173218,
+      "eval_runtime": 169.8477,
+      "eval_samples_per_second": 19.5,
+      "eval_steps_per_second": 9.75,
+      "step": 400
+    },
+    {
+      "epoch": 0.20395447898785682,
+      "grad_norm": 3.070576818993912,
+      "learning_rate": 1.9982343184067704e-05,
+      "loss": 0.9207,
+      "step": 401
+    },
+    {
+      "epoch": 0.20446309364867443,
+      "grad_norm": 3.0077776225656443,
+      "learning_rate": 1.9982008565550642e-05,
+      "loss": 0.8342,
+      "step": 402
+    },
+    {
+      "epoch": 0.20497170830949202,
+      "grad_norm": 2.8145678230119935,
+      "learning_rate": 1.998167080886386e-05,
+      "loss": 0.7832,
+      "step": 403
+    },
+    {
+      "epoch": 0.20548032297030963,
+      "grad_norm": 2.860629114169295,
+      "learning_rate": 1.9981329914113546e-05,
+      "loss": 0.841,
+      "step": 404
+    },
+    {
+      "epoch": 0.20598893763112722,
+      "grad_norm": 3.0310607536467105,
+      "learning_rate": 1.9980985881406872e-05,
+      "loss": 0.7525,
+      "step": 405
+    },
+    {
+      "epoch": 0.2064975522919448,
+      "grad_norm": 2.7194338409854404,
+      "learning_rate": 1.9980638710851995e-05,
+      "loss": 0.7702,
+      "step": 406
+    },
+    {
+      "epoch": 0.20700616695276242,
+      "grad_norm": 2.8958653144076734,
+      "learning_rate": 1.9980288402558056e-05,
+      "loss": 0.8114,
+      "step": 407
+    },
+    {
+      "epoch": 0.20751478161358,
+      "grad_norm": 3.6022922000079727,
+      "learning_rate": 1.997993495663519e-05,
+      "loss": 0.8684,
+      "step": 408
+    },
+    {
+      "epoch": 0.20802339627439762,
+      "grad_norm": 2.962303309611007,
+      "learning_rate": 1.997957837319451e-05,
+      "loss": 0.8475,
+      "step": 409
+    },
+    {
+      "epoch": 0.2085320109352152,
+      "grad_norm": 2.8507734487295626,
+      "learning_rate": 1.997921865234813e-05,
+      "loss": 0.802,
+      "step": 410
+    },
+    {
+      "epoch": 0.20904062559603281,
+      "grad_norm": 2.808808341071886,
+      "learning_rate": 1.997885579420913e-05,
+      "loss": 0.9074,
+      "step": 411
+    },
+    {
+      "epoch": 0.2095492402568504,
+      "grad_norm": 3.3682468810037665,
+      "learning_rate": 1.9978489798891584e-05,
+      "loss": 0.8447,
+      "step": 412
+    },
+    {
+      "epoch": 0.210057854917668,
+      "grad_norm": 2.748556798760612,
+      "learning_rate": 1.9978120666510564e-05,
+      "loss": 0.8627,
+      "step": 413
+    },
+    {
+      "epoch": 0.2105664695784856,
+      "grad_norm": 3.3573798353970705,
+      "learning_rate": 1.9977748397182112e-05,
+      "loss": 0.7056,
+      "step": 414
+    },
+    {
+      "epoch": 0.21107508423930318,
+      "grad_norm": 2.903235071434086,
+      "learning_rate": 1.9977372991023266e-05,
+      "loss": 0.8431,
+      "step": 415
+    },
+    {
+      "epoch": 0.2115836989001208,
+      "grad_norm": 3.1637686133769733,
+      "learning_rate": 1.9976994448152046e-05,
+      "loss": 0.8761,
+      "step": 416
+    },
+    {
+      "epoch": 0.21209231356093838,
+      "grad_norm": 3.63037605208345,
+      "learning_rate": 1.997661276868746e-05,
+      "loss": 0.7255,
+      "step": 417
+    },
+    {
+      "epoch": 0.212600928221756,
+      "grad_norm": 2.865667704323298,
+      "learning_rate": 1.99762279527495e-05,
+      "loss": 0.8234,
+      "step": 418
+    },
+    {
+      "epoch": 0.21310954288257358,
+      "grad_norm": 2.6356160888810702,
+      "learning_rate": 1.9975840000459148e-05,
+      "loss": 0.7229,
+      "step": 419
+    },
+    {
+      "epoch": 0.2136181575433912,
+      "grad_norm": 3.214769587350835,
+      "learning_rate": 1.9975448911938368e-05,
+      "loss": 0.8293,
+      "step": 420
+    },
+    {
+      "epoch": 0.21412677220420878,
+      "grad_norm": 2.787934232683153,
+      "learning_rate": 1.9975054687310108e-05,
+      "loss": 0.8628,
+      "step": 421
+    },
+    {
+      "epoch": 0.2146353868650264,
+      "grad_norm": 3.1075320509027247,
+      "learning_rate": 1.9974657326698313e-05,
+      "loss": 0.9038,
+      "step": 422
+    },
+    {
+      "epoch": 0.21514400152584398,
+      "grad_norm": 3.2965258372755395,
+      "learning_rate": 1.9974256830227904e-05,
+      "loss": 0.8858,
+      "step": 423
+    },
+    {
+      "epoch": 0.2156526161866616,
+      "grad_norm": 2.9582955443068597,
+      "learning_rate": 1.9973853198024783e-05,
+      "loss": 0.8563,
+      "step": 424
+    },
+    {
+      "epoch": 0.21616123084747918,
+      "grad_norm": 2.9701993917715157,
+      "learning_rate": 1.997344643021585e-05,
+      "loss": 0.8372,
+      "step": 425
+    },
+    {
+      "epoch": 0.21666984550829677,
+      "grad_norm": 3.1139036391679555,
+      "learning_rate": 1.9973036526928987e-05,
+      "loss": 0.7206,
+      "step": 426
+    },
+    {
+      "epoch": 0.21717846016911438,
+      "grad_norm": 2.545943388196383,
+      "learning_rate": 1.997262348829306e-05,
+      "loss": 0.8741,
+      "step": 427
+    },
+    {
+      "epoch": 0.21768707482993196,
+      "grad_norm": 2.621010064098098,
+      "learning_rate": 1.997220731443792e-05,
+      "loss": 0.8031,
+      "step": 428
+    },
+    {
+      "epoch": 0.21819568949074958,
+      "grad_norm": 3.866246827670306,
+      "learning_rate": 1.9971788005494402e-05,
+      "loss": 0.7929,
+      "step": 429
+    },
+    {
+      "epoch": 0.21870430415156716,
+      "grad_norm": 3.025849398065434,
+      "learning_rate": 1.9971365561594337e-05,
+      "loss": 0.8509,
+      "step": 430
+    },
+    {
+      "epoch": 0.21921291881238478,
+      "grad_norm": 3.412971514310531,
+      "learning_rate": 1.997093998287053e-05,
+      "loss": 0.7766,
+      "step": 431
+    },
+    {
+      "epoch": 0.21972153347320236,
+      "grad_norm": 6.995003688285024,
+      "learning_rate": 1.997051126945677e-05,
+      "loss": 0.899,
+      "step": 432
+    },
+    {
+      "epoch": 0.22023014813401998,
+      "grad_norm": 2.9696716484153356,
+      "learning_rate": 1.9970079421487843e-05,
+      "loss": 0.8592,
+      "step": 433
+    },
+    {
+      "epoch": 0.22073876279483756,
+      "grad_norm": 3.4966459552113927,
+      "learning_rate": 1.9969644439099518e-05,
+      "loss": 0.7792,
+      "step": 434
+    },
+    {
+      "epoch": 0.22124737745565515,
+      "grad_norm": 2.9516903253343436,
+      "learning_rate": 1.996920632242854e-05,
+      "loss": 0.8681,
+      "step": 435
+    },
+    {
+      "epoch": 0.22175599211647276,
+      "grad_norm": 3.0620797316562514,
+      "learning_rate": 1.9968765071612646e-05,
+      "loss": 0.9223,
+      "step": 436
+    },
+    {
+      "epoch": 0.22226460677729035,
+      "grad_norm": 2.5284527051084456,
+      "learning_rate": 1.9968320686790558e-05,
+      "loss": 0.706,
+      "step": 437
+    },
+    {
+      "epoch": 0.22277322143810796,
+      "grad_norm": 3.2625889148507707,
+      "learning_rate": 1.9967873168101984e-05,
+      "loss": 0.778,
+      "step": 438
+    },
+    {
+      "epoch": 0.22328183609892555,
+      "grad_norm": 2.8910179077353844,
+      "learning_rate": 1.9967422515687615e-05,
+      "loss": 0.9148,
+      "step": 439
+    },
+    {
+      "epoch": 0.22379045075974316,
+      "grad_norm": 3.0741967245279915,
+      "learning_rate": 1.996696872968913e-05,
+      "loss": 0.8634,
+      "step": 440
+    },
+    {
+      "epoch": 0.22429906542056074,
+      "grad_norm": 3.14454580741972,
+      "learning_rate": 1.9966511810249186e-05,
+      "loss": 0.8548,
+      "step": 441
+    },
+    {
+      "epoch": 0.22480768008137836,
+      "grad_norm": 2.831211829191332,
+      "learning_rate": 1.996605175751144e-05,
+      "loss": 0.7177,
+      "step": 442
+    },
+    {
+      "epoch": 0.22531629474219594,
+      "grad_norm": 2.5321147995546687,
+      "learning_rate": 1.9965588571620514e-05,
+      "loss": 0.7478,
+      "step": 443
+    },
+    {
+      "epoch": 0.22582490940301353,
+      "grad_norm": 2.696754136366398,
+      "learning_rate": 1.9965122252722034e-05,
+      "loss": 0.8852,
+      "step": 444
+    },
+    {
+      "epoch": 0.22633352406383114,
+      "grad_norm": 3.053951555268798,
+      "learning_rate": 1.99646528009626e-05,
+      "loss": 0.9272,
+      "step": 445
+    },
+    {
+      "epoch": 0.22684213872464873,
+      "grad_norm": 2.8338008008794375,
+      "learning_rate": 1.9964180216489793e-05,
+      "loss": 0.8499,
+      "step": 446
+    },
+    {
+      "epoch": 0.22735075338546634,
+      "grad_norm": 3.344354022364304,
+      "learning_rate": 1.9963704499452196e-05,
+      "loss": 0.9192,
+      "step": 447
+    },
+    {
+      "epoch": 0.22785936804628393,
+      "grad_norm": 3.3669116267861696,
+      "learning_rate": 1.996322564999936e-05,
+      "loss": 0.8697,
+      "step": 448
+    },
+    {
+      "epoch": 0.22836798270710154,
+      "grad_norm": 3.3618728222957834,
+      "learning_rate": 1.9962743668281828e-05,
+      "loss": 0.934,
+      "step": 449
+    },
+    {
+      "epoch": 0.22887659736791913,
+      "grad_norm": 2.9821838004089196,
+      "learning_rate": 1.9962258554451127e-05,
+      "loss": 0.8151,
+      "step": 450
+    },
+    {
+      "epoch": 0.22938521202873674,
+      "grad_norm": 3.421488700904903,
+      "learning_rate": 1.9961770308659765e-05,
+      "loss": 0.8787,
+      "step": 451
+    },
+    {
+      "epoch": 0.22989382668955433,
+      "grad_norm": 3.282373000024754,
+      "learning_rate": 1.9961278931061246e-05,
+      "loss": 0.7626,
+      "step": 452
+    },
+    {
+      "epoch": 0.2304024413503719,
+      "grad_norm": 2.6128584401399264,
+      "learning_rate": 1.996078442181004e-05,
+      "loss": 0.6905,
+      "step": 453
+    },
+    {
+      "epoch": 0.23091105601118952,
+      "grad_norm": 2.9232649327637863,
+      "learning_rate": 1.9960286781061625e-05,
+      "loss": 0.84,
+      "step": 454
+    },
+    {
+      "epoch": 0.2314196706720071,
+      "grad_norm": 2.819929186696176,
+      "learning_rate": 1.995978600897244e-05,
+      "loss": 0.833,
+      "step": 455
+    },
+    {
+      "epoch": 0.23192828533282472,
+      "grad_norm": 5.180628258152666,
+      "learning_rate": 1.995928210569992e-05,
+      "loss": 0.9353,
+      "step": 456
+    },
+    {
+      "epoch": 0.2324368999936423,
+      "grad_norm": 3.5280656888992286,
+      "learning_rate": 1.9958775071402486e-05,
+      "loss": 0.9315,
+      "step": 457
+    },
+    {
+      "epoch": 0.23294551465445992,
+      "grad_norm": 2.833167875358066,
+      "learning_rate": 1.9958264906239547e-05,
+      "loss": 0.7401,
+      "step": 458
+    },
+    {
+      "epoch": 0.2334541293152775,
+      "grad_norm": 2.928463611545397,
+      "learning_rate": 1.9957751610371477e-05,
+      "loss": 0.8281,
+      "step": 459
+    },
+    {
+      "epoch": 0.23396274397609512,
+      "grad_norm": 3.2674222682121443,
+      "learning_rate": 1.9957235183959654e-05,
+      "loss": 0.9141,
+      "step": 460
+    },
+    {
+      "epoch": 0.2344713586369127,
+      "grad_norm": 2.7201710637379617,
+      "learning_rate": 1.9956715627166436e-05,
+      "loss": 0.7933,
+      "step": 461
+    },
+    {
+      "epoch": 0.2349799732977303,
+      "grad_norm": 2.947113215182034,
+      "learning_rate": 1.995619294015516e-05,
+      "loss": 0.9543,
+      "step": 462
+    },
+    {
+      "epoch": 0.2354885879585479,
+      "grad_norm": 2.706447609961499,
+      "learning_rate": 1.995566712309015e-05,
+      "loss": 0.8224,
+      "step": 463
+    },
+    {
+      "epoch": 0.2359972026193655,
+      "grad_norm": 3.277175553453659,
+      "learning_rate": 1.995513817613671e-05,
+      "loss": 0.8414,
+      "step": 464
+    },
+    {
+      "epoch": 0.2365058172801831,
+      "grad_norm": 3.241881280249092,
+      "learning_rate": 1.9954606099461136e-05,
+      "loss": 0.9293,
+      "step": 465
+    },
+    {
+      "epoch": 0.2370144319410007,
+      "grad_norm": 2.90922672864997,
+      "learning_rate": 1.9954070893230704e-05,
+      "loss": 0.8311,
+      "step": 466
+    },
+    {
+      "epoch": 0.2375230466018183,
+      "grad_norm": 3.2550087992231327,
+      "learning_rate": 1.995353255761367e-05,
+      "loss": 0.8728,
+      "step": 467
+    },
+    {
+      "epoch": 0.2380316612626359,
+      "grad_norm": 2.452925655745303,
+      "learning_rate": 1.9952991092779285e-05,
+      "loss": 0.8145,
+      "step": 468
+    },
+    {
+      "epoch": 0.2385402759234535,
+      "grad_norm": 3.224850253391515,
+      "learning_rate": 1.9952446498897766e-05,
+      "loss": 0.9784,
+      "step": 469
+    },
+    {
+      "epoch": 0.2390488905842711,
+      "grad_norm": 2.8884998654627494,
+      "learning_rate": 1.9951898776140327e-05,
+      "loss": 0.8782,
+      "step": 470
+    },
+    {
+      "epoch": 0.2395575052450887,
+      "grad_norm": 2.7520562200681016,
+      "learning_rate": 1.995134792467917e-05,
+      "loss": 0.6955,
+      "step": 471
+    },
+    {
+      "epoch": 0.2400661199059063,
+      "grad_norm": 3.4334684832987277,
+      "learning_rate": 1.9950793944687463e-05,
+      "loss": 0.8898,
+      "step": 472
+    },
+    {
+      "epoch": 0.24057473456672387,
+      "grad_norm": 3.038065533521968,
+      "learning_rate": 1.9950236836339372e-05,
+      "loss": 0.9315,
+      "step": 473
+    },
+    {
+      "epoch": 0.2410833492275415,
+      "grad_norm": 3.2730414030286785,
+      "learning_rate": 1.994967659981004e-05,
+      "loss": 0.7867,
+      "step": 474
+    },
+    {
+      "epoch": 0.24159196388835907,
+      "grad_norm": 2.9543745031451216,
+      "learning_rate": 1.9949113235275602e-05,
+      "loss": 0.9246,
+      "step": 475
+    },
+    {
+      "epoch": 0.2421005785491767,
+      "grad_norm": 2.778316193428405,
+      "learning_rate": 1.9948546742913163e-05,
+      "loss": 0.9301,
+      "step": 476
+    },
+    {
+      "epoch": 0.24260919320999427,
+      "grad_norm": 2.8881792598894998,
+      "learning_rate": 1.9947977122900825e-05,
+      "loss": 0.8328,
+      "step": 477
+    },
+    {
+      "epoch": 0.2431178078708119,
+      "grad_norm": 4.236303812170896,
+      "learning_rate": 1.994740437541766e-05,
+      "loss": 0.9042,
+      "step": 478
+    },
+    {
+      "epoch": 0.24362642253162947,
+      "grad_norm": 3.17398512231889,
+      "learning_rate": 1.994682850064373e-05,
+      "loss": 0.8983,
+      "step": 479
+    },
+    {
+      "epoch": 0.24413503719244709,
+      "grad_norm": 2.5736114758141686,
+      "learning_rate": 1.994624949876008e-05,
+      "loss": 0.71,
+      "step": 480
+    },
+    {
+      "epoch": 0.24464365185326467,
+      "grad_norm": 2.8624636118879003,
+      "learning_rate": 1.9945667369948746e-05,
+      "loss": 0.9265,
+      "step": 481
+    },
+    {
+      "epoch": 0.24515226651408226,
+      "grad_norm": 2.6016989549013574,
+      "learning_rate": 1.9945082114392732e-05,
+      "loss": 0.762,
+      "step": 482
+    },
+    {
+      "epoch": 0.24566088117489987,
+      "grad_norm": 2.866606858834618,
+      "learning_rate": 1.9944493732276035e-05,
+      "loss": 0.8287,
+      "step": 483
+    },
+    {
+      "epoch": 0.24616949583571746,
+      "grad_norm": 2.898838724914507,
+      "learning_rate": 1.994390222378363e-05,
+      "loss": 0.7563,
+      "step": 484
+    },
+    {
+      "epoch": 0.24667811049653507,
+      "grad_norm": 3.6054797270536447,
+      "learning_rate": 1.994330758910148e-05,
+      "loss": 0.7718,
+      "step": 485
+    },
+    {
+      "epoch": 0.24718672515735265,
+      "grad_norm": 3.4258539068157483,
+      "learning_rate": 1.994270982841652e-05,
+      "loss": 0.7698,
+      "step": 486
+    },
+    {
+      "epoch": 0.24769533981817027,
+      "grad_norm": 2.806103192335877,
+      "learning_rate": 1.9942108941916685e-05,
+      "loss": 0.8647,
+      "step": 487
+    },
+    {
+      "epoch": 0.24820395447898785,
+      "grad_norm": 3.0130949327242242,
+      "learning_rate": 1.994150492979088e-05,
+      "loss": 0.9358,
+      "step": 488
+    },
+    {
+      "epoch": 0.24871256913980547,
+      "grad_norm": 3.2639120619652404,
+      "learning_rate": 1.9940897792228993e-05,
+      "loss": 0.792,
+      "step": 489
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 2.8543727456942802,
+      "learning_rate": 1.99402875294219e-05,
+      "loss": 0.8202,
+      "step": 490
+    },
+    {
+      "epoch": 0.24972979846144064,
+      "grad_norm": 2.4989864202025855,
+      "learning_rate": 1.993967414156146e-05,
+      "loss": 0.8005,
+      "step": 491
+    },
+    {
+      "epoch": 0.25023841312225825,
+      "grad_norm": 3.283422713507381,
+      "learning_rate": 1.993905762884051e-05,
+      "loss": 0.8854,
+      "step": 492
+    },
+    {
+      "epoch": 0.25074702778307584,
+      "grad_norm": 2.9792481168981237,
+      "learning_rate": 1.9938437991452864e-05,
+      "loss": 0.8507,
+      "step": 493
+    },
+    {
+      "epoch": 0.2512556424438934,
+      "grad_norm": 4.24041722350173,
+      "learning_rate": 1.993781522959333e-05,
+      "loss": 0.8011,
+      "step": 494
+    },
+    {
+      "epoch": 0.25176425710471106,
+      "grad_norm": 3.400520475604416,
+      "learning_rate": 1.99371893434577e-05,
+      "loss": 0.8356,
+      "step": 495
+    },
+    {
+      "epoch": 0.25227287176552865,
+      "grad_norm": 2.854212869608976,
+      "learning_rate": 1.993656033324273e-05,
+      "loss": 0.7519,
+      "step": 496
+    },
+    {
+      "epoch": 0.25278148642634624,
+      "grad_norm": 3.0241554846880554,
+      "learning_rate": 1.9935928199146177e-05,
+      "loss": 0.8592,
+      "step": 497
+    },
+    {
+      "epoch": 0.2532901010871638,
+      "grad_norm": 2.8704786822576143,
+      "learning_rate": 1.993529294136677e-05,
+      "loss": 0.8698,
+      "step": 498
+    },
+    {
+      "epoch": 0.25379871574798146,
+      "grad_norm": 3.2958929398102295,
+      "learning_rate": 1.993465456010423e-05,
+      "loss": 0.8166,
+      "step": 499
+    },
+    {
+      "epoch": 0.25430733040879905,
+      "grad_norm": 3.3326205127214115,
+      "learning_rate": 1.9934013055559244e-05,
+      "loss": 0.8051,
+      "step": 500
+    },
+    {
+      "epoch": 0.25430733040879905,
+      "eval_loss": 0.7910401225090027,
+      "eval_runtime": 169.8165,
+      "eval_samples_per_second": 19.503,
+      "eval_steps_per_second": 9.752,
+      "step": 500
+    },
+    {
+      "epoch": 0.25481594506961663,
+      "grad_norm": 3.473842422557162,
+      "learning_rate": 1.993336842793349e-05,
+      "loss": 0.794,
+      "step": 501
+    },
+    {
+      "epoch": 0.2553245597304342,
+      "grad_norm": 3.32588170343213,
+      "learning_rate": 1.9932720677429634e-05,
+      "loss": 0.8371,
+      "step": 502
+    },
+    {
+      "epoch": 0.2558331743912518,
+      "grad_norm": 3.7537032597464925,
+      "learning_rate": 1.9932069804251312e-05,
+      "loss": 0.9639,
+      "step": 503
+    },
+    {
+      "epoch": 0.25634178905206945,
+      "grad_norm": 2.492613561677559,
+      "learning_rate": 1.9931415808603152e-05,
+      "loss": 0.8552,
+      "step": 504
+    },
+    {
+      "epoch": 0.25685040371288703,
+      "grad_norm": 2.9264423759153746,
+      "learning_rate": 1.9930758690690758e-05,
+      "loss": 0.8011,
+      "step": 505
+    },
+    {
+      "epoch": 0.2573590183737046,
+      "grad_norm": 2.876080848862621,
+      "learning_rate": 1.993009845072071e-05,
+      "loss": 0.8377,
+      "step": 506
+    },
+    {
+      "epoch": 0.2578676330345222,
+      "grad_norm": 2.6664426133285826,
+      "learning_rate": 1.9929435088900585e-05,
+      "loss": 0.7296,
+      "step": 507
+    },
+    {
+      "epoch": 0.25837624769533984,
+      "grad_norm": 2.8616622365192597,
+      "learning_rate": 1.9928768605438927e-05,
+      "loss": 0.7628,
+      "step": 508
+    },
+    {
+      "epoch": 0.25888486235615743,
+      "grad_norm": 2.9863378650613726,
+      "learning_rate": 1.992809900054527e-05,
+      "loss": 0.9388,
+      "step": 509
+    },
+    {
+      "epoch": 0.259393477016975,
+      "grad_norm": 2.7489745222113444,
+      "learning_rate": 1.9927426274430123e-05,
+      "loss": 0.7747,
+      "step": 510
+    },
+    {
+      "epoch": 0.2599020916777926,
+      "grad_norm": 2.850341368191215,
+      "learning_rate": 1.9926750427304978e-05,
+      "loss": 0.945,
+      "step": 511
+    },
+    {
+      "epoch": 0.2604107063386102,
+      "grad_norm": 2.9014856807993707,
+      "learning_rate": 1.992607145938231e-05,
+      "loss": 0.9277,
+      "step": 512
+    },
+    {
+      "epoch": 0.26091932099942783,
+      "grad_norm": 2.9418014524305445,
+      "learning_rate": 1.9925389370875582e-05,
+      "loss": 0.7792,
+      "step": 513
+    },
+    {
+      "epoch": 0.2614279356602454,
+      "grad_norm": 2.981444832695261,
+      "learning_rate": 1.9924704161999225e-05,
+      "loss": 0.8282,
+      "step": 514
+    },
+    {
+      "epoch": 0.261936550321063,
+      "grad_norm": 3.205278816890404,
+      "learning_rate": 1.992401583296866e-05,
+      "loss": 0.9186,
+      "step": 515
+    },
+    {
+      "epoch": 0.2624451649818806,
+      "grad_norm": 2.9666081783515503,
+      "learning_rate": 1.9923324384000278e-05,
+      "loss": 0.8442,
+      "step": 516
+    },
+    {
+      "epoch": 0.2629537796426982,
+      "grad_norm": 2.880151362945031,
+      "learning_rate": 1.9922629815311462e-05,
+      "loss": 0.7789,
+      "step": 517
+    },
+    {
+      "epoch": 0.2634623943035158,
+      "grad_norm": 2.752469081445153,
+      "learning_rate": 1.9921932127120578e-05,
+      "loss": 0.7961,
+      "step": 518
+    },
+    {
+      "epoch": 0.2639710089643334,
+      "grad_norm": 3.0746688416344568,
+      "learning_rate": 1.992123131964696e-05,
+      "loss": 0.7985,
+      "step": 519
+    },
+    {
+      "epoch": 0.264479623625151,
+      "grad_norm": 2.575381952494214,
+      "learning_rate": 1.9920527393110936e-05,
+      "loss": 0.8483,
+      "step": 520
+    },
+    {
+      "epoch": 0.26498823828596857,
+      "grad_norm": 3.760299272192983,
+      "learning_rate": 1.99198203477338e-05,
+      "loss": 0.8869,
+      "step": 521
+    },
+    {
+      "epoch": 0.2654968529467862,
+      "grad_norm": 2.8537915022677334,
+      "learning_rate": 1.9919110183737843e-05,
+      "loss": 0.9113,
+      "step": 522
+    },
+    {
+      "epoch": 0.2660054676076038,
+      "grad_norm": 3.1691562670539497,
+      "learning_rate": 1.9918396901346324e-05,
+      "loss": 0.7839,
+      "step": 523
+    },
+    {
+      "epoch": 0.2665140822684214,
+      "grad_norm": 2.8748153958968983,
+      "learning_rate": 1.9917680500783486e-05,
+      "loss": 0.788,
+      "step": 524
+    },
+    {
+      "epoch": 0.26702269692923897,
+      "grad_norm": 2.487341111825957,
+      "learning_rate": 1.9916960982274557e-05,
+      "loss": 0.7233,
+      "step": 525
+    },
+    {
+      "epoch": 0.2675313115900566,
+      "grad_norm": 3.0328714025988885,
+      "learning_rate": 1.9916238346045734e-05,
+      "loss": 0.8976,
+      "step": 526
+    },
+    {
+      "epoch": 0.2680399262508742,
+      "grad_norm": 2.9127922664578434,
+      "learning_rate": 1.991551259232421e-05,
+      "loss": 0.8784,
+      "step": 527
+    },
+    {
+      "epoch": 0.2685485409116918,
+      "grad_norm": 3.2462459303603075,
+      "learning_rate": 1.991478372133814e-05,
+      "loss": 0.8422,
+      "step": 528
+    },
+    {
+      "epoch": 0.26905715557250937,
+      "grad_norm": 2.707490075710491,
+      "learning_rate": 1.9914051733316676e-05,
+      "loss": 0.764,
+      "step": 529
+    },
+    {
+      "epoch": 0.26956577023332695,
+      "grad_norm": 3.4342587535508056,
+      "learning_rate": 1.991331662848994e-05,
+      "loss": 0.9554,
+      "step": 530
+    },
+    {
+      "epoch": 0.2700743848941446,
+      "grad_norm": 3.549517751900549,
+      "learning_rate": 1.9912578407089035e-05,
+      "loss": 0.8475,
+      "step": 531
+    },
+    {
+      "epoch": 0.2705829995549622,
+      "grad_norm": 3.1729112257298193,
+      "learning_rate": 1.9911837069346048e-05,
+      "loss": 0.8843,
+      "step": 532
+    },
+    {
+      "epoch": 0.27109161421577976,
+      "grad_norm": 2.7543892944048114,
+      "learning_rate": 1.9911092615494033e-05,
+      "loss": 0.837,
+      "step": 533
+    },
+    {
+      "epoch": 0.27160022887659735,
+      "grad_norm": 3.8450108142794956,
+      "learning_rate": 1.9910345045767048e-05,
+      "loss": 0.8575,
+      "step": 534
+    },
+    {
+      "epoch": 0.272108843537415,
+      "grad_norm": 2.3539076355968995,
+      "learning_rate": 1.9909594360400105e-05,
+      "loss": 0.7775,
+      "step": 535
+    },
+    {
+      "epoch": 0.2726174581982326,
+      "grad_norm": 2.7149038184676586,
+      "learning_rate": 1.990884055962921e-05,
+      "loss": 0.9303,
+      "step": 536
+    },
+    {
+      "epoch": 0.27312607285905016,
+      "grad_norm": 2.9212671963500196,
+      "learning_rate": 1.9908083643691345e-05,
+      "loss": 0.7642,
+      "step": 537
+    },
+    {
+      "epoch": 0.27363468751986775,
+      "grad_norm": 2.6701772549544303,
+      "learning_rate": 1.9907323612824473e-05,
+      "loss": 0.9397,
+      "step": 538
+    },
+    {
+      "epoch": 0.27414330218068533,
+      "grad_norm": 2.9252199008802497,
+      "learning_rate": 1.990656046726753e-05,
+      "loss": 0.9612,
+      "step": 539
+    },
+    {
+      "epoch": 0.274651916841503,
+      "grad_norm": 5.249375805759924,
+      "learning_rate": 1.990579420726044e-05,
+      "loss": 0.8586,
+      "step": 540
+    },
+    {
+      "epoch": 0.27516053150232056,
+      "grad_norm": 2.638539771917551,
+      "learning_rate": 1.9905024833044095e-05,
+      "loss": 0.7672,
+      "step": 541
+    },
+    {
+      "epoch": 0.27566914616313815,
+      "grad_norm": 2.5074300979500816,
+      "learning_rate": 1.990425234486038e-05,
+      "loss": 0.8149,
+      "step": 542
+    },
+    {
+      "epoch": 0.27617776082395573,
+      "grad_norm": 2.842712212980779,
+      "learning_rate": 1.9903476742952152e-05,
+      "loss": 0.852,
+      "step": 543
+    },
+    {
+      "epoch": 0.2766863754847734,
+      "grad_norm": 2.4562148540593425,
+      "learning_rate": 1.9902698027563244e-05,
+      "loss": 0.847,
+      "step": 544
+    },
+    {
+      "epoch": 0.27719499014559096,
+      "grad_norm": 3.178311436828924,
+      "learning_rate": 1.9901916198938468e-05,
+      "loss": 1.0046,
+      "step": 545
+    },
+    {
+      "epoch": 0.27770360480640854,
+      "grad_norm": 2.7590096012513916,
+      "learning_rate": 1.9901131257323622e-05,
+      "loss": 0.8063,
+      "step": 546
+    },
+    {
+      "epoch": 0.27821221946722613,
+      "grad_norm": 2.612834168011121,
+      "learning_rate": 1.9900343202965477e-05,
+      "loss": 0.7542,
+      "step": 547
+    },
+    {
+      "epoch": 0.2787208341280437,
+      "grad_norm": 3.642853803659827,
+      "learning_rate": 1.989955203611178e-05,
+      "loss": 0.9115,
+      "step": 548
+    },
+    {
+      "epoch": 0.27922944878886136,
+      "grad_norm": 2.5873223137775265,
+      "learning_rate": 1.9898757757011267e-05,
+      "loss": 0.9384,
+      "step": 549
+    },
+    {
+      "epoch": 0.27973806344967894,
+      "grad_norm": 4.373648298443367,
+      "learning_rate": 1.9897960365913638e-05,
+      "loss": 0.8899,
+      "step": 550
+    },
+    {
+      "epoch": 0.2802466781104965,
+      "grad_norm": 2.588511496866332,
+      "learning_rate": 1.9897159863069583e-05,
+      "loss": 0.7645,
+      "step": 551
+    },
+    {
+      "epoch": 0.2807552927713141,
+      "grad_norm": 2.7059680783065807,
+      "learning_rate": 1.9896356248730765e-05,
+      "loss": 0.7475,
+      "step": 552
+    },
+    {
+      "epoch": 0.28126390743213175,
+      "grad_norm": 2.6449548559536815,
+      "learning_rate": 1.9895549523149825e-05,
+      "loss": 0.8247,
+      "step": 553
+    },
+    {
+      "epoch": 0.28177252209294934,
+      "grad_norm": 2.6473447632777822,
+      "learning_rate": 1.989473968658039e-05,
+      "loss": 0.7693,
+      "step": 554
+    },
+    {
+      "epoch": 0.2822811367537669,
+      "grad_norm": 2.4491599530671753,
+      "learning_rate": 1.9893926739277048e-05,
+      "loss": 0.8257,
+      "step": 555
+    },
+    {
+      "epoch": 0.2827897514145845,
+      "grad_norm": 2.7610962421094483,
+      "learning_rate": 1.989311068149539e-05,
+      "loss": 0.8102,
+      "step": 556
+    },
+    {
+      "epoch": 0.2832983660754021,
+      "grad_norm": 2.8208644675523344,
+      "learning_rate": 1.9892291513491953e-05,
+      "loss": 0.867,
+      "step": 557
+    },
+    {
+      "epoch": 0.28380698073621974,
+      "grad_norm": 3.0997741309296654,
+      "learning_rate": 1.989146923552428e-05,
+      "loss": 0.7987,
+      "step": 558
+    },
+    {
+      "epoch": 0.2843155953970373,
+      "grad_norm": 2.638397360639484,
+      "learning_rate": 1.989064384785088e-05,
+      "loss": 0.8216,
+      "step": 559
+    },
+    {
+      "epoch": 0.2848242100578549,
+      "grad_norm": 2.532035046568428,
+      "learning_rate": 1.988981535073124e-05,
+      "loss": 0.7557,
+      "step": 560
+    },
+    {
+      "epoch": 0.2853328247186725,
+      "grad_norm": 2.6077402441986774,
+      "learning_rate": 1.988898374442582e-05,
+      "loss": 0.713,
+      "step": 561
+    },
+    {
+      "epoch": 0.28584143937949014,
+      "grad_norm": 2.52353632588749,
+      "learning_rate": 1.9888149029196068e-05,
+      "loss": 0.8632,
+      "step": 562
+    },
+    {
+      "epoch": 0.2863500540403077,
+      "grad_norm": 3.074850906851458,
+      "learning_rate": 1.9887311205304403e-05,
+      "loss": 0.9381,
+      "step": 563
+    },
+    {
+      "epoch": 0.2868586687011253,
+      "grad_norm": 2.7700131189469452,
+      "learning_rate": 1.9886470273014224e-05,
+      "loss": 0.8329,
+      "step": 564
+    },
+    {
+      "epoch": 0.2873672833619429,
+      "grad_norm": 2.6184685603656837,
+      "learning_rate": 1.9885626232589903e-05,
+      "loss": 0.8319,
+      "step": 565
+    },
+    {
+      "epoch": 0.2878758980227605,
+      "grad_norm": 2.8707305055610988,
+      "learning_rate": 1.9884779084296792e-05,
+      "loss": 0.8764,
+      "step": 566
+    },
+    {
+      "epoch": 0.2883845126835781,
+      "grad_norm": 2.6897668371580363,
+      "learning_rate": 1.988392882840122e-05,
+      "loss": 0.7477,
+      "step": 567
+    },
+    {
+      "epoch": 0.2888931273443957,
+      "grad_norm": 2.904070785742158,
+      "learning_rate": 1.988307546517049e-05,
+      "loss": 0.8657,
+      "step": 568
+    },
+    {
+      "epoch": 0.2894017420052133,
+      "grad_norm": 2.675961052638251,
+      "learning_rate": 1.988221899487289e-05,
+      "loss": 0.8003,
+      "step": 569
+    },
+    {
+      "epoch": 0.2899103566660309,
+      "grad_norm": 2.882434851248488,
+      "learning_rate": 1.9881359417777673e-05,
+      "loss": 0.7731,
+      "step": 570
+    },
+    {
+      "epoch": 0.2904189713268485,
+      "grad_norm": 3.2158222401205125,
+      "learning_rate": 1.988049673415508e-05,
+      "loss": 0.9736,
+      "step": 571
+    },
+    {
+      "epoch": 0.2909275859876661,
+      "grad_norm": 2.5875555318134733,
+      "learning_rate": 1.987963094427632e-05,
+      "loss": 0.7596,
+      "step": 572
+    },
+    {
+      "epoch": 0.2914362006484837,
+      "grad_norm": 3.19758521401177,
+      "learning_rate": 1.9878762048413587e-05,
+      "loss": 0.9127,
+      "step": 573
+    },
+    {
+      "epoch": 0.2919448153093013,
+      "grad_norm": 2.6393678821609563,
+      "learning_rate": 1.9877890046840042e-05,
+      "loss": 0.8307,
+      "step": 574
+    },
+    {
+      "epoch": 0.29245342997011886,
+      "grad_norm": 2.429009149677858,
+      "learning_rate": 1.987701493982983e-05,
+      "loss": 0.7769,
+      "step": 575
+    },
+    {
+      "epoch": 0.2929620446309365,
+      "grad_norm": 2.5579227412101404,
+      "learning_rate": 1.987613672765807e-05,
+      "loss": 0.8558,
+      "step": 576
+    },
+    {
+      "epoch": 0.2934706592917541,
+      "grad_norm": 2.89793321690243,
+      "learning_rate": 1.987525541060085e-05,
+      "loss": 0.9917,
+      "step": 577
+    },
+    {
+      "epoch": 0.2939792739525717,
+      "grad_norm": 3.8669915660196987,
+      "learning_rate": 1.987437098893525e-05,
+      "loss": 0.849,
+      "step": 578
+    },
+    {
+      "epoch": 0.29448788861338926,
+      "grad_norm": 2.7120174488530306,
+      "learning_rate": 1.9873483462939314e-05,
+      "loss": 0.9175,
+      "step": 579
+    },
+    {
+      "epoch": 0.2949965032742069,
+      "grad_norm": 2.8856052011162205,
+      "learning_rate": 1.9872592832892062e-05,
+      "loss": 0.8072,
+      "step": 580
+    },
+    {
+      "epoch": 0.2955051179350245,
+      "grad_norm": 2.9749379768638073,
+      "learning_rate": 1.9871699099073492e-05,
+      "loss": 0.8335,
+      "step": 581
+    },
+    {
+      "epoch": 0.29601373259584207,
+      "grad_norm": 3.0435253021206243,
+      "learning_rate": 1.9870802261764584e-05,
+      "loss": 0.8413,
+      "step": 582
+    },
+    {
+      "epoch": 0.29652234725665966,
+      "grad_norm": 3.176251478679002,
+      "learning_rate": 1.986990232124728e-05,
+      "loss": 0.7443,
+      "step": 583
+    },
+    {
+      "epoch": 0.2970309619174773,
+      "grad_norm": 2.728048974165298,
+      "learning_rate": 1.9868999277804516e-05,
+      "loss": 0.8316,
+      "step": 584
+    },
+    {
+      "epoch": 0.2975395765782949,
+      "grad_norm": 2.564717333650217,
+      "learning_rate": 1.986809313172018e-05,
+      "loss": 0.8365,
+      "step": 585
+    },
+    {
+      "epoch": 0.29804819123911247,
+      "grad_norm": 2.9416108281902904,
+      "learning_rate": 1.986718388327916e-05,
+      "loss": 0.8311,
+      "step": 586
+    },
+    {
+      "epoch": 0.29855680589993006,
+      "grad_norm": 2.920523754334792,
+      "learning_rate": 1.9866271532767302e-05,
+      "loss": 0.8544,
+      "step": 587
+    },
+    {
+      "epoch": 0.29906542056074764,
+      "grad_norm": 2.5480379597158342,
+      "learning_rate": 1.9865356080471438e-05,
+      "loss": 0.7726,
+      "step": 588
+    },
+    {
+      "epoch": 0.2995740352215653,
+      "grad_norm": 2.7271594948815263,
+      "learning_rate": 1.986443752667936e-05,
+      "loss": 0.9192,
+      "step": 589
+    },
+    {
+      "epoch": 0.30008264988238287,
+      "grad_norm": 2.7385907617238057,
+      "learning_rate": 1.986351587167986e-05,
+      "loss": 0.914,
+      "step": 590
+    },
+    {
+      "epoch": 0.30059126454320045,
+      "grad_norm": 2.6335888179346987,
+      "learning_rate": 1.986259111576268e-05,
+      "loss": 0.978,
+      "step": 591
+    },
+    {
+      "epoch": 0.30109987920401804,
+      "grad_norm": 2.846184998022572,
+      "learning_rate": 1.9861663259218545e-05,
+      "loss": 0.7641,
+      "step": 592
+    },
+    {
+      "epoch": 0.3016084938648357,
+      "grad_norm": 2.7410836022037013,
+      "learning_rate": 1.9860732302339166e-05,
+      "loss": 0.8307,
+      "step": 593
+    },
+    {
+      "epoch": 0.30211710852565327,
+      "grad_norm": 2.8320100349420896,
+      "learning_rate": 1.9859798245417216e-05,
+      "loss": 0.8101,
+      "step": 594
+    },
+    {
+      "epoch": 0.30262572318647085,
+      "grad_norm": 3.451481078306983,
+      "learning_rate": 1.9858861088746343e-05,
+      "loss": 1.0823,
+      "step": 595
+    },
+    {
+      "epoch": 0.30313433784728844,
+      "grad_norm": 3.530911184615118,
+      "learning_rate": 1.985792083262118e-05,
+      "loss": 0.7673,
+      "step": 596
+    },
+    {
+      "epoch": 0.303642952508106,
+      "grad_norm": 2.7685138679199777,
+      "learning_rate": 1.9856977477337317e-05,
+      "loss": 0.8216,
+      "step": 597
+    },
+    {
+      "epoch": 0.30415156716892366,
+      "grad_norm": 2.398939247955408,
+      "learning_rate": 1.9856031023191334e-05,
+      "loss": 0.8069,
+      "step": 598
+    },
+    {
+      "epoch": 0.30466018182974125,
+      "grad_norm": 3.4947422590709643,
+      "learning_rate": 1.985508147048078e-05,
+      "loss": 0.9262,
+      "step": 599
+    },
+    {
+      "epoch": 0.30516879649055884,
+      "grad_norm": 2.5745089053386887,
+      "learning_rate": 1.985412881950418e-05,
+      "loss": 0.7747,
+      "step": 600
+    },
+    {
+      "epoch": 0.30516879649055884,
+      "eval_loss": 0.7779638171195984,
+      "eval_runtime": 169.3976,
+      "eval_samples_per_second": 19.552,
+      "eval_steps_per_second": 9.776,
+      "step": 600
+    },
+    {
+      "epoch": 0.3056774111513764,
+      "grad_norm": 3.3427093883087564,
+      "learning_rate": 1.985317307056103e-05,
+      "loss": 0.9285,
+      "step": 601
+    },
+    {
+      "epoch": 0.30618602581219406,
+      "grad_norm": 2.695577326493883,
+      "learning_rate": 1.9852214223951796e-05,
+      "loss": 0.7492,
+      "step": 602
+    },
+    {
+      "epoch": 0.30669464047301165,
+      "grad_norm": 2.5468051643392315,
+      "learning_rate": 1.9851252279977928e-05,
+      "loss": 0.8157,
+      "step": 603
+    },
+    {
+      "epoch": 0.30720325513382923,
+      "grad_norm": 2.630814697067507,
+      "learning_rate": 1.9850287238941838e-05,
+      "loss": 0.8495,
+      "step": 604
+    },
+    {
+      "epoch": 0.3077118697946468,
+      "grad_norm": 2.9295428024162686,
+      "learning_rate": 1.9849319101146927e-05,
+      "loss": 0.8773,
+      "step": 605
+    },
+    {
+      "epoch": 0.3082204844554644,
+      "grad_norm": 3.160699106938921,
+      "learning_rate": 1.9848347866897553e-05,
+      "loss": 0.8752,
+      "step": 606
+    },
+    {
+      "epoch": 0.30872909911628205,
+      "grad_norm": 2.572076719584434,
+      "learning_rate": 1.984737353649906e-05,
+      "loss": 0.7689,
+      "step": 607
+    },
+    {
+      "epoch": 0.30923771377709963,
+      "grad_norm": 2.88481752714396,
+      "learning_rate": 1.9846396110257756e-05,
+      "loss": 0.7381,
+      "step": 608
+    },
+    {
+      "epoch": 0.3097463284379172,
+      "grad_norm": 3.1448842484319126,
+      "learning_rate": 1.9845415588480933e-05,
+      "loss": 0.9054,
+      "step": 609
+    },
+    {
+      "epoch": 0.3102549430987348,
+      "grad_norm": 2.788418564604436,
+      "learning_rate": 1.9844431971476843e-05,
+      "loss": 0.9105,
+      "step": 610
+    },
+    {
+      "epoch": 0.31076355775955244,
+      "grad_norm": 2.742307133252207,
+      "learning_rate": 1.9843445259554724e-05,
+      "loss": 0.7839,
+      "step": 611
+    },
+    {
+      "epoch": 0.31127217242037003,
+      "grad_norm": 2.540629117805175,
+      "learning_rate": 1.984245545302478e-05,
+      "loss": 0.7865,
+      "step": 612
+    },
+    {
+      "epoch": 0.3117807870811876,
+      "grad_norm": 2.4535203360038755,
+      "learning_rate": 1.9841462552198182e-05,
+      "loss": 0.7452,
+      "step": 613
+    },
+    {
+      "epoch": 0.3122894017420052,
+      "grad_norm": 2.896296530437469,
+      "learning_rate": 1.984046655738709e-05,
+      "loss": 0.6764,
+      "step": 614
+    },
+    {
+      "epoch": 0.3127980164028228,
+      "grad_norm": 2.589967106201555,
+      "learning_rate": 1.9839467468904624e-05,
+      "loss": 0.9089,
+      "step": 615
+    },
+    {
+      "epoch": 0.31330663106364043,
+      "grad_norm": 2.885243948898138,
+      "learning_rate": 1.983846528706488e-05,
+      "loss": 0.7393,
+      "step": 616
+    },
+    {
+      "epoch": 0.313815245724458,
+      "grad_norm": 2.6829589803761182,
+      "learning_rate": 1.9837460012182927e-05,
+      "loss": 0.7233,
+      "step": 617
+    },
+    {
+      "epoch": 0.3143238603852756,
+      "grad_norm": 2.6174520330397026,
+      "learning_rate": 1.9836451644574805e-05,
+      "loss": 0.8009,
+      "step": 618
+    },
+    {
+      "epoch": 0.3148324750460932,
+      "grad_norm": 2.4984788081301175,
+      "learning_rate": 1.9835440184557525e-05,
+      "loss": 0.7529,
+      "step": 619
+    },
+    {
+      "epoch": 0.3153410897069108,
+      "grad_norm": 2.8455749357173348,
+      "learning_rate": 1.9834425632449076e-05,
+      "loss": 0.8223,
+      "step": 620
+    },
+    {
+      "epoch": 0.3158497043677284,
+      "grad_norm": 3.4534670284956737,
+      "learning_rate": 1.9833407988568414e-05,
+      "loss": 0.8304,
+      "step": 621
+    },
+    {
+      "epoch": 0.316358319028546,
+      "grad_norm": 2.8314989734171547,
+      "learning_rate": 1.9832387253235466e-05,
+      "loss": 0.7549,
+      "step": 622
+    },
+    {
+      "epoch": 0.3168669336893636,
+      "grad_norm": 2.8111384168897793,
+      "learning_rate": 1.9831363426771144e-05,
+      "loss": 0.7952,
+      "step": 623
+    },
+    {
+      "epoch": 0.31737554835018117,
+      "grad_norm": 2.8037282849515575,
+      "learning_rate": 1.9830336509497308e-05,
+      "loss": 0.8019,
+      "step": 624
+    },
+    {
+      "epoch": 0.3178841630109988,
+      "grad_norm": 5.43016996367062,
+      "learning_rate": 1.982930650173681e-05,
+      "loss": 0.7408,
+      "step": 625
+    },
+    {
+      "epoch": 0.3183927776718164,
+      "grad_norm": 4.177556772902112,
+      "learning_rate": 1.982827340381346e-05,
+      "loss": 0.9135,
+      "step": 626
+    },
+    {
+      "epoch": 0.318901392332634,
+      "grad_norm": 2.977275772940291,
+      "learning_rate": 1.982723721605206e-05,
+      "loss": 0.8743,
+      "step": 627
+    },
+    {
+      "epoch": 0.31941000699345157,
+      "grad_norm": 2.4971434485892137,
+      "learning_rate": 1.9826197938778354e-05,
+      "loss": 0.7503,
+      "step": 628
+    },
+    {
+      "epoch": 0.3199186216542692,
+      "grad_norm": 3.2095827089306033,
+      "learning_rate": 1.9825155572319083e-05,
+      "loss": 0.823,
+      "step": 629
+    },
+    {
+      "epoch": 0.3204272363150868,
+      "grad_norm": 2.7770949626506454,
+      "learning_rate": 1.982411011700194e-05,
+      "loss": 0.8806,
+      "step": 630
+    },
+    {
+      "epoch": 0.3209358509759044,
+      "grad_norm": 2.715868595905297,
+      "learning_rate": 1.9823061573155608e-05,
+      "loss": 0.8465,
+      "step": 631
+    },
+    {
+      "epoch": 0.32144446563672197,
+      "grad_norm": 2.7849725392492757,
+      "learning_rate": 1.982200994110972e-05,
+      "loss": 0.8504,
+      "step": 632
+    },
+    {
+      "epoch": 0.32195308029753955,
+      "grad_norm": 2.4900846309144216,
+      "learning_rate": 1.98209552211949e-05,
+      "loss": 0.7237,
+      "step": 633
+    },
+    {
+      "epoch": 0.3224616949583572,
+      "grad_norm": 3.3378371964894504,
+      "learning_rate": 1.9819897413742728e-05,
+      "loss": 0.8028,
+      "step": 634
+    },
+    {
+      "epoch": 0.3229703096191748,
+      "grad_norm": 2.6238754225036303,
+      "learning_rate": 1.9818836519085764e-05,
+      "loss": 0.7079,
+      "step": 635
+    },
+    {
+      "epoch": 0.32347892427999236,
+      "grad_norm": 3.0236428361111387,
+      "learning_rate": 1.9817772537557536e-05,
+      "loss": 0.828,
+      "step": 636
+    },
+    {
+      "epoch": 0.32398753894080995,
+      "grad_norm": 2.74529418591821,
+      "learning_rate": 1.9816705469492534e-05,
+      "loss": 0.8366,
+      "step": 637
+    },
+    {
+      "epoch": 0.3244961536016276,
+      "grad_norm": 3.1312841173700363,
+      "learning_rate": 1.981563531522623e-05,
+      "loss": 0.9035,
+      "step": 638
+    },
+    {
+      "epoch": 0.3250047682624452,
+      "grad_norm": 2.6345615633136936,
+      "learning_rate": 1.9814562075095066e-05,
+      "loss": 0.8646,
+      "step": 639
+    },
+    {
+      "epoch": 0.32551338292326276,
+      "grad_norm": 3.0348941881314904,
+      "learning_rate": 1.9813485749436444e-05,
+      "loss": 0.9318,
+      "step": 640
+    },
+    {
+      "epoch": 0.32602199758408035,
+      "grad_norm": 2.8484690988266337,
+      "learning_rate": 1.9812406338588742e-05,
+      "loss": 0.847,
+      "step": 641
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 2.74301178544503,
+      "learning_rate": 1.9811323842891313e-05,
+      "loss": 0.833,
+      "step": 642
+    },
+    {
+      "epoch": 0.3270392269057156,
+      "grad_norm": 3.0829713110564883,
+      "learning_rate": 1.9810238262684473e-05,
+      "loss": 0.8053,
+      "step": 643
+    },
+    {
+      "epoch": 0.32754784156653316,
+      "grad_norm": 3.0087530870029537,
+      "learning_rate": 1.9809149598309502e-05,
+      "loss": 0.754,
+      "step": 644
+    },
+    {
+      "epoch": 0.32805645622735075,
+      "grad_norm": 2.7309534955082824,
+      "learning_rate": 1.980805785010867e-05,
+      "loss": 0.8864,
+      "step": 645
+    },
+    {
+      "epoch": 0.32856507088816833,
+      "grad_norm": 2.6316460535628474,
+      "learning_rate": 1.980696301842519e-05,
+      "loss": 0.7913,
+      "step": 646
+    },
+    {
+      "epoch": 0.329073685548986,
+      "grad_norm": 2.454656135266429,
+      "learning_rate": 1.9805865103603274e-05,
+      "loss": 0.8706,
+      "step": 647
+    },
+    {
+      "epoch": 0.32958230020980356,
+      "grad_norm": 3.2880927534177204,
+      "learning_rate": 1.980476410598807e-05,
+      "loss": 0.9705,
+      "step": 648
+    },
+    {
+      "epoch": 0.33009091487062114,
+      "grad_norm": 2.62063908506539,
+      "learning_rate": 1.9803660025925726e-05,
+      "loss": 0.8774,
+      "step": 649
+    },
+    {
+      "epoch": 0.33059952953143873,
+      "grad_norm": 3.1727160010667976,
+      "learning_rate": 1.980255286376334e-05,
+      "loss": 0.9398,
+      "step": 650
+    },
+    {
+      "epoch": 0.3311081441922563,
+      "grad_norm": 2.7565508758261776,
+      "learning_rate": 1.9801442619848985e-05,
+      "loss": 0.7895,
+      "step": 651
+    },
+    {
+      "epoch": 0.33161675885307396,
+      "grad_norm": 2.6950122610743925,
+      "learning_rate": 1.9800329294531703e-05,
+      "loss": 0.9246,
+      "step": 652
+    },
+    {
+      "epoch": 0.33212537351389154,
+      "grad_norm": 2.725415882257643,
+      "learning_rate": 1.9799212888161497e-05,
+      "loss": 0.8215,
+      "step": 653
+    },
+    {
+      "epoch": 0.3326339881747091,
+      "grad_norm": 3.2746731194588934,
+      "learning_rate": 1.979809340108936e-05,
+      "loss": 0.8942,
+      "step": 654
+    },
+    {
+      "epoch": 0.3331426028355267,
+      "grad_norm": 3.458421330692074,
+      "learning_rate": 1.979697083366723e-05,
+      "loss": 0.7618,
+      "step": 655
+    },
+    {
+      "epoch": 0.33365121749634435,
+      "grad_norm": 2.74894173807327,
+      "learning_rate": 1.979584518624802e-05,
+      "loss": 0.6321,
+      "step": 656
+    },
+    {
+      "epoch": 0.33415983215716194,
+      "grad_norm": 2.9083221595644244,
+      "learning_rate": 1.979471645918562e-05,
+      "loss": 0.9667,
+      "step": 657
+    },
+    {
+      "epoch": 0.3346684468179795,
+      "grad_norm": 2.8285823894745246,
+      "learning_rate": 1.9793584652834882e-05,
+      "loss": 0.812,
+      "step": 658
+    },
+    {
+      "epoch": 0.3351770614787971,
+      "grad_norm": 4.709280048266194,
+      "learning_rate": 1.9792449767551622e-05,
+      "loss": 0.8644,
+      "step": 659
+    },
+    {
+      "epoch": 0.3356856761396147,
+      "grad_norm": 4.273230399460694,
+      "learning_rate": 1.979131180369263e-05,
+      "loss": 0.9151,
+      "step": 660
+    },
+    {
+      "epoch": 0.33619429080043234,
+      "grad_norm": 2.96151220536764,
+      "learning_rate": 1.9790170761615665e-05,
+      "loss": 0.8454,
+      "step": 661
+    },
+    {
+      "epoch": 0.3367029054612499,
+      "grad_norm": 3.181467818501953,
+      "learning_rate": 1.9789026641679446e-05,
+      "loss": 0.8406,
+      "step": 662
+    },
+    {
+      "epoch": 0.3372115201220675,
+      "grad_norm": 2.5172110349856087,
+      "learning_rate": 1.9787879444243668e-05,
+      "loss": 0.7432,
+      "step": 663
+    },
+    {
+      "epoch": 0.3377201347828851,
+      "grad_norm": 2.5777225064604026,
+      "learning_rate": 1.978672916966899e-05,
+      "loss": 0.7224,
+      "step": 664
+    },
+    {
+      "epoch": 0.33822874944370274,
+      "grad_norm": 2.7889650472076375,
+      "learning_rate": 1.9785575818317034e-05,
+      "loss": 0.855,
+      "step": 665
+    },
+    {
+      "epoch": 0.3387373641045203,
+      "grad_norm": 2.475199038276165,
+      "learning_rate": 1.9784419390550397e-05,
+      "loss": 0.6718,
+      "step": 666
+    },
+    {
+      "epoch": 0.3392459787653379,
+      "grad_norm": 3.0341653863418196,
+      "learning_rate": 1.978325988673264e-05,
+      "loss": 0.8083,
+      "step": 667
+    },
+    {
+      "epoch": 0.3397545934261555,
+      "grad_norm": 2.7424222405536676,
+      "learning_rate": 1.9782097307228293e-05,
+      "loss": 0.7916,
+      "step": 668
+    },
+    {
+      "epoch": 0.34026320808697313,
+      "grad_norm": 2.5056546158062045,
+      "learning_rate": 1.9780931652402842e-05,
+      "loss": 0.8795,
+      "step": 669
+    },
+    {
+      "epoch": 0.3407718227477907,
+      "grad_norm": 3.1022622498580823,
+      "learning_rate": 1.977976292262276e-05,
+      "loss": 0.7693,
+      "step": 670
+    },
+    {
+      "epoch": 0.3412804374086083,
+      "grad_norm": 2.881629266828978,
+      "learning_rate": 1.9778591118255465e-05,
+      "loss": 0.7356,
+      "step": 671
+    },
+    {
+      "epoch": 0.3417890520694259,
+      "grad_norm": 2.778765136293138,
+      "learning_rate": 1.977741623966936e-05,
+      "loss": 0.8223,
+      "step": 672
+    },
+    {
+      "epoch": 0.3422976667302435,
+      "grad_norm": 2.5514438580543413,
+      "learning_rate": 1.97762382872338e-05,
+      "loss": 0.7725,
+      "step": 673
+    },
+    {
+      "epoch": 0.3428062813910611,
+      "grad_norm": 2.675961052638251,
+      "learning_rate": 1.9775057261319116e-05,
+      "loss": 0.7259,
+      "step": 674
+    },
+    {
+      "epoch": 0.3433148960518787,
+      "grad_norm": 3.1951943074870512,
+      "learning_rate": 1.9773873162296605e-05,
+      "loss": 0.862,
+      "step": 675
+    },
+    {
+      "epoch": 0.3438235107126963,
+      "grad_norm": 2.8574587545146546,
+      "learning_rate": 1.977268599053852e-05,
+      "loss": 0.7941,
+      "step": 676
+    },
+    {
+      "epoch": 0.3443321253735139,
+      "grad_norm": 2.5919770765291665,
+      "learning_rate": 1.9771495746418093e-05,
+      "loss": 0.7542,
+      "step": 677
+    },
+    {
+      "epoch": 0.3448407400343315,
+      "grad_norm": 2.841192919337759,
+      "learning_rate": 1.9770302430309517e-05,
+      "loss": 0.9419,
+      "step": 678
+    },
+    {
+      "epoch": 0.3453493546951491,
+      "grad_norm": 2.6228003595535476,
+      "learning_rate": 1.976910604258794e-05,
+      "loss": 0.7372,
+      "step": 679
+    },
+    {
+      "epoch": 0.3458579693559667,
+      "grad_norm": 2.607081336025282,
+      "learning_rate": 1.9767906583629496e-05,
+      "loss": 0.8434,
+      "step": 680
+    },
+    {
+      "epoch": 0.3463665840167843,
+      "grad_norm": 2.7749975840240135,
+      "learning_rate": 1.976670405381127e-05,
+      "loss": 0.8543,
+      "step": 681
+    },
+    {
+      "epoch": 0.34687519867760186,
+      "grad_norm": 2.6980228673193496,
+      "learning_rate": 1.9765498453511317e-05,
+      "loss": 0.9023,
+      "step": 682
+    },
+    {
+      "epoch": 0.3473838133384195,
+      "grad_norm": 2.504119721119131,
+      "learning_rate": 1.9764289783108658e-05,
+      "loss": 0.8394,
+      "step": 683
+    },
+    {
+      "epoch": 0.3478924279992371,
+      "grad_norm": 2.847905905497902,
+      "learning_rate": 1.9763078042983273e-05,
+      "loss": 0.953,
+      "step": 684
+    },
+    {
+      "epoch": 0.34840104266005467,
+      "grad_norm": 2.785284994394649,
+      "learning_rate": 1.9761863233516118e-05,
+      "loss": 0.7947,
+      "step": 685
+    },
+    {
+      "epoch": 0.34890965732087226,
+      "grad_norm": 3.7344582520466467,
+      "learning_rate": 1.97606453550891e-05,
+      "loss": 0.8478,
+      "step": 686
+    },
+    {
+      "epoch": 0.3494182719816899,
+      "grad_norm": 2.6397519445924256,
+      "learning_rate": 1.975942440808511e-05,
+      "loss": 0.908,
+      "step": 687
+    },
+    {
+      "epoch": 0.3499268866425075,
+      "grad_norm": 2.3758043633653507,
+      "learning_rate": 1.9758200392887986e-05,
+      "loss": 0.7087,
+      "step": 688
+    },
+    {
+      "epoch": 0.35043550130332507,
+      "grad_norm": 2.902748870300068,
+      "learning_rate": 1.9756973309882535e-05,
+      "loss": 0.9096,
+      "step": 689
+    },
+    {
+      "epoch": 0.35094411596414266,
+      "grad_norm": 2.826164282503191,
+      "learning_rate": 1.9755743159454537e-05,
+      "loss": 0.842,
+      "step": 690
+    },
+    {
+      "epoch": 0.35145273062496024,
+      "grad_norm": 2.802477462637098,
+      "learning_rate": 1.9754509941990724e-05,
+      "loss": 0.6991,
+      "step": 691
+    },
+    {
+      "epoch": 0.3519613452857779,
+      "grad_norm": 3.298713063231653,
+      "learning_rate": 1.97532736578788e-05,
+      "loss": 0.8347,
+      "step": 692
+    },
+    {
+      "epoch": 0.35246995994659547,
+      "grad_norm": 2.627668659085406,
+      "learning_rate": 1.975203430750743e-05,
+      "loss": 0.7581,
+      "step": 693
+    },
+    {
+      "epoch": 0.35297857460741305,
+      "grad_norm": 2.6349037083326325,
+      "learning_rate": 1.9750791891266247e-05,
+      "loss": 0.8786,
+      "step": 694
+    },
+    {
+      "epoch": 0.35348718926823064,
+      "grad_norm": 2.911360482543164,
+      "learning_rate": 1.9749546409545845e-05,
+      "loss": 0.865,
+      "step": 695
+    },
+    {
+      "epoch": 0.3539958039290483,
+      "grad_norm": 2.727279787343453,
+      "learning_rate": 1.974829786273778e-05,
+      "loss": 0.8436,
+      "step": 696
+    },
+    {
+      "epoch": 0.35450441858986587,
+      "grad_norm": 2.7458152141368775,
+      "learning_rate": 1.974704625123457e-05,
+      "loss": 0.8916,
+      "step": 697
+    },
+    {
+      "epoch": 0.35501303325068345,
+      "grad_norm": 2.3974370702732553,
+      "learning_rate": 1.9745791575429707e-05,
+      "loss": 0.6801,
+      "step": 698
+    },
+    {
+      "epoch": 0.35552164791150104,
+      "grad_norm": 2.5293388215935337,
+      "learning_rate": 1.9744533835717635e-05,
+      "loss": 0.8189,
+      "step": 699
+    },
+    {
+      "epoch": 0.3560302625723186,
+      "grad_norm": 2.601853546266719,
+      "learning_rate": 1.9743273032493767e-05,
+      "loss": 0.8654,
+      "step": 700
+    },
+    {
+      "epoch": 0.3560302625723186,
+      "eval_loss": 0.7728118300437927,
+      "eval_runtime": 169.6795,
+      "eval_samples_per_second": 19.519,
+      "eval_steps_per_second": 9.76,
+      "step": 700
+    },
+    {
+      "epoch": 0.35653887723313626,
+      "grad_norm": 2.5223848491657703,
+      "learning_rate": 1.9742009166154477e-05,
+      "loss": 0.7916,
+      "step": 701
+    },
+    {
+      "epoch": 0.35704749189395385,
+      "grad_norm": 2.684527150150928,
+      "learning_rate": 1.9740742237097103e-05,
+      "loss": 0.7982,
+      "step": 702
+    },
+    {
+      "epoch": 0.35755610655477144,
+      "grad_norm": 2.5149320038842755,
+      "learning_rate": 1.9739472245719945e-05,
+      "loss": 0.7106,
+      "step": 703
+    },
+    {
+      "epoch": 0.358064721215589,
+      "grad_norm": 2.9297360022026795,
+      "learning_rate": 1.9738199192422264e-05,
+      "loss": 0.8303,
+      "step": 704
+    },
+    {
+      "epoch": 0.35857333587640666,
+      "grad_norm": 2.8040124617447972,
+      "learning_rate": 1.9736923077604285e-05,
+      "loss": 0.8678,
+      "step": 705
+    },
+    {
+      "epoch": 0.35908195053722425,
+      "grad_norm": 2.800608269152631,
+      "learning_rate": 1.9735643901667205e-05,
+      "loss": 0.7591,
+      "step": 706
+    },
+    {
+      "epoch": 0.35959056519804183,
+      "grad_norm": 2.581386098899069,
+      "learning_rate": 1.9734361665013165e-05,
+      "loss": 0.841,
+      "step": 707
+    },
+    {
+      "epoch": 0.3600991798588594,
+      "grad_norm": 2.8126702998842275,
+      "learning_rate": 1.9733076368045286e-05,
+      "loss": 0.858,
+      "step": 708
+    },
+    {
+      "epoch": 0.360607794519677,
+      "grad_norm": 2.6905688130430203,
+      "learning_rate": 1.9731788011167635e-05,
+      "loss": 0.7923,
+      "step": 709
+    },
+    {
+      "epoch": 0.36111640918049465,
+      "grad_norm": 3.740245593977952,
+      "learning_rate": 1.9730496594785254e-05,
+      "loss": 0.8485,
+      "step": 710
+    },
+    {
+      "epoch": 0.36162502384131223,
+      "grad_norm": 3.0204595049483305,
+      "learning_rate": 1.972920211930414e-05,
+      "loss": 0.9207,
+      "step": 711
+    },
+    {
+      "epoch": 0.3621336385021298,
+      "grad_norm": 2.514396794144337,
+      "learning_rate": 1.9727904585131253e-05,
+      "loss": 0.7782,
+      "step": 712
+    },
+    {
+      "epoch": 0.3626422531629474,
+      "grad_norm": 2.9045208124408997,
+      "learning_rate": 1.972660399267452e-05,
+      "loss": 0.8178,
+      "step": 713
+    },
+    {
+      "epoch": 0.36315086782376504,
+      "grad_norm": 2.916815308915819,
+      "learning_rate": 1.972530034234282e-05,
+      "loss": 0.8792,
+      "step": 714
+    },
+    {
+      "epoch": 0.36365948248458263,
+      "grad_norm": 2.744003086105035,
+      "learning_rate": 1.9723993634545995e-05,
+      "loss": 0.8858,
+      "step": 715
+    },
+    {
+      "epoch": 0.3641680971454002,
+      "grad_norm": 2.6095891624782297,
+      "learning_rate": 1.9722683869694857e-05,
+      "loss": 0.8448,
+      "step": 716
+    },
+    {
+      "epoch": 0.3646767118062178,
+      "grad_norm": 3.32150182628567,
+      "learning_rate": 1.9721371048201173e-05,
+      "loss": 0.9535,
+      "step": 717
+    },
+    {
+      "epoch": 0.3651853264670354,
+      "grad_norm": 2.406979338851643,
+      "learning_rate": 1.9720055170477667e-05,
+      "loss": 0.87,
+      "step": 718
+    },
+    {
+      "epoch": 0.36569394112785303,
+      "grad_norm": 4.846365536268681,
+      "learning_rate": 1.971873623693803e-05,
+      "loss": 0.8823,
+      "step": 719
+    },
+    {
+      "epoch": 0.3662025557886706,
+      "grad_norm": 2.5674264584179944,
+      "learning_rate": 1.9717414247996912e-05,
+      "loss": 0.7688,
+      "step": 720
+    },
+    {
+      "epoch": 0.3667111704494882,
+      "grad_norm": 2.842302226330054,
+      "learning_rate": 1.9716089204069928e-05,
+      "loss": 0.9135,
+      "step": 721
+    },
+    {
+      "epoch": 0.3672197851103058,
+      "grad_norm": 2.5666649628505582,
+      "learning_rate": 1.971476110557364e-05,
+      "loss": 0.7785,
+      "step": 722
+    },
+    {
+      "epoch": 0.3677283997711234,
+      "grad_norm": 2.6943342562120978,
+      "learning_rate": 1.9713429952925587e-05,
+      "loss": 0.8883,
+      "step": 723
+    },
+    {
+      "epoch": 0.368237014431941,
+      "grad_norm": 2.80660290782329,
+      "learning_rate": 1.9712095746544255e-05,
+      "loss": 0.8789,
+      "step": 724
+    },
+    {
+      "epoch": 0.3687456290927586,
+      "grad_norm": 2.8622188913169313,
+      "learning_rate": 1.9710758486849096e-05,
+      "loss": 0.8495,
+      "step": 725
+    },
+    {
+      "epoch": 0.3692542437535762,
+      "grad_norm": 2.492558179690955,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.7114,
+      "step": 726
+    },
+    {
+      "epoch": 0.36976285841439377,
+      "grad_norm": 3.076572394560968,
+      "learning_rate": 1.9708074809199903e-05,
+      "loss": 0.8345,
+      "step": 727
+    },
+    {
+      "epoch": 0.3702714730752114,
+      "grad_norm": 3.0306279438843906,
+      "learning_rate": 1.9706728392089575e-05,
+      "loss": 0.8504,
+      "step": 728
+    },
+    {
+      "epoch": 0.370780087736029,
+      "grad_norm": 2.5781599562615973,
+      "learning_rate": 1.970537892335282e-05,
+      "loss": 0.8402,
+      "step": 729
+    },
+    {
+      "epoch": 0.3712887023968466,
+      "grad_norm": 2.5898703549314557,
+      "learning_rate": 1.970402640341389e-05,
+      "loss": 0.7412,
+      "step": 730
+    },
+    {
+      "epoch": 0.37179731705766417,
+      "grad_norm": 2.8943401484655413,
+      "learning_rate": 1.9702670832697997e-05,
+      "loss": 0.8828,
+      "step": 731
+    },
+    {
+      "epoch": 0.3723059317184818,
+      "grad_norm": 2.2687431188221554,
+      "learning_rate": 1.9701312211631304e-05,
+      "loss": 0.7144,
+      "step": 732
+    },
+    {
+      "epoch": 0.3728145463792994,
+      "grad_norm": 2.4934555224025368,
+      "learning_rate": 1.9699950540640945e-05,
+      "loss": 0.8284,
+      "step": 733
+    },
+    {
+      "epoch": 0.373323161040117,
+      "grad_norm": 3.2705610287455307,
+      "learning_rate": 1.9698585820155e-05,
+      "loss": 0.8707,
+      "step": 734
+    },
+    {
+      "epoch": 0.37383177570093457,
+      "grad_norm": 3.0923187335956306,
+      "learning_rate": 1.9697218050602515e-05,
+      "loss": 0.754,
+      "step": 735
+    },
+    {
+      "epoch": 0.37434039036175215,
+      "grad_norm": 2.880630121969353,
+      "learning_rate": 1.9695847232413488e-05,
+      "loss": 0.8286,
+      "step": 736
+    },
+    {
+      "epoch": 0.3748490050225698,
+      "grad_norm": 2.5098515952262623,
+      "learning_rate": 1.9694473366018887e-05,
+      "loss": 0.9182,
+      "step": 737
+    },
+    {
+      "epoch": 0.3753576196833874,
+      "grad_norm": 3.962945854061612,
+      "learning_rate": 1.9693096451850632e-05,
+      "loss": 0.699,
+      "step": 738
+    },
+    {
+      "epoch": 0.37586623434420496,
+      "grad_norm": 2.412495323532395,
+      "learning_rate": 1.9691716490341593e-05,
+      "loss": 0.698,
+      "step": 739
+    },
+    {
+      "epoch": 0.37637484900502255,
+      "grad_norm": 3.0205297558712307,
+      "learning_rate": 1.9690333481925616e-05,
+      "loss": 0.9651,
+      "step": 740
+    },
+    {
+      "epoch": 0.3768834636658402,
+      "grad_norm": 2.524239994149761,
+      "learning_rate": 1.968894742703749e-05,
+      "loss": 0.8567,
+      "step": 741
+    },
+    {
+      "epoch": 0.3773920783266578,
+      "grad_norm": 2.530539330454778,
+      "learning_rate": 1.9687558326112963e-05,
+      "loss": 0.831,
+      "step": 742
+    },
+    {
+      "epoch": 0.37790069298747536,
+      "grad_norm": 2.531893518851455,
+      "learning_rate": 1.9686166179588754e-05,
+      "loss": 0.863,
+      "step": 743
+    },
+    {
+      "epoch": 0.37840930764829295,
+      "grad_norm": 2.560327958468081,
+      "learning_rate": 1.968477098790252e-05,
+      "loss": 0.8475,
+      "step": 744
+    },
+    {
+      "epoch": 0.37891792230911053,
+      "grad_norm": 2.5534106688909928,
+      "learning_rate": 1.9683372751492884e-05,
+      "loss": 0.7562,
+      "step": 745
+    },
+    {
+      "epoch": 0.3794265369699282,
+      "grad_norm": 2.3950912542618195,
+      "learning_rate": 1.9681971470799435e-05,
+      "loss": 0.7454,
+      "step": 746
+    },
+    {
+      "epoch": 0.37993515163074576,
+      "grad_norm": 2.3323288412966323,
+      "learning_rate": 1.9680567146262704e-05,
+      "loss": 0.838,
+      "step": 747
+    },
+    {
+      "epoch": 0.38044376629156335,
+      "grad_norm": 2.6370376507093227,
+      "learning_rate": 1.9679159778324197e-05,
+      "loss": 0.7495,
+      "step": 748
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 2.488128896986653,
+      "learning_rate": 1.9677749367426355e-05,
+      "loss": 0.7742,
+      "step": 749
+    },
+    {
+      "epoch": 0.3814609956131986,
+      "grad_norm": 2.523822011085595,
+      "learning_rate": 1.9676335914012588e-05,
+      "loss": 0.7738,
+      "step": 750
+    },
+    {
+      "epoch": 0.38196961027401616,
+      "grad_norm": 2.7607525507189647,
+      "learning_rate": 1.967491941852727e-05,
+      "loss": 0.8003,
+      "step": 751
+    },
+    {
+      "epoch": 0.38247822493483374,
+      "grad_norm": 2.7074003422114346,
+      "learning_rate": 1.9673499881415713e-05,
+      "loss": 0.867,
+      "step": 752
+    },
+    {
+      "epoch": 0.38298683959565133,
+      "grad_norm": 2.513806556585185,
+      "learning_rate": 1.9672077303124196e-05,
+      "loss": 0.8217,
+      "step": 753
+    },
+    {
+      "epoch": 0.3834954542564689,
+      "grad_norm": 2.599238460226402,
+      "learning_rate": 1.9670651684099957e-05,
+      "loss": 0.7814,
+      "step": 754
+    },
+    {
+      "epoch": 0.38400406891728656,
+      "grad_norm": 2.5887721449689027,
+      "learning_rate": 1.966922302479118e-05,
+      "loss": 0.7459,
+      "step": 755
+    },
+    {
+      "epoch": 0.38451268357810414,
+      "grad_norm": 2.540538933716716,
+      "learning_rate": 1.9667791325647018e-05,
+      "loss": 0.7469,
+      "step": 756
+    },
+    {
+      "epoch": 0.3850212982389217,
+      "grad_norm": 2.75947188037743,
+      "learning_rate": 1.966635658711757e-05,
+      "loss": 0.8233,
+      "step": 757
+    },
+    {
+      "epoch": 0.3855299128997393,
+      "grad_norm": 2.557737344000031,
+      "learning_rate": 1.966491880965389e-05,
+      "loss": 0.8708,
+      "step": 758
+    },
+    {
+      "epoch": 0.38603852756055695,
+      "grad_norm": 2.719969465000195,
+      "learning_rate": 1.9663477993707987e-05,
+      "loss": 0.8857,
+      "step": 759
+    },
+    {
+      "epoch": 0.38654714222137454,
+      "grad_norm": 2.6679270169972193,
+      "learning_rate": 1.966203413973284e-05,
+      "loss": 0.7932,
+      "step": 760
+    },
+    {
+      "epoch": 0.3870557568821921,
+      "grad_norm": 2.8049720896047994,
+      "learning_rate": 1.966058724818236e-05,
+      "loss": 0.737,
+      "step": 761
+    },
+    {
+      "epoch": 0.3875643715430097,
+      "grad_norm": 3.377947615095893,
+      "learning_rate": 1.965913731951143e-05,
+      "loss": 0.8168,
+      "step": 762
+    },
+    {
+      "epoch": 0.38807298620382735,
+      "grad_norm": 2.8737131638922553,
+      "learning_rate": 1.9657684354175883e-05,
+      "loss": 0.8626,
+      "step": 763
+    },
+    {
+      "epoch": 0.38858160086464494,
+      "grad_norm": 2.7724037500431233,
+      "learning_rate": 1.96562283526325e-05,
+      "loss": 0.8786,
+      "step": 764
+    },
+    {
+      "epoch": 0.3890902155254625,
+      "grad_norm": 2.6883342246589987,
+      "learning_rate": 1.965476931533903e-05,
+      "loss": 0.7079,
+      "step": 765
+    },
+    {
+      "epoch": 0.3895988301862801,
+      "grad_norm": 2.4373672644836106,
+      "learning_rate": 1.9653307242754164e-05,
+      "loss": 0.8007,
+      "step": 766
+    },
+    {
+      "epoch": 0.3901074448470977,
+      "grad_norm": 2.4540840752460915,
+      "learning_rate": 1.9651842135337554e-05,
+      "loss": 0.7609,
+      "step": 767
+    },
+    {
+      "epoch": 0.39061605950791534,
+      "grad_norm": 2.707905065469876,
+      "learning_rate": 1.9650373993549806e-05,
+      "loss": 0.8061,
+      "step": 768
+    },
+    {
+      "epoch": 0.3911246741687329,
+      "grad_norm": 2.3305452810593215,
+      "learning_rate": 1.964890281785248e-05,
+      "loss": 0.7662,
+      "step": 769
+    },
+    {
+      "epoch": 0.3916332888295505,
+      "grad_norm": 2.7581677086499483,
+      "learning_rate": 1.964742860870808e-05,
+      "loss": 0.9251,
+      "step": 770
+    },
+    {
+      "epoch": 0.3921419034903681,
+      "grad_norm": 2.5445457057524687,
+      "learning_rate": 1.9645951366580075e-05,
+      "loss": 0.7927,
+      "step": 771
+    },
+    {
+      "epoch": 0.39265051815118573,
+      "grad_norm": 2.491248835952418,
+      "learning_rate": 1.9644471091932886e-05,
+      "loss": 0.7968,
+      "step": 772
+    },
+    {
+      "epoch": 0.3931591328120033,
+      "grad_norm": 2.8379563220952866,
+      "learning_rate": 1.9642987785231888e-05,
+      "loss": 0.8256,
+      "step": 773
+    },
+    {
+      "epoch": 0.3936677474728209,
+      "grad_norm": 3.1991670299257224,
+      "learning_rate": 1.9641501446943404e-05,
+      "loss": 0.9771,
+      "step": 774
+    },
+    {
+      "epoch": 0.3941763621336385,
+      "grad_norm": 2.5513838659338695,
+      "learning_rate": 1.9640012077534707e-05,
+      "loss": 0.9148,
+      "step": 775
+    },
+    {
+      "epoch": 0.3946849767944561,
+      "grad_norm": 2.4365614037623184,
+      "learning_rate": 1.9638519677474038e-05,
+      "loss": 0.7444,
+      "step": 776
+    },
+    {
+      "epoch": 0.3951935914552737,
+      "grad_norm": 2.4657205270651277,
+      "learning_rate": 1.9637024247230578e-05,
+      "loss": 0.7639,
+      "step": 777
+    },
+    {
+      "epoch": 0.3957022061160913,
+      "grad_norm": 2.5457074818406307,
+      "learning_rate": 1.9635525787274465e-05,
+      "loss": 0.7719,
+      "step": 778
+    },
+    {
+      "epoch": 0.3962108207769089,
+      "grad_norm": 2.231974666253432,
+      "learning_rate": 1.9634024298076787e-05,
+      "loss": 0.8665,
+      "step": 779
+    },
+    {
+      "epoch": 0.3967194354377265,
+      "grad_norm": 2.4092520280421565,
+      "learning_rate": 1.963251978010959e-05,
+      "loss": 0.6592,
+      "step": 780
+    },
+    {
+      "epoch": 0.3972280500985441,
+      "grad_norm": 3.0714377944909197,
+      "learning_rate": 1.963101223384586e-05,
+      "loss": 0.7548,
+      "step": 781
+    },
+    {
+      "epoch": 0.3977366647593617,
+      "grad_norm": 2.542492706228351,
+      "learning_rate": 1.9629501659759555e-05,
+      "loss": 0.8191,
+      "step": 782
+    },
+    {
+      "epoch": 0.3982452794201793,
+      "grad_norm": 2.6638892332767834,
+      "learning_rate": 1.9627988058325564e-05,
+      "loss": 0.745,
+      "step": 783
+    },
+    {
+      "epoch": 0.3987538940809969,
+      "grad_norm": 2.5562575062685866,
+      "learning_rate": 1.962647143001974e-05,
+      "loss": 0.8401,
+      "step": 784
+    },
+    {
+      "epoch": 0.39926250874181446,
+      "grad_norm": 2.3330041335257725,
+      "learning_rate": 1.962495177531888e-05,
+      "loss": 0.943,
+      "step": 785
+    },
+    {
+      "epoch": 0.3997711234026321,
+      "grad_norm": 2.786053057042833,
+      "learning_rate": 1.9623429094700745e-05,
+      "loss": 0.6912,
+      "step": 786
+    },
+    {
+      "epoch": 0.4002797380634497,
+      "grad_norm": 2.482832999892015,
+      "learning_rate": 1.9621903388644037e-05,
+      "loss": 0.8413,
+      "step": 787
+    },
+    {
+      "epoch": 0.40078835272426727,
+      "grad_norm": 2.518190012130702,
+      "learning_rate": 1.962037465762841e-05,
+      "loss": 0.8,
+      "step": 788
+    },
+    {
+      "epoch": 0.40129696738508486,
+      "grad_norm": 2.540317635987954,
+      "learning_rate": 1.9618842902134467e-05,
+      "loss": 0.9279,
+      "step": 789
+    },
+    {
+      "epoch": 0.4018055820459025,
+      "grad_norm": 3.1285801025787965,
+      "learning_rate": 1.9617308122643772e-05,
+      "loss": 0.7106,
+      "step": 790
+    },
+    {
+      "epoch": 0.4023141967067201,
+      "grad_norm": 3.5793140551262903,
+      "learning_rate": 1.961577031963883e-05,
+      "loss": 0.7879,
+      "step": 791
+    },
+    {
+      "epoch": 0.40282281136753767,
+      "grad_norm": 2.860650116983354,
+      "learning_rate": 1.9614229493603095e-05,
+      "loss": 0.8436,
+      "step": 792
+    },
+    {
+      "epoch": 0.40333142602835526,
+      "grad_norm": 2.62237226569689,
+      "learning_rate": 1.9612685645020985e-05,
+      "loss": 0.8164,
+      "step": 793
+    },
+    {
+      "epoch": 0.40384004068917284,
+      "grad_norm": 2.4722242414868667,
+      "learning_rate": 1.961113877437785e-05,
+      "loss": 0.7134,
+      "step": 794
+    },
+    {
+      "epoch": 0.4043486553499905,
+      "grad_norm": 2.466393227070069,
+      "learning_rate": 1.9609588882160007e-05,
+      "loss": 0.7448,
+      "step": 795
+    },
+    {
+      "epoch": 0.40485727001080807,
+      "grad_norm": 2.4766901506805463,
+      "learning_rate": 1.9608035968854708e-05,
+      "loss": 0.7543,
+      "step": 796
+    },
+    {
+      "epoch": 0.40536588467162565,
+      "grad_norm": 2.510148431105061,
+      "learning_rate": 1.960648003495017e-05,
+      "loss": 0.8228,
+      "step": 797
+    },
+    {
+      "epoch": 0.40587449933244324,
+      "grad_norm": 2.4070623438534504,
+      "learning_rate": 1.9604921080935543e-05,
+      "loss": 0.8378,
+      "step": 798
+    },
+    {
+      "epoch": 0.4063831139932609,
+      "grad_norm": 2.431838380669753,
+      "learning_rate": 1.960335910730094e-05,
+      "loss": 0.7104,
+      "step": 799
+    },
+    {
+      "epoch": 0.40689172865407847,
+      "grad_norm": 2.9158256044709665,
+      "learning_rate": 1.9601794114537414e-05,
+      "loss": 0.8148,
+      "step": 800
+    },
+    {
+      "epoch": 0.40689172865407847,
+      "eval_loss": 0.7690863013267517,
+      "eval_runtime": 170.0707,
+      "eval_samples_per_second": 19.474,
+      "eval_steps_per_second": 9.737,
+      "step": 800
+    },
+    {
+      "epoch": 0.40740034331489605,
+      "grad_norm": 2.1988613996971536,
+      "learning_rate": 1.9600226103136978e-05,
+      "loss": 0.7855,
+      "step": 801
+    },
+    {
+      "epoch": 0.40790895797571364,
+      "grad_norm": 2.7127906116261267,
+      "learning_rate": 1.9598655073592583e-05,
+      "loss": 0.8347,
+      "step": 802
+    },
+    {
+      "epoch": 0.4084175726365312,
+      "grad_norm": 2.7485028439681085,
+      "learning_rate": 1.9597081026398138e-05,
+      "loss": 0.8565,
+      "step": 803
+    },
+    {
+      "epoch": 0.40892618729734886,
+      "grad_norm": 2.480125391466621,
+      "learning_rate": 1.959550396204849e-05,
+      "loss": 0.7956,
+      "step": 804
+    },
+    {
+      "epoch": 0.40943480195816645,
+      "grad_norm": 3.279340206402967,
+      "learning_rate": 1.959392388103944e-05,
+      "loss": 0.8317,
+      "step": 805
+    },
+    {
+      "epoch": 0.40994341661898404,
+      "grad_norm": 2.382469177527058,
+      "learning_rate": 1.9592340783867747e-05,
+      "loss": 0.9162,
+      "step": 806
+    },
+    {
+      "epoch": 0.4104520312798016,
+      "grad_norm": 2.2441740905019434,
+      "learning_rate": 1.95907546710311e-05,
+      "loss": 0.7934,
+      "step": 807
+    },
+    {
+      "epoch": 0.41096064594061926,
+      "grad_norm": 2.8457523883647693,
+      "learning_rate": 1.9589165543028147e-05,
+      "loss": 0.7585,
+      "step": 808
+    },
+    {
+      "epoch": 0.41146926060143685,
+      "grad_norm": 2.421484638871931,
+      "learning_rate": 1.958757340035849e-05,
+      "loss": 0.7084,
+      "step": 809
+    },
+    {
+      "epoch": 0.41197787526225443,
+      "grad_norm": 2.421209822337622,
+      "learning_rate": 1.958597824352266e-05,
+      "loss": 0.8252,
+      "step": 810
+    },
+    {
+      "epoch": 0.412486489923072,
+      "grad_norm": 2.658036831510891,
+      "learning_rate": 1.9584380073022152e-05,
+      "loss": 0.8526,
+      "step": 811
+    },
+    {
+      "epoch": 0.4129951045838896,
+      "grad_norm": 2.8721742426367376,
+      "learning_rate": 1.9582778889359402e-05,
+      "loss": 0.7484,
+      "step": 812
+    },
+    {
+      "epoch": 0.41350371924470725,
+      "grad_norm": 2.7499923705948857,
+      "learning_rate": 1.9581174693037794e-05,
+      "loss": 0.7938,
+      "step": 813
+    },
+    {
+      "epoch": 0.41401233390552483,
+      "grad_norm": 2.6891943447634032,
+      "learning_rate": 1.957956748456166e-05,
+      "loss": 0.9328,
+      "step": 814
+    },
+    {
+      "epoch": 0.4145209485663424,
+      "grad_norm": 2.2144783700906,
+      "learning_rate": 1.957795726443628e-05,
+      "loss": 0.8163,
+      "step": 815
+    },
+    {
+      "epoch": 0.41502956322716,
+      "grad_norm": 2.4458594152391036,
+      "learning_rate": 1.9576344033167877e-05,
+      "loss": 0.8415,
+      "step": 816
+    },
+    {
+      "epoch": 0.41553817788797764,
+      "grad_norm": 2.555565546080738,
+      "learning_rate": 1.9574727791263622e-05,
+      "loss": 0.7966,
+      "step": 817
+    },
+    {
+      "epoch": 0.41604679254879523,
+      "grad_norm": 2.592750448638668,
+      "learning_rate": 1.9573108539231636e-05,
+      "loss": 0.8754,
+      "step": 818
+    },
+    {
+      "epoch": 0.4165554072096128,
+      "grad_norm": 2.8304036271144812,
+      "learning_rate": 1.9571486277580976e-05,
+      "loss": 0.849,
+      "step": 819
+    },
+    {
+      "epoch": 0.4170640218704304,
+      "grad_norm": 2.402651486419046,
+      "learning_rate": 1.956986100682167e-05,
+      "loss": 0.7645,
+      "step": 820
+    },
+    {
+      "epoch": 0.417572636531248,
+      "grad_norm": 2.6953556665474507,
+      "learning_rate": 1.9568232727464656e-05,
+      "loss": 0.8696,
+      "step": 821
+    },
+    {
+      "epoch": 0.41808125119206563,
+      "grad_norm": 2.4723334079080836,
+      "learning_rate": 1.9566601440021845e-05,
+      "loss": 0.8847,
+      "step": 822
+    },
+    {
+      "epoch": 0.4185898658528832,
+      "grad_norm": 3.6895700883475047,
+      "learning_rate": 1.9564967145006087e-05,
+      "loss": 0.9179,
+      "step": 823
+    },
+    {
+      "epoch": 0.4190984805137008,
+      "grad_norm": 2.4555385889164345,
+      "learning_rate": 1.9563329842931176e-05,
+      "loss": 0.8165,
+      "step": 824
+    },
+    {
+      "epoch": 0.4196070951745184,
+      "grad_norm": 2.335954238678531,
+      "learning_rate": 1.9561689534311845e-05,
+      "loss": 0.8073,
+      "step": 825
+    },
+    {
+      "epoch": 0.420115709835336,
+      "grad_norm": 2.9537233422650813,
+      "learning_rate": 1.9560046219663783e-05,
+      "loss": 0.9297,
+      "step": 826
+    },
+    {
+      "epoch": 0.4206243244961536,
+      "grad_norm": 2.5265132712886267,
+      "learning_rate": 1.9558399899503622e-05,
+      "loss": 0.8322,
+      "step": 827
+    },
+    {
+      "epoch": 0.4211329391569712,
+      "grad_norm": 2.30260743494211,
+      "learning_rate": 1.955675057434893e-05,
+      "loss": 0.7741,
+      "step": 828
+    },
+    {
+      "epoch": 0.4216415538177888,
+      "grad_norm": 2.3244658523122412,
+      "learning_rate": 1.955509824471823e-05,
+      "loss": 0.8367,
+      "step": 829
+    },
+    {
+      "epoch": 0.42215016847860637,
+      "grad_norm": 2.230153397215058,
+      "learning_rate": 1.955344291113099e-05,
+      "loss": 0.7999,
+      "step": 830
+    },
+    {
+      "epoch": 0.422658783139424,
+      "grad_norm": 2.411700330638582,
+      "learning_rate": 1.955178457410761e-05,
+      "loss": 0.8194,
+      "step": 831
+    },
+    {
+      "epoch": 0.4231673978002416,
+      "grad_norm": 2.2220615871969307,
+      "learning_rate": 1.9550123234169445e-05,
+      "loss": 0.81,
+      "step": 832
+    },
+    {
+      "epoch": 0.4236760124610592,
+      "grad_norm": 2.3575692245336124,
+      "learning_rate": 1.9548458891838794e-05,
+      "loss": 0.8872,
+      "step": 833
+    },
+    {
+      "epoch": 0.42418462712187677,
+      "grad_norm": 2.3250114563690465,
+      "learning_rate": 1.954679154763889e-05,
+      "loss": 0.6911,
+      "step": 834
+    },
+    {
+      "epoch": 0.4246932417826944,
+      "grad_norm": 2.3763343676459905,
+      "learning_rate": 1.9545121202093927e-05,
+      "loss": 0.8004,
+      "step": 835
+    },
+    {
+      "epoch": 0.425201856443512,
+      "grad_norm": 2.263014242837935,
+      "learning_rate": 1.9543447855729026e-05,
+      "loss": 0.8008,
+      "step": 836
+    },
+    {
+      "epoch": 0.4257104711043296,
+      "grad_norm": 2.711358645044378,
+      "learning_rate": 1.954177150907026e-05,
+      "loss": 0.7448,
+      "step": 837
+    },
+    {
+      "epoch": 0.42621908576514717,
+      "grad_norm": 2.417697379763892,
+      "learning_rate": 1.9540092162644645e-05,
+      "loss": 0.7585,
+      "step": 838
+    },
+    {
+      "epoch": 0.42672770042596475,
+      "grad_norm": 2.4130337692914128,
+      "learning_rate": 1.9538409816980134e-05,
+      "loss": 0.6713,
+      "step": 839
+    },
+    {
+      "epoch": 0.4272363150867824,
+      "grad_norm": 2.420146300567365,
+      "learning_rate": 1.953672447260563e-05,
+      "loss": 0.8216,
+      "step": 840
+    },
+    {
+      "epoch": 0.4277449297476,
+      "grad_norm": 2.2804001570785353,
+      "learning_rate": 1.9535036130050976e-05,
+      "loss": 0.8878,
+      "step": 841
+    },
+    {
+      "epoch": 0.42825354440841756,
+      "grad_norm": 2.7094754280740236,
+      "learning_rate": 1.9533344789846956e-05,
+      "loss": 0.8595,
+      "step": 842
+    },
+    {
+      "epoch": 0.42876215906923515,
+      "grad_norm": 3.451418631971514,
+      "learning_rate": 1.95316504525253e-05,
+      "loss": 0.8507,
+      "step": 843
+    },
+    {
+      "epoch": 0.4292707737300528,
+      "grad_norm": 2.682314283158343,
+      "learning_rate": 1.9529953118618676e-05,
+      "loss": 0.788,
+      "step": 844
+    },
+    {
+      "epoch": 0.4297793883908704,
+      "grad_norm": 2.5122573773350463,
+      "learning_rate": 1.9528252788660694e-05,
+      "loss": 0.7582,
+      "step": 845
+    },
+    {
+      "epoch": 0.43028800305168796,
+      "grad_norm": 2.6751325948776823,
+      "learning_rate": 1.9526549463185914e-05,
+      "loss": 0.8144,
+      "step": 846
+    },
+    {
+      "epoch": 0.43079661771250555,
+      "grad_norm": 2.554432520842944,
+      "learning_rate": 1.9524843142729833e-05,
+      "loss": 0.7787,
+      "step": 847
+    },
+    {
+      "epoch": 0.4313052323733232,
+      "grad_norm": 2.3492986484713843,
+      "learning_rate": 1.952313382782888e-05,
+      "loss": 0.8027,
+      "step": 848
+    },
+    {
+      "epoch": 0.4318138470341408,
+      "grad_norm": 3.1361013448552066,
+      "learning_rate": 1.952142151902044e-05,
+      "loss": 0.8544,
+      "step": 849
+    },
+    {
+      "epoch": 0.43232246169495836,
+      "grad_norm": 2.849813348112397,
+      "learning_rate": 1.9519706216842832e-05,
+      "loss": 0.7823,
+      "step": 850
+    },
+    {
+      "epoch": 0.43283107635577595,
+      "grad_norm": 3.6308426961777354,
+      "learning_rate": 1.9517987921835314e-05,
+      "loss": 0.9469,
+      "step": 851
+    },
+    {
+      "epoch": 0.43333969101659353,
+      "grad_norm": 2.2482690510798973,
+      "learning_rate": 1.9516266634538095e-05,
+      "loss": 0.8032,
+      "step": 852
+    },
+    {
+      "epoch": 0.4338483056774112,
+      "grad_norm": 2.2681320525853623,
+      "learning_rate": 1.951454235549231e-05,
+      "loss": 0.7067,
+      "step": 853
+    },
+    {
+      "epoch": 0.43435692033822876,
+      "grad_norm": 2.3268839965613606,
+      "learning_rate": 1.9512815085240048e-05,
+      "loss": 0.8297,
+      "step": 854
+    },
+    {
+      "epoch": 0.43486553499904634,
+      "grad_norm": 2.179986748611352,
+      "learning_rate": 1.951108482432433e-05,
+      "loss": 0.81,
+      "step": 855
+    },
+    {
+      "epoch": 0.43537414965986393,
+      "grad_norm": 3.1293972358805013,
+      "learning_rate": 1.950935157328912e-05,
+      "loss": 0.8914,
+      "step": 856
+    },
+    {
+      "epoch": 0.43588276432068157,
+      "grad_norm": 2.378835291814264,
+      "learning_rate": 1.9507615332679322e-05,
+      "loss": 0.8072,
+      "step": 857
+    },
+    {
+      "epoch": 0.43639137898149916,
+      "grad_norm": 2.7926049055511415,
+      "learning_rate": 1.950587610304078e-05,
+      "loss": 0.8025,
+      "step": 858
+    },
+    {
+      "epoch": 0.43689999364231674,
+      "grad_norm": 2.669037678778122,
+      "learning_rate": 1.950413388492028e-05,
+      "loss": 0.8419,
+      "step": 859
+    },
+    {
+      "epoch": 0.4374086083031343,
+      "grad_norm": 2.3450145107706897,
+      "learning_rate": 1.950238867886554e-05,
+      "loss": 0.8582,
+      "step": 860
+    },
+    {
+      "epoch": 0.4379172229639519,
+      "grad_norm": 3.0290717724432192,
+      "learning_rate": 1.9500640485425224e-05,
+      "loss": 0.6746,
+      "step": 861
+    },
+    {
+      "epoch": 0.43842583762476955,
+      "grad_norm": 2.3648385961833704,
+      "learning_rate": 1.949888930514894e-05,
+      "loss": 0.7607,
+      "step": 862
+    },
+    {
+      "epoch": 0.43893445228558714,
+      "grad_norm": 2.5539544129869904,
+      "learning_rate": 1.9497135138587217e-05,
+      "loss": 0.7031,
+      "step": 863
+    },
+    {
+      "epoch": 0.4394430669464047,
+      "grad_norm": 2.3760209148013076,
+      "learning_rate": 1.9495377986291544e-05,
+      "loss": 0.8107,
+      "step": 864
+    },
+    {
+      "epoch": 0.4399516816072223,
+      "grad_norm": 2.9495893030869174,
+      "learning_rate": 1.9493617848814338e-05,
+      "loss": 0.8675,
+      "step": 865
+    },
+    {
+      "epoch": 0.44046029626803995,
+      "grad_norm": 2.3903382571918907,
+      "learning_rate": 1.949185472670895e-05,
+      "loss": 0.707,
+      "step": 866
+    },
+    {
+      "epoch": 0.44096891092885754,
+      "grad_norm": 2.242872499043705,
+      "learning_rate": 1.9490088620529678e-05,
+      "loss": 0.8225,
+      "step": 867
+    },
+    {
+      "epoch": 0.4414775255896751,
+      "grad_norm": 2.5650398878896294,
+      "learning_rate": 1.948831953083176e-05,
+      "loss": 0.7722,
+      "step": 868
+    },
+    {
+      "epoch": 0.4419861402504927,
+      "grad_norm": 2.632592438236845,
+      "learning_rate": 1.9486547458171355e-05,
+      "loss": 0.8316,
+      "step": 869
+    },
+    {
+      "epoch": 0.4424947549113103,
+      "grad_norm": 3.081841099122682,
+      "learning_rate": 1.948477240310559e-05,
+      "loss": 0.7881,
+      "step": 870
+    },
+    {
+      "epoch": 0.44300336957212794,
+      "grad_norm": 2.5567892685613662,
+      "learning_rate": 1.948299436619249e-05,
+      "loss": 0.6519,
+      "step": 871
+    },
+    {
+      "epoch": 0.4435119842329455,
+      "grad_norm": 2.4918974226264967,
+      "learning_rate": 1.9481213347991053e-05,
+      "loss": 0.7153,
+      "step": 872
+    },
+    {
+      "epoch": 0.4440205988937631,
+      "grad_norm": 2.3966862252312326,
+      "learning_rate": 1.9479429349061198e-05,
+      "loss": 0.7135,
+      "step": 873
+    },
+    {
+      "epoch": 0.4445292135545807,
+      "grad_norm": 3.3782383958838413,
+      "learning_rate": 1.947764236996378e-05,
+      "loss": 0.85,
+      "step": 874
+    },
+    {
+      "epoch": 0.44503782821539833,
+      "grad_norm": 3.223440363887785,
+      "learning_rate": 1.9475852411260598e-05,
+      "loss": 0.7444,
+      "step": 875
+    },
+    {
+      "epoch": 0.4455464428762159,
+      "grad_norm": 2.3129267943095306,
+      "learning_rate": 1.947405947351438e-05,
+      "loss": 0.8849,
+      "step": 876
+    },
+    {
+      "epoch": 0.4460550575370335,
+      "grad_norm": 2.4957493886273396,
+      "learning_rate": 1.9472263557288796e-05,
+      "loss": 0.7693,
+      "step": 877
+    },
+    {
+      "epoch": 0.4465636721978511,
+      "grad_norm": 2.4322640351407996,
+      "learning_rate": 1.9470464663148455e-05,
+      "loss": 0.7705,
+      "step": 878
+    },
+    {
+      "epoch": 0.4470722868586687,
+      "grad_norm": 2.5676379913124734,
+      "learning_rate": 1.9468662791658888e-05,
+      "loss": 0.9015,
+      "step": 879
+    },
+    {
+      "epoch": 0.4475809015194863,
+      "grad_norm": 2.494013485098864,
+      "learning_rate": 1.946685794338658e-05,
+      "loss": 0.7585,
+      "step": 880
+    },
+    {
+      "epoch": 0.4480895161803039,
+      "grad_norm": 2.557909319238901,
+      "learning_rate": 1.9465050118898944e-05,
+      "loss": 0.6651,
+      "step": 881
+    },
+    {
+      "epoch": 0.4485981308411215,
+      "grad_norm": 2.2460739002179957,
+      "learning_rate": 1.9463239318764323e-05,
+      "loss": 0.8178,
+      "step": 882
+    },
+    {
+      "epoch": 0.4491067455019391,
+      "grad_norm": 2.4467100592081943,
+      "learning_rate": 1.9461425543552008e-05,
+      "loss": 0.72,
+      "step": 883
+    },
+    {
+      "epoch": 0.4496153601627567,
+      "grad_norm": 2.4013633154007517,
+      "learning_rate": 1.9459608793832208e-05,
+      "loss": 0.8474,
+      "step": 884
+    },
+    {
+      "epoch": 0.4501239748235743,
+      "grad_norm": 2.3218316137026207,
+      "learning_rate": 1.9457789070176095e-05,
+      "loss": 0.7436,
+      "step": 885
+    },
+    {
+      "epoch": 0.4506325894843919,
+      "grad_norm": 2.4052645163887543,
+      "learning_rate": 1.945596637315574e-05,
+      "loss": 0.8113,
+      "step": 886
+    },
+    {
+      "epoch": 0.4511412041452095,
+      "grad_norm": 2.4905724150157242,
+      "learning_rate": 1.9454140703344175e-05,
+      "loss": 0.8644,
+      "step": 887
+    },
+    {
+      "epoch": 0.45164981880602706,
+      "grad_norm": 2.5624516180169596,
+      "learning_rate": 1.9452312061315363e-05,
+      "loss": 0.972,
+      "step": 888
+    },
+    {
+      "epoch": 0.4521584334668447,
+      "grad_norm": 2.2024065293602924,
+      "learning_rate": 1.945048044764419e-05,
+      "loss": 0.719,
+      "step": 889
+    },
+    {
+      "epoch": 0.4526670481276623,
+      "grad_norm": 2.430399603306108,
+      "learning_rate": 1.9448645862906485e-05,
+      "loss": 0.8254,
+      "step": 890
+    },
+    {
+      "epoch": 0.45317566278847987,
+      "grad_norm": 2.5379904963538613,
+      "learning_rate": 1.9446808307679013e-05,
+      "loss": 0.7624,
+      "step": 891
+    },
+    {
+      "epoch": 0.45368427744929746,
+      "grad_norm": 2.219860067854664,
+      "learning_rate": 1.9444967782539466e-05,
+      "loss": 0.6326,
+      "step": 892
+    },
+    {
+      "epoch": 0.4541928921101151,
+      "grad_norm": 2.3247504643896058,
+      "learning_rate": 1.944312428806648e-05,
+      "loss": 0.7666,
+      "step": 893
+    },
+    {
+      "epoch": 0.4547015067709327,
+      "grad_norm": 2.735654956426138,
+      "learning_rate": 1.9441277824839603e-05,
+      "loss": 0.7949,
+      "step": 894
+    },
+    {
+      "epoch": 0.45521012143175027,
+      "grad_norm": 2.6123383649990384,
+      "learning_rate": 1.9439428393439342e-05,
+      "loss": 0.7486,
+      "step": 895
+    },
+    {
+      "epoch": 0.45571873609256786,
+      "grad_norm": 2.659418886138602,
+      "learning_rate": 1.943757599444713e-05,
+      "loss": 0.7962,
+      "step": 896
+    },
+    {
+      "epoch": 0.45622735075338544,
+      "grad_norm": 2.9217248561857954,
+      "learning_rate": 1.9435720628445317e-05,
+      "loss": 0.8533,
+      "step": 897
+    },
+    {
+      "epoch": 0.4567359654142031,
+      "grad_norm": 2.489999414466881,
+      "learning_rate": 1.943386229601721e-05,
+      "loss": 0.7494,
+      "step": 898
+    },
+    {
+      "epoch": 0.45724458007502067,
+      "grad_norm": 2.69568788429886,
+      "learning_rate": 1.943200099774702e-05,
+      "loss": 0.7234,
+      "step": 899
+    },
+    {
+      "epoch": 0.45775319473583825,
+      "grad_norm": 3.098852554353867,
+      "learning_rate": 1.9430136734219923e-05,
+      "loss": 0.8527,
+      "step": 900
+    },
+    {
+      "epoch": 0.45775319473583825,
+      "eval_loss": 0.7568135261535645,
+      "eval_runtime": 170.1451,
+      "eval_samples_per_second": 19.466,
+      "eval_steps_per_second": 9.733,
+      "step": 900
+    },
+    {
+      "epoch": 0.45826180939665584,
+      "grad_norm": 3.0004478756051793,
+      "learning_rate": 1.9428269506022003e-05,
+      "loss": 0.9014,
+      "step": 901
+    },
+    {
+      "epoch": 0.4587704240574735,
+      "grad_norm": 2.5061794204477033,
+      "learning_rate": 1.9426399313740287e-05,
+      "loss": 0.8343,
+      "step": 902
+    },
+    {
+      "epoch": 0.45927903871829107,
+      "grad_norm": 2.243489913955165,
+      "learning_rate": 1.9424526157962732e-05,
+      "loss": 0.8112,
+      "step": 903
+    },
+    {
+      "epoch": 0.45978765337910865,
+      "grad_norm": 2.2771964442670445,
+      "learning_rate": 1.9422650039278218e-05,
+      "loss": 0.7859,
+      "step": 904
+    },
+    {
+      "epoch": 0.46029626803992624,
+      "grad_norm": 2.518727350251469,
+      "learning_rate": 1.9420770958276573e-05,
+      "loss": 0.831,
+      "step": 905
+    },
+    {
+      "epoch": 0.4608048827007438,
+      "grad_norm": 2.4868401347072595,
+      "learning_rate": 1.9418888915548543e-05,
+      "loss": 0.7735,
+      "step": 906
+    },
+    {
+      "epoch": 0.46131349736156146,
+      "grad_norm": 2.5786860000131497,
+      "learning_rate": 1.941700391168581e-05,
+      "loss": 0.773,
+      "step": 907
+    },
+    {
+      "epoch": 0.46182211202237905,
+      "grad_norm": 2.1083862247793586,
+      "learning_rate": 1.941511594728098e-05,
+      "loss": 0.7275,
+      "step": 908
+    },
+    {
+      "epoch": 0.46233072668319664,
+      "grad_norm": 2.6420753604095033,
+      "learning_rate": 1.9413225022927614e-05,
+      "loss": 0.8463,
+      "step": 909
+    },
+    {
+      "epoch": 0.4628393413440142,
+      "grad_norm": 2.9865706590423464,
+      "learning_rate": 1.9411331139220168e-05,
+      "loss": 0.8087,
+      "step": 910
+    },
+    {
+      "epoch": 0.46334795600483186,
+      "grad_norm": 3.557446291659509,
+      "learning_rate": 1.940943429675405e-05,
+      "loss": 0.8226,
+      "step": 911
+    },
+    {
+      "epoch": 0.46385657066564945,
+      "grad_norm": 2.3095258196361783,
+      "learning_rate": 1.94075344961256e-05,
+      "loss": 0.737,
+      "step": 912
+    },
+    {
+      "epoch": 0.46436518532646703,
+      "grad_norm": 2.442428301693037,
+      "learning_rate": 1.940563173793208e-05,
+      "loss": 0.8575,
+      "step": 913
+    },
+    {
+      "epoch": 0.4648737999872846,
+      "grad_norm": 2.3843700883847267,
+      "learning_rate": 1.9403726022771684e-05,
+      "loss": 0.7967,
+      "step": 914
+    },
+    {
+      "epoch": 0.4653824146481022,
+      "grad_norm": 2.4952344295779154,
+      "learning_rate": 1.940181735124353e-05,
+      "loss": 0.7692,
+      "step": 915
+    },
+    {
+      "epoch": 0.46589102930891985,
+      "grad_norm": 2.5679754978389533,
+      "learning_rate": 1.939990572394768e-05,
+      "loss": 0.8225,
+      "step": 916
+    },
+    {
+      "epoch": 0.46639964396973743,
+      "grad_norm": 2.427223943694825,
+      "learning_rate": 1.9397991141485112e-05,
+      "loss": 0.8209,
+      "step": 917
+    },
+    {
+      "epoch": 0.466908258630555,
+      "grad_norm": 2.209395447114377,
+      "learning_rate": 1.9396073604457736e-05,
+      "loss": 0.7584,
+      "step": 918
+    },
+    {
+      "epoch": 0.4674168732913726,
+      "grad_norm": 2.531898792148466,
+      "learning_rate": 1.939415311346839e-05,
+      "loss": 0.7701,
+      "step": 919
+    },
+    {
+      "epoch": 0.46792548795219024,
+      "grad_norm": 2.5188060101133734,
+      "learning_rate": 1.9392229669120853e-05,
+      "loss": 0.7801,
+      "step": 920
+    },
+    {
+      "epoch": 0.46843410261300783,
+      "grad_norm": 2.2099220990566444,
+      "learning_rate": 1.939030327201981e-05,
+      "loss": 0.8724,
+      "step": 921
+    },
+    {
+      "epoch": 0.4689427172738254,
+      "grad_norm": 2.3138344497148573,
+      "learning_rate": 1.9388373922770898e-05,
+      "loss": 0.7804,
+      "step": 922
+    },
+    {
+      "epoch": 0.469451331934643,
+      "grad_norm": 2.528179519876486,
+      "learning_rate": 1.9386441621980664e-05,
+      "loss": 0.7556,
+      "step": 923
+    },
+    {
+      "epoch": 0.4699599465954606,
+      "grad_norm": 2.3529673084042084,
+      "learning_rate": 1.938450637025659e-05,
+      "loss": 0.7245,
+      "step": 924
+    },
+    {
+      "epoch": 0.47046856125627823,
+      "grad_norm": 2.8924836652685206,
+      "learning_rate": 1.9382568168207088e-05,
+      "loss": 0.8763,
+      "step": 925
+    },
+    {
+      "epoch": 0.4709771759170958,
+      "grad_norm": 2.4644865088469885,
+      "learning_rate": 1.9380627016441495e-05,
+      "loss": 0.8111,
+      "step": 926
+    },
+    {
+      "epoch": 0.4714857905779134,
+      "grad_norm": 2.554075395240468,
+      "learning_rate": 1.937868291557007e-05,
+      "loss": 0.8182,
+      "step": 927
+    },
+    {
+      "epoch": 0.471994405238731,
+      "grad_norm": 2.3668278084103007,
+      "learning_rate": 1.9376735866204013e-05,
+      "loss": 0.8584,
+      "step": 928
+    },
+    {
+      "epoch": 0.4725030198995486,
+      "grad_norm": 2.7311193648812737,
+      "learning_rate": 1.9374785868955436e-05,
+      "loss": 0.8557,
+      "step": 929
+    },
+    {
+      "epoch": 0.4730116345603662,
+      "grad_norm": 2.5451049270434107,
+      "learning_rate": 1.9372832924437386e-05,
+      "loss": 0.8139,
+      "step": 930
+    },
+    {
+      "epoch": 0.4735202492211838,
+      "grad_norm": 2.6259462376805196,
+      "learning_rate": 1.9370877033263836e-05,
+      "loss": 0.8116,
+      "step": 931
+    },
+    {
+      "epoch": 0.4740288638820014,
+      "grad_norm": 2.5485665724278577,
+      "learning_rate": 1.936891819604968e-05,
+      "loss": 0.7895,
+      "step": 932
+    },
+    {
+      "epoch": 0.47453747854281897,
+      "grad_norm": 3.022096637223369,
+      "learning_rate": 1.936695641341075e-05,
+      "loss": 0.8186,
+      "step": 933
+    },
+    {
+      "epoch": 0.4750460932036366,
+      "grad_norm": 2.433115610633213,
+      "learning_rate": 1.9364991685963788e-05,
+      "loss": 0.8381,
+      "step": 934
+    },
+    {
+      "epoch": 0.4755547078644542,
+      "grad_norm": 2.3985748748702607,
+      "learning_rate": 1.9363024014326478e-05,
+      "loss": 0.7842,
+      "step": 935
+    },
+    {
+      "epoch": 0.4760633225252718,
+      "grad_norm": 2.4852548636640313,
+      "learning_rate": 1.9361053399117413e-05,
+      "loss": 0.7718,
+      "step": 936
+    },
+    {
+      "epoch": 0.47657193718608937,
+      "grad_norm": 2.8219545608744547,
+      "learning_rate": 1.9359079840956124e-05,
+      "loss": 0.719,
+      "step": 937
+    },
+    {
+      "epoch": 0.477080551846907,
+      "grad_norm": 2.881421422927217,
+      "learning_rate": 1.9357103340463067e-05,
+      "loss": 0.8462,
+      "step": 938
+    },
+    {
+      "epoch": 0.4775891665077246,
+      "grad_norm": 2.4970361306689597,
+      "learning_rate": 1.9355123898259616e-05,
+      "loss": 0.8513,
+      "step": 939
+    },
+    {
+      "epoch": 0.4780977811685422,
+      "grad_norm": 5.674876866286154,
+      "learning_rate": 1.9353141514968075e-05,
+      "loss": 0.7344,
+      "step": 940
+    },
+    {
+      "epoch": 0.47860639582935977,
+      "grad_norm": 2.23543123475726,
+      "learning_rate": 1.9351156191211665e-05,
+      "loss": 0.7614,
+      "step": 941
+    },
+    {
+      "epoch": 0.4791150104901774,
+      "grad_norm": 2.4418095370037447,
+      "learning_rate": 1.9349167927614547e-05,
+      "loss": 0.8382,
+      "step": 942
+    },
+    {
+      "epoch": 0.479623625150995,
+      "grad_norm": 2.417811374750328,
+      "learning_rate": 1.9347176724801788e-05,
+      "loss": 0.8912,
+      "step": 943
+    },
+    {
+      "epoch": 0.4801322398118126,
+      "grad_norm": 2.4293182190544598,
+      "learning_rate": 1.934518258339939e-05,
+      "loss": 0.7463,
+      "step": 944
+    },
+    {
+      "epoch": 0.48064085447263016,
+      "grad_norm": 2.3237355475261006,
+      "learning_rate": 1.9343185504034277e-05,
+      "loss": 0.8519,
+      "step": 945
+    },
+    {
+      "epoch": 0.48114946913344775,
+      "grad_norm": 2.639021527710273,
+      "learning_rate": 1.9341185487334297e-05,
+      "loss": 0.7721,
+      "step": 946
+    },
+    {
+      "epoch": 0.4816580837942654,
+      "grad_norm": 2.2965907744928873,
+      "learning_rate": 1.9339182533928215e-05,
+      "loss": 0.8174,
+      "step": 947
+    },
+    {
+      "epoch": 0.482166698455083,
+      "grad_norm": 2.411275100063204,
+      "learning_rate": 1.9337176644445736e-05,
+      "loss": 0.8842,
+      "step": 948
+    },
+    {
+      "epoch": 0.48267531311590056,
+      "grad_norm": 2.216703518821686,
+      "learning_rate": 1.9335167819517462e-05,
+      "loss": 0.8501,
+      "step": 949
+    },
+    {
+      "epoch": 0.48318392777671815,
+      "grad_norm": 2.3301321504633696,
+      "learning_rate": 1.9333156059774945e-05,
+      "loss": 0.7669,
+      "step": 950
+    },
+    {
+      "epoch": 0.4836925424375358,
+      "grad_norm": 2.4875510205734583,
+      "learning_rate": 1.933114136585064e-05,
+      "loss": 0.7837,
+      "step": 951
+    },
+    {
+      "epoch": 0.4842011570983534,
+      "grad_norm": 2.577215600737654,
+      "learning_rate": 1.9329123738377936e-05,
+      "loss": 0.7953,
+      "step": 952
+    },
+    {
+      "epoch": 0.48470977175917096,
+      "grad_norm": 2.306566823478009,
+      "learning_rate": 1.9327103177991138e-05,
+      "loss": 0.7618,
+      "step": 953
+    },
+    {
+      "epoch": 0.48521838641998855,
+      "grad_norm": 2.524270785192975,
+      "learning_rate": 1.9325079685325476e-05,
+      "loss": 0.7578,
+      "step": 954
+    },
+    {
+      "epoch": 0.48572700108080613,
+      "grad_norm": 2.4381823440532053,
+      "learning_rate": 1.93230532610171e-05,
+      "loss": 0.8342,
+      "step": 955
+    },
+    {
+      "epoch": 0.4862356157416238,
+      "grad_norm": 2.445361444635469,
+      "learning_rate": 1.9321023905703077e-05,
+      "loss": 0.8193,
+      "step": 956
+    },
+    {
+      "epoch": 0.48674423040244136,
+      "grad_norm": 2.2650647753280446,
+      "learning_rate": 1.9318991620021414e-05,
+      "loss": 0.6999,
+      "step": 957
+    },
+    {
+      "epoch": 0.48725284506325894,
+      "grad_norm": 2.601249423964151,
+      "learning_rate": 1.931695640461101e-05,
+      "loss": 0.7755,
+      "step": 958
+    },
+    {
+      "epoch": 0.48776145972407653,
+      "grad_norm": 2.507032231431426,
+      "learning_rate": 1.931491826011172e-05,
+      "loss": 0.8113,
+      "step": 959
+    },
+    {
+      "epoch": 0.48827007438489417,
+      "grad_norm": 2.7300169302167374,
+      "learning_rate": 1.9312877187164286e-05,
+      "loss": 0.8777,
+      "step": 960
+    },
+    {
+      "epoch": 0.48877868904571176,
+      "grad_norm": 2.602530634367216,
+      "learning_rate": 1.931083318641039e-05,
+      "loss": 0.9668,
+      "step": 961
+    },
+    {
+      "epoch": 0.48928730370652934,
+      "grad_norm": 2.180826197335455,
+      "learning_rate": 1.9308786258492635e-05,
+      "loss": 0.7425,
+      "step": 962
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 2.6183837889486803,
+      "learning_rate": 1.930673640405453e-05,
+      "loss": 0.8532,
+      "step": 963
+    },
+    {
+      "epoch": 0.4903045330281645,
+      "grad_norm": 2.3423412921437894,
+      "learning_rate": 1.9304683623740526e-05,
+      "loss": 0.7269,
+      "step": 964
+    },
+    {
+      "epoch": 0.49081314768898215,
+      "grad_norm": 2.9730889986627926,
+      "learning_rate": 1.930262791819597e-05,
+      "loss": 0.7709,
+      "step": 965
+    },
+    {
+      "epoch": 0.49132176234979974,
+      "grad_norm": 2.4548583497555416,
+      "learning_rate": 1.9300569288067145e-05,
+      "loss": 0.8346,
+      "step": 966
+    },
+    {
+      "epoch": 0.4918303770106173,
+      "grad_norm": 3.088836775903484,
+      "learning_rate": 1.929850773400125e-05,
+      "loss": 0.8243,
+      "step": 967
+    },
+    {
+      "epoch": 0.4923389916714349,
+      "grad_norm": 2.4253895515516932,
+      "learning_rate": 1.92964432566464e-05,
+      "loss": 0.7914,
+      "step": 968
+    },
+    {
+      "epoch": 0.49284760633225255,
+      "grad_norm": 2.4832674832755437,
+      "learning_rate": 1.9294375856651627e-05,
+      "loss": 0.8906,
+      "step": 969
+    },
+    {
+      "epoch": 0.49335622099307014,
+      "grad_norm": 2.477800414871032,
+      "learning_rate": 1.9292305534666897e-05,
+      "loss": 0.7838,
+      "step": 970
+    },
+    {
+      "epoch": 0.4938648356538877,
+      "grad_norm": 2.6896024065550432,
+      "learning_rate": 1.929023229134307e-05,
+      "loss": 0.8424,
+      "step": 971
+    },
+    {
+      "epoch": 0.4943734503147053,
+      "grad_norm": 2.5020017239498356,
+      "learning_rate": 1.928815612733194e-05,
+      "loss": 0.7979,
+      "step": 972
+    },
+    {
+      "epoch": 0.4948820649755229,
+      "grad_norm": 2.255508038793372,
+      "learning_rate": 1.9286077043286225e-05,
+      "loss": 0.8569,
+      "step": 973
+    },
+    {
+      "epoch": 0.49539067963634054,
+      "grad_norm": 3.0228087882929437,
+      "learning_rate": 1.9283995039859544e-05,
+      "loss": 0.7153,
+      "step": 974
+    },
+    {
+      "epoch": 0.4958992942971581,
+      "grad_norm": 2.8484823234910923,
+      "learning_rate": 1.928191011770645e-05,
+      "loss": 0.7494,
+      "step": 975
+    },
+    {
+      "epoch": 0.4964079089579757,
+      "grad_norm": 2.5176192251421137,
+      "learning_rate": 1.92798222774824e-05,
+      "loss": 0.84,
+      "step": 976
+    },
+    {
+      "epoch": 0.4969165236187933,
+      "grad_norm": 2.3549761660396364,
+      "learning_rate": 1.927773151984378e-05,
+      "loss": 0.9197,
+      "step": 977
+    },
+    {
+      "epoch": 0.49742513827961093,
+      "grad_norm": 2.3930555964896896,
+      "learning_rate": 1.927563784544788e-05,
+      "loss": 0.8771,
+      "step": 978
+    },
+    {
+      "epoch": 0.4979337529404285,
+      "grad_norm": 2.4906826917219917,
+      "learning_rate": 1.9273541254952922e-05,
+      "loss": 0.8743,
+      "step": 979
+    },
+    {
+      "epoch": 0.4984423676012461,
+      "grad_norm": 2.3781083496152493,
+      "learning_rate": 1.9271441749018037e-05,
+      "loss": 0.7596,
+      "step": 980
+    },
+    {
+      "epoch": 0.4989509822620637,
+      "grad_norm": 2.2747297776259625,
+      "learning_rate": 1.926933932830327e-05,
+      "loss": 0.8164,
+      "step": 981
+    },
+    {
+      "epoch": 0.4994595969228813,
+      "grad_norm": 2.540272022609496,
+      "learning_rate": 1.9267233993469587e-05,
+      "loss": 0.8239,
+      "step": 982
+    },
+    {
+      "epoch": 0.4999682115836989,
+      "grad_norm": 2.006462741873765,
+      "learning_rate": 1.9265125745178866e-05,
+      "loss": 0.8182,
+      "step": 983
+    },
+    {
+      "epoch": 0.5004768262445165,
+      "grad_norm": 2.550317882844872,
+      "learning_rate": 1.926301458409391e-05,
+      "loss": 0.828,
+      "step": 984
+    },
+    {
+      "epoch": 0.5009854409053341,
+      "grad_norm": 2.121929418571094,
+      "learning_rate": 1.9260900510878426e-05,
+      "loss": 0.8394,
+      "step": 985
+    },
+    {
+      "epoch": 0.5014940555661517,
+      "grad_norm": 2.4581616444090786,
+      "learning_rate": 1.9258783526197043e-05,
+      "loss": 0.8026,
+      "step": 986
+    },
+    {
+      "epoch": 0.5020026702269693,
+      "grad_norm": 2.374224787230879,
+      "learning_rate": 1.9256663630715305e-05,
+      "loss": 0.8199,
+      "step": 987
+    },
+    {
+      "epoch": 0.5025112848877868,
+      "grad_norm": 2.689638218712434,
+      "learning_rate": 1.925454082509967e-05,
+      "loss": 0.7241,
+      "step": 988
+    },
+    {
+      "epoch": 0.5030198995486045,
+      "grad_norm": 2.4221043631940415,
+      "learning_rate": 1.9252415110017504e-05,
+      "loss": 0.7302,
+      "step": 989
+    },
+    {
+      "epoch": 0.5035285142094221,
+      "grad_norm": 2.2192044934343476,
+      "learning_rate": 1.9250286486137107e-05,
+      "loss": 0.8001,
+      "step": 990
+    },
+    {
+      "epoch": 0.5040371288702397,
+      "grad_norm": 2.545560439031253,
+      "learning_rate": 1.9248154954127677e-05,
+      "loss": 0.6876,
+      "step": 991
+    },
+    {
+      "epoch": 0.5045457435310573,
+      "grad_norm": 2.838861644570182,
+      "learning_rate": 1.9246020514659326e-05,
+      "loss": 1.0042,
+      "step": 992
+    },
+    {
+      "epoch": 0.5050543581918748,
+      "grad_norm": 2.654820225694315,
+      "learning_rate": 1.9243883168403093e-05,
+      "loss": 0.8595,
+      "step": 993
+    },
+    {
+      "epoch": 0.5055629728526925,
+      "grad_norm": 2.332025626988876,
+      "learning_rate": 1.9241742916030913e-05,
+      "loss": 0.7745,
+      "step": 994
+    },
+    {
+      "epoch": 0.5060715875135101,
+      "grad_norm": 2.4288574959864,
+      "learning_rate": 1.923959975821565e-05,
+      "loss": 0.7711,
+      "step": 995
+    },
+    {
+      "epoch": 0.5065802021743276,
+      "grad_norm": 2.245579403574402,
+      "learning_rate": 1.923745369563107e-05,
+      "loss": 0.8926,
+      "step": 996
+    },
+    {
+      "epoch": 0.5070888168351453,
+      "grad_norm": 3.2654518191034057,
+      "learning_rate": 1.9235304728951868e-05,
+      "loss": 0.8833,
+      "step": 997
+    },
+    {
+      "epoch": 0.5075974314959629,
+      "grad_norm": 2.173510477690609,
+      "learning_rate": 1.9233152858853634e-05,
+      "loss": 0.8877,
+      "step": 998
+    },
+    {
+      "epoch": 0.5081060461567805,
+      "grad_norm": 2.209882720493597,
+      "learning_rate": 1.9230998086012876e-05,
+      "loss": 0.7036,
+      "step": 999
+    },
+    {
+      "epoch": 0.5086146608175981,
+      "grad_norm": 2.216188268258961,
+      "learning_rate": 1.9228840411107025e-05,
+      "loss": 0.7452,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5086146608175981,
+      "eval_loss": 0.7460782527923584,
+      "eval_runtime": 170.0897,
+      "eval_samples_per_second": 19.472,
+      "eval_steps_per_second": 9.736,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5091232754784156,
+      "grad_norm": 2.310259893915683,
+      "learning_rate": 1.922667983481441e-05,
+      "loss": 0.778,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5096318901392333,
+      "grad_norm": 2.3972384658426202,
+      "learning_rate": 1.9224516357814282e-05,
+      "loss": 0.7815,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5101405048000509,
+      "grad_norm": 2.3042311394370607,
+      "learning_rate": 1.92223499807868e-05,
+      "loss": 0.8147,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5106491194608684,
+      "grad_norm": 2.480136158189542,
+      "learning_rate": 1.922018070441303e-05,
+      "loss": 0.9088,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5111577341216861,
+      "grad_norm": 2.4472425357737118,
+      "learning_rate": 1.9218008529374965e-05,
+      "loss": 0.7806,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5116663487825036,
+      "grad_norm": 2.3593405411584936,
+      "learning_rate": 1.9215833456355494e-05,
+      "loss": 0.8068,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5121749634433213,
+      "grad_norm": 2.2643962816780316,
+      "learning_rate": 1.9213655486038417e-05,
+      "loss": 0.8607,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5126835781041389,
+      "grad_norm": 2.2315194616372978,
+      "learning_rate": 1.9211474619108456e-05,
+      "loss": 0.7442,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5131921927649564,
+      "grad_norm": 2.2370435877898927,
+      "learning_rate": 1.9209290856251236e-05,
+      "loss": 0.7293,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5137008074257741,
+      "grad_norm": 2.239673973437585,
+      "learning_rate": 1.9207104198153297e-05,
+      "loss": 0.8008,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5142094220865916,
+      "grad_norm": 2.0387512203285407,
+      "learning_rate": 1.920491464550208e-05,
+      "loss": 0.7304,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5147180367474092,
+      "grad_norm": 2.7754143345442235,
+      "learning_rate": 1.9202722198985947e-05,
+      "loss": 0.7937,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5152266514082269,
+      "grad_norm": 2.3048561163821315,
+      "learning_rate": 1.9200526859294164e-05,
+      "loss": 0.8031,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5157352660690444,
+      "grad_norm": 2.3224436416604313,
+      "learning_rate": 1.9198328627116908e-05,
+      "loss": 0.76,
+      "step": 1014
+    },
+    {
+      "epoch": 0.516243880729862,
+      "grad_norm": 2.5743658229278115,
+      "learning_rate": 1.919612750314526e-05,
+      "loss": 0.801,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5167524953906797,
+      "grad_norm": 2.505836730521711,
+      "learning_rate": 1.9193923488071222e-05,
+      "loss": 0.7773,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5172611100514972,
+      "grad_norm": 2.4786497159477126,
+      "learning_rate": 1.9191716582587694e-05,
+      "loss": 0.7817,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5177697247123149,
+      "grad_norm": 2.228266763123114,
+      "learning_rate": 1.9189506787388488e-05,
+      "loss": 0.6696,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5182783393731324,
+      "grad_norm": 3.162242526101418,
+      "learning_rate": 1.9187294103168334e-05,
+      "loss": 0.7035,
+      "step": 1019
+    },
+    {
+      "epoch": 0.51878695403395,
+      "grad_norm": 2.740876060826273,
+      "learning_rate": 1.9185078530622848e-05,
+      "loss": 0.7701,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5192955686947677,
+      "grad_norm": 2.4008647552873845,
+      "learning_rate": 1.9182860070448573e-05,
+      "loss": 0.7979,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5198041833555852,
+      "grad_norm": 2.8283632294021364,
+      "learning_rate": 1.918063872334296e-05,
+      "loss": 0.8589,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5203127980164028,
+      "grad_norm": 2.6592048814737934,
+      "learning_rate": 1.9178414490004356e-05,
+      "loss": 0.8548,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5208214126772204,
+      "grad_norm": 3.826071301200185,
+      "learning_rate": 1.917618737113202e-05,
+      "loss": 0.8452,
+      "step": 1024
+    },
+    {
+      "epoch": 0.521330027338038,
+      "grad_norm": 2.3507916273735514,
+      "learning_rate": 1.917395736742612e-05,
+      "loss": 0.8014,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5218386419988557,
+      "grad_norm": 2.5248446010699355,
+      "learning_rate": 1.9171724479587738e-05,
+      "loss": 0.7582,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5223472566596732,
+      "grad_norm": 2.599517109850739,
+      "learning_rate": 1.916948870831884e-05,
+      "loss": 0.7667,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5228558713204908,
+      "grad_norm": 2.353408139888678,
+      "learning_rate": 1.9167250054322332e-05,
+      "loss": 0.8707,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5233644859813084,
+      "grad_norm": 3.2510085375042146,
+      "learning_rate": 1.916500851830199e-05,
+      "loss": 0.7348,
+      "step": 1029
+    },
+    {
+      "epoch": 0.523873100642126,
+      "grad_norm": 2.2356524250416214,
+      "learning_rate": 1.9162764100962526e-05,
+      "loss": 0.8221,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5243817153029436,
+      "grad_norm": 2.272976438562175,
+      "learning_rate": 1.9160516803009543e-05,
+      "loss": 0.7826,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5248903299637612,
+      "grad_norm": 2.5629458155923057,
+      "learning_rate": 1.915826662514955e-05,
+      "loss": 0.8825,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5253989446245788,
+      "grad_norm": 2.544871004054353,
+      "learning_rate": 1.9156013568089968e-05,
+      "loss": 0.7956,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5259075592853965,
+      "grad_norm": 2.426942998624607,
+      "learning_rate": 1.915375763253911e-05,
+      "loss": 0.7977,
+      "step": 1034
+    },
+    {
+      "epoch": 0.526416173946214,
+      "grad_norm": 2.4269357289848634,
+      "learning_rate": 1.9151498819206213e-05,
+      "loss": 0.8532,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5269247886070316,
+      "grad_norm": 2.3028525078702176,
+      "learning_rate": 1.9149237128801403e-05,
+      "loss": 0.78,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5274334032678492,
+      "grad_norm": 3.0157701526554477,
+      "learning_rate": 1.914697256203572e-05,
+      "loss": 0.8688,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5279420179286668,
+      "grad_norm": 2.3697812817244697,
+      "learning_rate": 1.91447051196211e-05,
+      "loss": 0.8607,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5284506325894844,
+      "grad_norm": 2.6907316782645765,
+      "learning_rate": 1.9142434802270388e-05,
+      "loss": 0.7806,
+      "step": 1039
+    },
+    {
+      "epoch": 0.528959247250302,
+      "grad_norm": 2.315721999387769,
+      "learning_rate": 1.9140161610697335e-05,
+      "loss": 0.7148,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5294678619111196,
+      "grad_norm": 2.2865090563686725,
+      "learning_rate": 1.9137885545616592e-05,
+      "loss": 0.8098,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5299764765719371,
+      "grad_norm": 2.2595439462560263,
+      "learning_rate": 1.9135606607743714e-05,
+      "loss": 0.776,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5304850912327548,
+      "grad_norm": 3.3419319986105944,
+      "learning_rate": 1.9133324797795157e-05,
+      "loss": 0.8224,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5309937058935724,
+      "grad_norm": 2.3410911228348756,
+      "learning_rate": 1.9131040116488284e-05,
+      "loss": 0.6388,
+      "step": 1044
+    },
+    {
+      "epoch": 0.53150232055439,
+      "grad_norm": 2.300077113641786,
+      "learning_rate": 1.912875256454136e-05,
+      "loss": 0.7045,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5320109352152076,
+      "grad_norm": 2.4322411956016086,
+      "learning_rate": 1.9126462142673555e-05,
+      "loss": 0.8289,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5325195498760251,
+      "grad_norm": 2.9282025859247356,
+      "learning_rate": 1.912416885160493e-05,
+      "loss": 0.7772,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5330281645368428,
+      "grad_norm": 2.3927113523174985,
+      "learning_rate": 1.912187269205646e-05,
+      "loss": 0.759,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5335367791976604,
+      "grad_norm": 2.5483252018195235,
+      "learning_rate": 1.911957366475002e-05,
+      "loss": 0.8202,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5340453938584779,
+      "grad_norm": 2.459011910960797,
+      "learning_rate": 1.9117271770408373e-05,
+      "loss": 0.8835,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5345540085192956,
+      "grad_norm": 2.6545837561693295,
+      "learning_rate": 1.911496700975521e-05,
+      "loss": 0.8626,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5350626231801132,
+      "grad_norm": 3.1699432353524006,
+      "learning_rate": 1.9112659383515097e-05,
+      "loss": 0.8145,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5355712378409307,
+      "grad_norm": 2.3602608539795185,
+      "learning_rate": 1.9110348892413517e-05,
+      "loss": 0.8455,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5360798525017484,
+      "grad_norm": 2.872760729929623,
+      "learning_rate": 1.9108035537176843e-05,
+      "loss": 0.8253,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5365884671625659,
+      "grad_norm": 3.2427264765203185,
+      "learning_rate": 1.9105719318532362e-05,
+      "loss": 0.8647,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5370970818233836,
+      "grad_norm": 2.2771818911579023,
+      "learning_rate": 1.9103400237208244e-05,
+      "loss": 0.704,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5376056964842012,
+      "grad_norm": 2.5229410444909433,
+      "learning_rate": 1.9101078293933574e-05,
+      "loss": 0.7911,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5381143111450187,
+      "grad_norm": 2.239877394658696,
+      "learning_rate": 1.9098753489438323e-05,
+      "loss": 0.7106,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5386229258058364,
+      "grad_norm": 2.6223683562570757,
+      "learning_rate": 1.9096425824453377e-05,
+      "loss": 0.7532,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5391315404666539,
+      "grad_norm": 2.778149709173282,
+      "learning_rate": 1.909409529971051e-05,
+      "loss": 0.7228,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5396401551274715,
+      "grad_norm": 2.3397740527234356,
+      "learning_rate": 1.9091761915942397e-05,
+      "loss": 0.7596,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5401487697882892,
+      "grad_norm": 2.2732048833286793,
+      "learning_rate": 1.9089425673882617e-05,
+      "loss": 0.8392,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5406573844491067,
+      "grad_norm": 2.3624240802494856,
+      "learning_rate": 1.908708657426564e-05,
+      "loss": 0.7947,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5411659991099244,
+      "grad_norm": 2.2175483135822067,
+      "learning_rate": 1.908474461782684e-05,
+      "loss": 0.7078,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5416746137707419,
+      "grad_norm": 2.6447701508251718,
+      "learning_rate": 1.9082399805302486e-05,
+      "loss": 0.8451,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5421832284315595,
+      "grad_norm": 2.2832991920016337,
+      "learning_rate": 1.9080052137429753e-05,
+      "loss": 0.8181,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5426918430923772,
+      "grad_norm": 2.2462082808775246,
+      "learning_rate": 1.9077701614946697e-05,
+      "loss": 0.7196,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5432004577531947,
+      "grad_norm": 2.2956467023571934,
+      "learning_rate": 1.9075348238592286e-05,
+      "loss": 0.8577,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5437090724140123,
+      "grad_norm": 2.3234629195464,
+      "learning_rate": 1.907299200910638e-05,
+      "loss": 0.6514,
+      "step": 1069
+    },
+    {
+      "epoch": 0.54421768707483,
+      "grad_norm": 2.2703196326588104,
+      "learning_rate": 1.907063292722974e-05,
+      "loss": 0.728,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5447263017356475,
+      "grad_norm": 2.805582142904302,
+      "learning_rate": 1.9068270993704013e-05,
+      "loss": 0.8748,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5452349163964652,
+      "grad_norm": 3.4071725418474825,
+      "learning_rate": 1.9065906209271758e-05,
+      "loss": 0.8936,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5457435310572827,
+      "grad_norm": 2.272009437078031,
+      "learning_rate": 1.9063538574676413e-05,
+      "loss": 0.9586,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5462521457181003,
+      "grad_norm": 2.376640656620067,
+      "learning_rate": 1.906116809066233e-05,
+      "loss": 0.7316,
+      "step": 1074
+    },
+    {
+      "epoch": 0.546760760378918,
+      "grad_norm": 2.102242512216342,
+      "learning_rate": 1.9058794757974743e-05,
+      "loss": 0.6872,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5472693750397355,
+      "grad_norm": 2.1682406968020866,
+      "learning_rate": 1.9056418577359787e-05,
+      "loss": 0.8111,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5477779897005531,
+      "grad_norm": 2.3700765473441985,
+      "learning_rate": 1.905403954956449e-05,
+      "loss": 0.7785,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5482866043613707,
+      "grad_norm": 2.492420245456644,
+      "learning_rate": 1.9051657675336782e-05,
+      "loss": 0.8109,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5487952190221883,
+      "grad_norm": 2.7293375752833353,
+      "learning_rate": 1.904927295542548e-05,
+      "loss": 0.7566,
+      "step": 1079
+    },
+    {
+      "epoch": 0.549303833683006,
+      "grad_norm": 2.849487887767586,
+      "learning_rate": 1.90468853905803e-05,
+      "loss": 0.9098,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5498124483438235,
+      "grad_norm": 2.3876831304009274,
+      "learning_rate": 1.9044494981551843e-05,
+      "loss": 0.7533,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5503210630046411,
+      "grad_norm": 2.3441660193930485,
+      "learning_rate": 1.9042101729091624e-05,
+      "loss": 0.8104,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5508296776654588,
+      "grad_norm": 2.2642740365372345,
+      "learning_rate": 1.903970563395203e-05,
+      "loss": 0.8642,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5513382923262763,
+      "grad_norm": 2.156324191821154,
+      "learning_rate": 1.903730669688636e-05,
+      "loss": 0.852,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5518469069870939,
+      "grad_norm": 3.193058926041889,
+      "learning_rate": 1.9034904918648794e-05,
+      "loss": 0.8086,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5523555216479115,
+      "grad_norm": 2.6671563136060277,
+      "learning_rate": 1.9032500299994405e-05,
+      "loss": 0.8727,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5528641363087291,
+      "grad_norm": 2.4355864227192145,
+      "learning_rate": 1.9030092841679168e-05,
+      "loss": 0.7401,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5533727509695467,
+      "grad_norm": 2.553672752313699,
+      "learning_rate": 1.9027682544459946e-05,
+      "loss": 0.7398,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5538813656303643,
+      "grad_norm": 2.696240782175945,
+      "learning_rate": 1.9025269409094495e-05,
+      "loss": 0.8157,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5543899802911819,
+      "grad_norm": 3.057504433155355,
+      "learning_rate": 1.9022853436341462e-05,
+      "loss": 0.7784,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5548985949519994,
+      "grad_norm": 2.3631578649673575,
+      "learning_rate": 1.9020434626960386e-05,
+      "loss": 0.9105,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5554072096128171,
+      "grad_norm": 2.7761461382609967,
+      "learning_rate": 1.9018012981711697e-05,
+      "loss": 0.8468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5559158242736347,
+      "grad_norm": 2.1031217215117115,
+      "learning_rate": 1.901558850135672e-05,
+      "loss": 0.797,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5564244389344523,
+      "grad_norm": 2.644062671253346,
+      "learning_rate": 1.901316118665767e-05,
+      "loss": 1.0108,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5569330535952699,
+      "grad_norm": 2.1809688616475187,
+      "learning_rate": 1.901073103837765e-05,
+      "loss": 0.7699,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5574416682560874,
+      "grad_norm": 4.512410957113063,
+      "learning_rate": 1.9008298057280657e-05,
+      "loss": 0.8535,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5579502829169051,
+      "grad_norm": 2.3487555394455013,
+      "learning_rate": 1.9005862244131582e-05,
+      "loss": 0.7239,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5584588975777227,
+      "grad_norm": 2.115129998362546,
+      "learning_rate": 1.90034235996962e-05,
+      "loss": 0.8206,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5589675122385402,
+      "grad_norm": 2.182841108130577,
+      "learning_rate": 1.9000982124741177e-05,
+      "loss": 0.7469,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5594761268993579,
+      "grad_norm": 2.6101210207304346,
+      "learning_rate": 1.899853782003407e-05,
+      "loss": 0.7916,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5594761268993579,
+      "eval_loss": 0.7452298402786255,
+      "eval_runtime": 169.5688,
+      "eval_samples_per_second": 19.532,
+      "eval_steps_per_second": 9.766,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5599847415601755,
+      "grad_norm": 2.2835853850570045,
+      "learning_rate": 1.8996090686343328e-05,
+      "loss": 0.7983,
+      "step": 1101
+    },
+    {
+      "epoch": 0.560493356220993,
+      "grad_norm": 2.2495324390766203,
+      "learning_rate": 1.8993640724438284e-05,
+      "loss": 0.7895,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5610019708818107,
+      "grad_norm": 2.39233875556421,
+      "learning_rate": 1.8991187935089168e-05,
+      "loss": 0.8279,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5615105855426282,
+      "grad_norm": 2.5850275288339004,
+      "learning_rate": 1.898873231906709e-05,
+      "loss": 0.818,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5620192002034459,
+      "grad_norm": 2.3007087113193667,
+      "learning_rate": 1.8986273877144055e-05,
+      "loss": 0.7038,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5625278148642635,
+      "grad_norm": 2.2472534370323936,
+      "learning_rate": 1.8983812610092952e-05,
+      "loss": 0.6553,
+      "step": 1106
+    },
+    {
+      "epoch": 0.563036429525081,
+      "grad_norm": 2.0930402321336485,
+      "learning_rate": 1.8981348518687563e-05,
+      "loss": 0.8078,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5635450441858987,
+      "grad_norm": 2.185713119933942,
+      "learning_rate": 1.8978881603702554e-05,
+      "loss": 0.8256,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5640536588467162,
+      "grad_norm": 2.485861087873891,
+      "learning_rate": 1.8976411865913485e-05,
+      "loss": 0.6715,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5645622735075339,
+      "grad_norm": 2.467896036137731,
+      "learning_rate": 1.897393930609679e-05,
+      "loss": 0.7236,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5650708881683515,
+      "grad_norm": 2.432016905813686,
+      "learning_rate": 1.8971463925029806e-05,
+      "loss": 0.9963,
+      "step": 1111
+    },
+    {
+      "epoch": 0.565579502829169,
+      "grad_norm": 2.1300356047972198,
+      "learning_rate": 1.8968985723490745e-05,
+      "loss": 0.7371,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5660881174899867,
+      "grad_norm": 2.580309687813068,
+      "learning_rate": 1.8966504702258716e-05,
+      "loss": 0.9575,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5665967321508042,
+      "grad_norm": 2.417483673987347,
+      "learning_rate": 1.8964020862113703e-05,
+      "loss": 0.824,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5671053468116218,
+      "grad_norm": 2.176146663204061,
+      "learning_rate": 1.8961534203836582e-05,
+      "loss": 0.779,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5676139614724395,
+      "grad_norm": 3.8024246411000857,
+      "learning_rate": 1.8959044728209122e-05,
+      "loss": 0.7203,
+      "step": 1116
+    },
+    {
+      "epoch": 0.568122576133257,
+      "grad_norm": 2.4221580093905586,
+      "learning_rate": 1.8956552436013963e-05,
+      "loss": 0.9167,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5686311907940746,
+      "grad_norm": 1.9245453557740007,
+      "learning_rate": 1.895405732803464e-05,
+      "loss": 0.7126,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5691398054548923,
+      "grad_norm": 2.165616025259294,
+      "learning_rate": 1.8951559405055578e-05,
+      "loss": 0.7098,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5696484201157098,
+      "grad_norm": 2.505542714306202,
+      "learning_rate": 1.8949058667862073e-05,
+      "loss": 0.7725,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5701570347765275,
+      "grad_norm": 2.251275336902976,
+      "learning_rate": 1.8946555117240317e-05,
+      "loss": 0.6768,
+      "step": 1121
+    },
+    {
+      "epoch": 0.570665649437345,
+      "grad_norm": 2.5672267026223983,
+      "learning_rate": 1.8944048753977377e-05,
+      "loss": 0.7974,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5711742640981626,
+      "grad_norm": 2.2516985945434493,
+      "learning_rate": 1.894153957886122e-05,
+      "loss": 0.7283,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5716828787589803,
+      "grad_norm": 2.2838700807568646,
+      "learning_rate": 1.893902759268068e-05,
+      "loss": 0.8208,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5721914934197978,
+      "grad_norm": 2.5276288162141514,
+      "learning_rate": 1.8936512796225482e-05,
+      "loss": 0.7775,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5727001080806154,
+      "grad_norm": 2.550360792478971,
+      "learning_rate": 1.8933995190286234e-05,
+      "loss": 0.8383,
+      "step": 1126
+    },
+    {
+      "epoch": 0.573208722741433,
+      "grad_norm": 2.1563859979615407,
+      "learning_rate": 1.893147477565443e-05,
+      "loss": 0.7311,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5737173374022506,
+      "grad_norm": 2.4610650892334407,
+      "learning_rate": 1.8928951553122442e-05,
+      "loss": 0.884,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5742259520630683,
+      "grad_norm": 2.2171395058037735,
+      "learning_rate": 1.8926425523483533e-05,
+      "loss": 0.8017,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5747345667238858,
+      "grad_norm": 2.1096242121664317,
+      "learning_rate": 1.8923896687531838e-05,
+      "loss": 0.6714,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5752431813847034,
+      "grad_norm": 2.4598603328351114,
+      "learning_rate": 1.8921365046062376e-05,
+      "loss": 0.7001,
+      "step": 1131
+    },
+    {
+      "epoch": 0.575751796045521,
+      "grad_norm": 2.4535890371159184,
+      "learning_rate": 1.8918830599871055e-05,
+      "loss": 0.7407,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5762604107063386,
+      "grad_norm": 2.733686960031594,
+      "learning_rate": 1.891629334975466e-05,
+      "loss": 0.8461,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5767690253671562,
+      "grad_norm": 2.2353030324911676,
+      "learning_rate": 1.8913753296510863e-05,
+      "loss": 0.9561,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5772776400279738,
+      "grad_norm": 2.5904904762009062,
+      "learning_rate": 1.89112104409382e-05,
+      "loss": 0.7597,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5777862546887914,
+      "grad_norm": 2.307696121163763,
+      "learning_rate": 1.8908664783836116e-05,
+      "loss": 0.7909,
+      "step": 1136
+    },
+    {
+      "epoch": 0.578294869349609,
+      "grad_norm": 2.7578524405637332,
+      "learning_rate": 1.8906116326004908e-05,
+      "loss": 0.8004,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5788034840104266,
+      "grad_norm": 2.4851735109355175,
+      "learning_rate": 1.8903565068245774e-05,
+      "loss": 0.8043,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5793120986712442,
+      "grad_norm": 3.299032797824697,
+      "learning_rate": 1.8901011011360783e-05,
+      "loss": 0.773,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5798207133320618,
+      "grad_norm": 2.2572705103421593,
+      "learning_rate": 1.8898454156152888e-05,
+      "loss": 0.8522,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5803293279928794,
+      "grad_norm": 2.347592573524177,
+      "learning_rate": 1.8895894503425918e-05,
+      "loss": 0.8487,
+      "step": 1141
+    },
+    {
+      "epoch": 0.580837942653697,
+      "grad_norm": 2.227895451474372,
+      "learning_rate": 1.8893332053984586e-05,
+      "loss": 0.8391,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5813465573145146,
+      "grad_norm": 2.7910832796646248,
+      "learning_rate": 1.8890766808634476e-05,
+      "loss": 0.9081,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5818551719753322,
+      "grad_norm": 2.405785156532243,
+      "learning_rate": 1.888819876818206e-05,
+      "loss": 0.8213,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5823637866361497,
+      "grad_norm": 2.5059231209618913,
+      "learning_rate": 1.8885627933434683e-05,
+      "loss": 0.7578,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5828724012969674,
+      "grad_norm": 2.126112029876701,
+      "learning_rate": 1.8883054305200573e-05,
+      "loss": 0.9348,
+      "step": 1146
+    },
+    {
+      "epoch": 0.583381015957785,
+      "grad_norm": 2.409399275403861,
+      "learning_rate": 1.8880477884288835e-05,
+      "loss": 0.7448,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5838896306186026,
+      "grad_norm": 2.436982906953132,
+      "learning_rate": 1.887789867150945e-05,
+      "loss": 0.8545,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5843982452794202,
+      "grad_norm": 3.3906951756400763,
+      "learning_rate": 1.8875316667673276e-05,
+      "loss": 0.7674,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5849068599402377,
+      "grad_norm": 2.351340552720287,
+      "learning_rate": 1.8872731873592053e-05,
+      "loss": 0.8637,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5854154746010554,
+      "grad_norm": 2.7187756065828963,
+      "learning_rate": 1.887014429007839e-05,
+      "loss": 0.7817,
+      "step": 1151
+    },
+    {
+      "epoch": 0.585924089261873,
+      "grad_norm": 2.184953460721486,
+      "learning_rate": 1.886755391794578e-05,
+      "loss": 0.818,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5864327039226905,
+      "grad_norm": 2.1742809334736934,
+      "learning_rate": 1.8864960758008592e-05,
+      "loss": 0.8968,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5869413185835082,
+      "grad_norm": 2.9799372247830087,
+      "learning_rate": 1.8862364811082076e-05,
+      "loss": 0.7903,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5874499332443258,
+      "grad_norm": 2.579456979397647,
+      "learning_rate": 1.885976607798234e-05,
+      "loss": 0.6797,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5879585479051433,
+      "grad_norm": 2.5026166573542565,
+      "learning_rate": 1.885716455952639e-05,
+      "loss": 0.9662,
+      "step": 1156
+    },
+    {
+      "epoch": 0.588467162565961,
+      "grad_norm": 3.1838253007243074,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.7406,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5889757772267785,
+      "grad_norm": 9.652680857756524,
+      "learning_rate": 1.885195316981821e-05,
+      "loss": 0.9095,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5894843918875962,
+      "grad_norm": 2.716815654983168,
+      "learning_rate": 1.8849343300204347e-05,
+      "loss": 0.8688,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5899930065484138,
+      "grad_norm": 3.2996906713525935,
+      "learning_rate": 1.8846730648511005e-05,
+      "loss": 0.7235,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5905016212092313,
+      "grad_norm": 2.4007031801380103,
+      "learning_rate": 1.884411521555956e-05,
+      "loss": 0.7344,
+      "step": 1161
+    },
+    {
+      "epoch": 0.591010235870049,
+      "grad_norm": 2.5346929423645683,
+      "learning_rate": 1.884149700217226e-05,
+      "loss": 0.8029,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5915188505308665,
+      "grad_norm": 2.713459084414082,
+      "learning_rate": 1.883887600917222e-05,
+      "loss": 0.7182,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5920274651916841,
+      "grad_norm": 2.923825026353151,
+      "learning_rate": 1.8836252237383436e-05,
+      "loss": 0.7354,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5925360798525018,
+      "grad_norm": 2.004029149382186,
+      "learning_rate": 1.883362568763078e-05,
+      "loss": 0.7902,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5930446945133193,
+      "grad_norm": 2.7769785451805467,
+      "learning_rate": 1.883099636073999e-05,
+      "loss": 0.8612,
+      "step": 1166
+    },
+    {
+      "epoch": 0.593553309174137,
+      "grad_norm": 2.2388648958185295,
+      "learning_rate": 1.8828364257537683e-05,
+      "loss": 0.7041,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5940619238349546,
+      "grad_norm": 3.0785150426063717,
+      "learning_rate": 1.8825729378851347e-05,
+      "loss": 0.8356,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5945705384957721,
+      "grad_norm": 2.5233360246503715,
+      "learning_rate": 1.8823091725509335e-05,
+      "loss": 0.6971,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5950791531565898,
+      "grad_norm": 2.638316392424465,
+      "learning_rate": 1.8820451298340887e-05,
+      "loss": 0.7582,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5955877678174073,
+      "grad_norm": 2.251939573354715,
+      "learning_rate": 1.8817808098176106e-05,
+      "loss": 0.7993,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5960963824782249,
+      "grad_norm": 2.6785846364740165,
+      "learning_rate": 1.8815162125845967e-05,
+      "loss": 0.8739,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5966049971390426,
+      "grad_norm": 2.266396121633963,
+      "learning_rate": 1.8812513382182315e-05,
+      "loss": 0.8029,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5971136117998601,
+      "grad_norm": 2.6339507316784,
+      "learning_rate": 1.8809861868017874e-05,
+      "loss": 0.8873,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5976222264606778,
+      "grad_norm": 2.1759301619191644,
+      "learning_rate": 1.8807207584186234e-05,
+      "loss": 0.8522,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5981308411214953,
+      "grad_norm": 2.3352272204133526,
+      "learning_rate": 1.8804550531521847e-05,
+      "loss": 0.8004,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5986394557823129,
+      "grad_norm": 2.266812033292793,
+      "learning_rate": 1.8801890710860054e-05,
+      "loss": 0.7427,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5991480704431306,
+      "grad_norm": 2.272575083752241,
+      "learning_rate": 1.879922812303705e-05,
+      "loss": 0.8537,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5996566851039481,
+      "grad_norm": 2.411616397782809,
+      "learning_rate": 1.879656276888991e-05,
+      "loss": 0.8872,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6001652997647657,
+      "grad_norm": 2.4972440788190444,
+      "learning_rate": 1.8793894649256575e-05,
+      "loss": 0.783,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6006739144255833,
+      "grad_norm": 2.317257703770997,
+      "learning_rate": 1.879122376497585e-05,
+      "loss": 0.8748,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6011825290864009,
+      "grad_norm": 2.55446528130303,
+      "learning_rate": 1.8788550116887424e-05,
+      "loss": 0.7243,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6016911437472185,
+      "grad_norm": 2.4918013605393474,
+      "learning_rate": 1.8785873705831835e-05,
+      "loss": 0.8018,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6021997584080361,
+      "grad_norm": 2.4025069018531764,
+      "learning_rate": 1.878319453265051e-05,
+      "loss": 0.8383,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6027083730688537,
+      "grad_norm": 3.4723740625030373,
+      "learning_rate": 1.8780512598185727e-05,
+      "loss": 0.7492,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6032169877296714,
+      "grad_norm": 2.043103186303328,
+      "learning_rate": 1.877782790328064e-05,
+      "loss": 0.682,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6037256023904889,
+      "grad_norm": 2.42788138523744,
+      "learning_rate": 1.877514044877927e-05,
+      "loss": 0.646,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6042342170513065,
+      "grad_norm": 2.2618639014910706,
+      "learning_rate": 1.8772450235526513e-05,
+      "loss": 0.7202,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6047428317121241,
+      "grad_norm": 2.7895480035240197,
+      "learning_rate": 1.876975726436812e-05,
+      "loss": 0.7297,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6052514463729417,
+      "grad_norm": 2.4574112105604318,
+      "learning_rate": 1.8767061536150716e-05,
+      "loss": 0.867,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6057600610337593,
+      "grad_norm": 2.2780616113035705,
+      "learning_rate": 1.8764363051721785e-05,
+      "loss": 0.7826,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6062686756945769,
+      "grad_norm": 2.9614420035752094,
+      "learning_rate": 1.8761661811929686e-05,
+      "loss": 0.8516,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6067772903553945,
+      "grad_norm": 2.3245914961739995,
+      "learning_rate": 1.8758957817623652e-05,
+      "loss": 0.7317,
+      "step": 1193
+    },
+    {
+      "epoch": 0.607285905016212,
+      "grad_norm": 2.1709767412005454,
+      "learning_rate": 1.875625106965376e-05,
+      "loss": 0.7887,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6077945196770297,
+      "grad_norm": 2.206832903836197,
+      "learning_rate": 1.8753541568870968e-05,
+      "loss": 0.8165,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6083031343378473,
+      "grad_norm": 2.3742853394459376,
+      "learning_rate": 1.8750829316127098e-05,
+      "loss": 0.7595,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6088117489986649,
+      "grad_norm": 2.121960991309434,
+      "learning_rate": 1.874811431227483e-05,
+      "loss": 0.6613,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6093203636594825,
+      "grad_norm": 2.069663698271797,
+      "learning_rate": 1.8745396558167716e-05,
+      "loss": 0.7434,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6098289783203,
+      "grad_norm": 2.405096198555101,
+      "learning_rate": 1.8742676054660175e-05,
+      "loss": 0.8795,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6103375929811177,
+      "grad_norm": 3.028120017662713,
+      "learning_rate": 1.873995280260748e-05,
+      "loss": 0.7608,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6103375929811177,
+      "eval_loss": 0.7444096803665161,
+      "eval_runtime": 169.8372,
+      "eval_samples_per_second": 19.501,
+      "eval_steps_per_second": 9.751,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6108462076419353,
+      "grad_norm": 3.2428719039494616,
+      "learning_rate": 1.8737226802865778e-05,
+      "loss": 0.7476,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6113548223027528,
+      "grad_norm": 2.0021274695425553,
+      "learning_rate": 1.8734498056292074e-05,
+      "loss": 0.7236,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6118634369635705,
+      "grad_norm": 2.027970586461856,
+      "learning_rate": 1.8731766563744238e-05,
+      "loss": 0.7663,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6123720516243881,
+      "grad_norm": 2.278223721598271,
+      "learning_rate": 1.8729032326081004e-05,
+      "loss": 0.7572,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6128806662852057,
+      "grad_norm": 2.315483025073666,
+      "learning_rate": 1.8726295344161972e-05,
+      "loss": 0.7216,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6133892809460233,
+      "grad_norm": 2.1879298196587462,
+      "learning_rate": 1.8723555618847596e-05,
+      "loss": 0.6846,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6138978956068408,
+      "grad_norm": 2.7117773508539096,
+      "learning_rate": 1.8720813150999202e-05,
+      "loss": 0.7472,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6144065102676585,
+      "grad_norm": 2.4614943101481264,
+      "learning_rate": 1.8718067941478973e-05,
+      "loss": 0.9084,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6149151249284761,
+      "grad_norm": 2.51983470420044,
+      "learning_rate": 1.8715319991149954e-05,
+      "loss": 0.7696,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6154237395892936,
+      "grad_norm": 2.117318575134219,
+      "learning_rate": 1.8712569300876053e-05,
+      "loss": 0.7207,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6159323542501113,
+      "grad_norm": 2.170489630944389,
+      "learning_rate": 1.870981587152204e-05,
+      "loss": 0.7181,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6164409689109288,
+      "grad_norm": 2.5786452259814747,
+      "learning_rate": 1.8707059703953545e-05,
+      "loss": 0.797,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6169495835717465,
+      "grad_norm": 2.1630192524639122,
+      "learning_rate": 1.8704300799037062e-05,
+      "loss": 0.8371,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6174581982325641,
+      "grad_norm": 2.8152887717293047,
+      "learning_rate": 1.8701539157639937e-05,
+      "loss": 0.8711,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6179668128933816,
+      "grad_norm": 2.111210229565699,
+      "learning_rate": 1.8698774780630386e-05,
+      "loss": 0.7422,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6184754275541993,
+      "grad_norm": 2.0752123161068092,
+      "learning_rate": 1.8696007668877478e-05,
+      "loss": 0.8336,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6189840422150168,
+      "grad_norm": 2.0099857902973506,
+      "learning_rate": 1.8693237823251146e-05,
+      "loss": 0.8091,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6194926568758344,
+      "grad_norm": 2.359748457094891,
+      "learning_rate": 1.8690465244622184e-05,
+      "loss": 0.7486,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6200012715366521,
+      "grad_norm": 2.3844780775915404,
+      "learning_rate": 1.8687689933862238e-05,
+      "loss": 0.7739,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6205098861974696,
+      "grad_norm": 2.4924311503827328,
+      "learning_rate": 1.868491189184382e-05,
+      "loss": 0.7947,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6210185008582872,
+      "grad_norm": 2.1269094079098823,
+      "learning_rate": 1.8682131119440302e-05,
+      "loss": 0.737,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6215271155191049,
+      "grad_norm": 2.077188151580263,
+      "learning_rate": 1.8679347617525905e-05,
+      "loss": 0.7438,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6220357301799224,
+      "grad_norm": 2.6935341980913927,
+      "learning_rate": 1.8676561386975713e-05,
+      "loss": 0.8546,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6225443448407401,
+      "grad_norm": 2.2923707978459538,
+      "learning_rate": 1.8673772428665666e-05,
+      "loss": 0.7251,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6230529595015576,
+      "grad_norm": 2.3073063857826805,
+      "learning_rate": 1.8670980743472575e-05,
+      "loss": 0.9788,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6235615741623752,
+      "grad_norm": 2.428663031669673,
+      "learning_rate": 1.866818633227408e-05,
+      "loss": 0.6999,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6240701888231929,
+      "grad_norm": 2.310612604233354,
+      "learning_rate": 1.8665389195948715e-05,
+      "loss": 0.8253,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6245788034840104,
+      "grad_norm": 2.308396900646751,
+      "learning_rate": 1.866258933537584e-05,
+      "loss": 0.7914,
+      "step": 1228
+    },
+    {
+      "epoch": 0.625087418144828,
+      "grad_norm": 2.335476730985372,
+      "learning_rate": 1.8659786751435683e-05,
+      "loss": 0.8129,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6255960328056456,
+      "grad_norm": 2.2285434410238167,
+      "learning_rate": 1.8656981445009327e-05,
+      "loss": 0.6812,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6261046474664632,
+      "grad_norm": 1.941571848386482,
+      "learning_rate": 1.8654173416978712e-05,
+      "loss": 0.7504,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6266132621272809,
+      "grad_norm": 2.0875061080751887,
+      "learning_rate": 1.865136266822664e-05,
+      "loss": 0.7865,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6271218767880984,
+      "grad_norm": 2.4689468112189896,
+      "learning_rate": 1.8648549199636748e-05,
+      "loss": 0.8565,
+      "step": 1233
+    },
+    {
+      "epoch": 0.627630491448916,
+      "grad_norm": 2.25710774018423,
+      "learning_rate": 1.864573301209355e-05,
+      "loss": 0.7634,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6281391061097336,
+      "grad_norm": 2.404549232954566,
+      "learning_rate": 1.864291410648241e-05,
+      "loss": 0.7918,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6286477207705512,
+      "grad_norm": 2.3587814493917985,
+      "learning_rate": 1.8640092483689534e-05,
+      "loss": 0.7645,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6291563354313688,
+      "grad_norm": 2.244979874110758,
+      "learning_rate": 1.8637268144601996e-05,
+      "loss": 0.848,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6296649500921864,
+      "grad_norm": 7.371130493913452,
+      "learning_rate": 1.8634441090107722e-05,
+      "loss": 0.8901,
+      "step": 1238
+    },
+    {
+      "epoch": 0.630173564753004,
+      "grad_norm": 2.304726603548033,
+      "learning_rate": 1.8631611321095482e-05,
+      "loss": 0.835,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6306821794138217,
+      "grad_norm": 2.120365137108877,
+      "learning_rate": 1.862877883845491e-05,
+      "loss": 0.7926,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6311907940746392,
+      "grad_norm": 3.5207440216353914,
+      "learning_rate": 1.862594364307649e-05,
+      "loss": 0.7847,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6316994087354568,
+      "grad_norm": 2.277164511063972,
+      "learning_rate": 1.8623105735851553e-05,
+      "loss": 0.757,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6322080233962744,
+      "grad_norm": 2.3904316144527633,
+      "learning_rate": 1.8620265117672293e-05,
+      "loss": 0.7896,
+      "step": 1243
+    },
+    {
+      "epoch": 0.632716638057092,
+      "grad_norm": 2.3290782231535014,
+      "learning_rate": 1.8617421789431745e-05,
+      "loss": 0.8401,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6332252527179096,
+      "grad_norm": 2.154602637774047,
+      "learning_rate": 1.8614575752023814e-05,
+      "loss": 0.7395,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6337338673787272,
+      "grad_norm": 2.3605974074880405,
+      "learning_rate": 1.861172700634323e-05,
+      "loss": 0.6496,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6342424820395448,
+      "grad_norm": 2.270389151805207,
+      "learning_rate": 1.86088755532856e-05,
+      "loss": 0.8323,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6347510967003623,
+      "grad_norm": 2.223814972329858,
+      "learning_rate": 1.8606021393747364e-05,
+      "loss": 0.8094,
+      "step": 1248
+    },
+    {
+      "epoch": 0.63525971136118,
+      "grad_norm": 6.719633390326486,
+      "learning_rate": 1.8603164528625822e-05,
+      "loss": 0.9702,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6357683260219976,
+      "grad_norm": 2.439272700556095,
+      "learning_rate": 1.8600304958819127e-05,
+      "loss": 0.8659,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6362769406828152,
+      "grad_norm": 2.1847835567999585,
+      "learning_rate": 1.8597442685226274e-05,
+      "loss": 0.8374,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6367855553436328,
+      "grad_norm": 2.2898067787645155,
+      "learning_rate": 1.8594577708747107e-05,
+      "loss": 0.7325,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6372941700044504,
+      "grad_norm": 2.6804481071938113,
+      "learning_rate": 1.8591710030282337e-05,
+      "loss": 0.7687,
+      "step": 1253
+    },
+    {
+      "epoch": 0.637802784665268,
+      "grad_norm": 2.110378676509067,
+      "learning_rate": 1.8588839650733503e-05,
+      "loss": 0.8186,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6383113993260856,
+      "grad_norm": 2.577281837147583,
+      "learning_rate": 1.8585966571003004e-05,
+      "loss": 0.8882,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6388200139869031,
+      "grad_norm": 2.4773104046779215,
+      "learning_rate": 1.8583090791994086e-05,
+      "loss": 0.866,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6393286286477208,
+      "grad_norm": 2.3310596991789967,
+      "learning_rate": 1.8580212314610847e-05,
+      "loss": 0.7138,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6398372433085384,
+      "grad_norm": 2.2184156783842,
+      "learning_rate": 1.8577331139758228e-05,
+      "loss": 0.7171,
+      "step": 1258
+    },
+    {
+      "epoch": 0.640345857969356,
+      "grad_norm": 3.0695257756278336,
+      "learning_rate": 1.857444726834202e-05,
+      "loss": 0.799,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6408544726301736,
+      "grad_norm": 2.8026477760467134,
+      "learning_rate": 1.857156070126886e-05,
+      "loss": 0.821,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6413630872909911,
+      "grad_norm": 2.3628527544454974,
+      "learning_rate": 1.8568671439446237e-05,
+      "loss": 0.7825,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6418717019518088,
+      "grad_norm": 1.9899639810724152,
+      "learning_rate": 1.8565779483782485e-05,
+      "loss": 0.7647,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6423803166126264,
+      "grad_norm": 2.3992425518041514,
+      "learning_rate": 1.8562884835186786e-05,
+      "loss": 0.717,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6428889312734439,
+      "grad_norm": 2.410957487645244,
+      "learning_rate": 1.8559987494569165e-05,
+      "loss": 0.7071,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6433975459342616,
+      "grad_norm": 2.5908701892225987,
+      "learning_rate": 1.8557087462840493e-05,
+      "loss": 0.938,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6439061605950791,
+      "grad_norm": 2.4080304581246184,
+      "learning_rate": 1.8554184740912494e-05,
+      "loss": 0.7029,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6444147752558967,
+      "grad_norm": 2.4642950496141354,
+      "learning_rate": 1.8551279329697728e-05,
+      "loss": 0.6958,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6449233899167144,
+      "grad_norm": 1.9437969386084806,
+      "learning_rate": 1.854837123010961e-05,
+      "loss": 0.7333,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6454320045775319,
+      "grad_norm": 2.1082332205342715,
+      "learning_rate": 1.8545460443062397e-05,
+      "loss": 0.741,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6459406192383496,
+      "grad_norm": 2.1860564782332115,
+      "learning_rate": 1.8542546969471184e-05,
+      "loss": 0.7727,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6464492338991672,
+      "grad_norm": 2.3417506527574496,
+      "learning_rate": 1.8539630810251924e-05,
+      "loss": 0.7462,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6469578485599847,
+      "grad_norm": 2.603242715324043,
+      "learning_rate": 1.85367119663214e-05,
+      "loss": 0.8303,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6474664632208024,
+      "grad_norm": 2.312338024343427,
+      "learning_rate": 1.853379043859725e-05,
+      "loss": 0.8254,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6479750778816199,
+      "grad_norm": 2.0386754395716347,
+      "learning_rate": 1.8530866227997945e-05,
+      "loss": 0.7491,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6484836925424375,
+      "grad_norm": 2.686392711258498,
+      "learning_rate": 1.8527939335442814e-05,
+      "loss": 0.8144,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6489923072032552,
+      "grad_norm": 2.2316264072781933,
+      "learning_rate": 1.852500976185202e-05,
+      "loss": 0.7709,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6495009218640727,
+      "grad_norm": 2.289390045791411,
+      "learning_rate": 1.852207750814656e-05,
+      "loss": 0.7006,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6500095365248904,
+      "grad_norm": 2.3873150419322613,
+      "learning_rate": 1.85191425752483e-05,
+      "loss": 0.7938,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6505181511857079,
+      "grad_norm": 2.4250108659638463,
+      "learning_rate": 1.851620496407992e-05,
+      "loss": 0.8135,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6510267658465255,
+      "grad_norm": 2.2388198497446363,
+      "learning_rate": 1.8513264675564953e-05,
+      "loss": 0.7615,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6515353805073432,
+      "grad_norm": 2.955183976541068,
+      "learning_rate": 1.851032171062778e-05,
+      "loss": 0.7952,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6520439951681607,
+      "grad_norm": 2.4228686847566085,
+      "learning_rate": 1.8507376070193616e-05,
+      "loss": 0.7534,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6525526098289783,
+      "grad_norm": 4.215811603281773,
+      "learning_rate": 1.850442775518852e-05,
+      "loss": 0.8192,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 2.3095199353649547,
+      "learning_rate": 1.8501476766539396e-05,
+      "loss": 0.8484,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6535698391506135,
+      "grad_norm": 2.249193046823114,
+      "learning_rate": 1.8498523105173972e-05,
+      "loss": 0.7503,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6540784538114311,
+      "grad_norm": 2.321378828920088,
+      "learning_rate": 1.849556677202083e-05,
+      "loss": 0.7752,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6545870684722487,
+      "grad_norm": 2.2378547088942384,
+      "learning_rate": 1.8492607768009397e-05,
+      "loss": 0.7588,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6550956831330663,
+      "grad_norm": 2.381967863876634,
+      "learning_rate": 1.848964609406993e-05,
+      "loss": 0.8644,
+      "step": 1288
+    },
+    {
+      "epoch": 0.655604297793884,
+      "grad_norm": 2.5283785887330774,
+      "learning_rate": 1.8486681751133525e-05,
+      "loss": 0.7544,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6561129124547015,
+      "grad_norm": 2.2408629436056873,
+      "learning_rate": 1.848371474013212e-05,
+      "loss": 0.7405,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6566215271155191,
+      "grad_norm": 2.1553010925814093,
+      "learning_rate": 1.8480745061998492e-05,
+      "loss": 0.7875,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6571301417763367,
+      "grad_norm": 2.358425934751709,
+      "learning_rate": 1.8477772717666257e-05,
+      "loss": 0.8342,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6576387564371543,
+      "grad_norm": 2.5268463635856486,
+      "learning_rate": 1.847479770806987e-05,
+      "loss": 0.7645,
+      "step": 1293
+    },
+    {
+      "epoch": 0.658147371097972,
+      "grad_norm": 2.432707452190484,
+      "learning_rate": 1.8471820034144616e-05,
+      "loss": 0.6782,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6586559857587895,
+      "grad_norm": 2.119282716830438,
+      "learning_rate": 1.846883969682663e-05,
+      "loss": 0.7942,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6591646004196071,
+      "grad_norm": 2.6359797141375587,
+      "learning_rate": 1.846585669705288e-05,
+      "loss": 0.7918,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6596732150804246,
+      "grad_norm": 2.398866894583709,
+      "learning_rate": 1.846287103576116e-05,
+      "loss": 0.7837,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6601818297412423,
+      "grad_norm": 2.534440467073578,
+      "learning_rate": 1.845988271389012e-05,
+      "loss": 0.6615,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6606904444020599,
+      "grad_norm": 2.225819460379174,
+      "learning_rate": 1.8456891732379226e-05,
+      "loss": 0.5588,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6611990590628775,
+      "grad_norm": 2.543159442554939,
+      "learning_rate": 1.8453898092168803e-05,
+      "loss": 0.8138,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6611990590628775,
+      "eval_loss": 0.7401166558265686,
+      "eval_runtime": 170.051,
+      "eval_samples_per_second": 19.477,
+      "eval_steps_per_second": 9.738,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6617076737236951,
+      "grad_norm": 2.4178430281140777,
+      "learning_rate": 1.8450901794199988e-05,
+      "loss": 0.7512,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6622162883845126,
+      "grad_norm": 2.203322273242025,
+      "learning_rate": 1.844790283941477e-05,
+      "loss": 0.8166,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6627249030453303,
+      "grad_norm": 2.218177184186777,
+      "learning_rate": 1.844490122875597e-05,
+      "loss": 0.7542,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6632335177061479,
+      "grad_norm": 2.308906443633891,
+      "learning_rate": 1.8441896963167235e-05,
+      "loss": 0.8058,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6637421323669654,
+      "grad_norm": 2.6741134457885787,
+      "learning_rate": 1.8438890043593062e-05,
+      "loss": 0.8599,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6642507470277831,
+      "grad_norm": 2.1113827792706337,
+      "learning_rate": 1.843588047097877e-05,
+      "loss": 0.8423,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6647593616886007,
+      "grad_norm": 2.453797072174293,
+      "learning_rate": 1.843286824627052e-05,
+      "loss": 0.6286,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6652679763494183,
+      "grad_norm": 2.418435687951576,
+      "learning_rate": 1.8429853370415296e-05,
+      "loss": 0.7867,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6657765910102359,
+      "grad_norm": 2.2766876584263294,
+      "learning_rate": 1.8426835844360928e-05,
+      "loss": 0.8963,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6662852056710534,
+      "grad_norm": 2.386778485788358,
+      "learning_rate": 1.842381566905607e-05,
+      "loss": 0.8327,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6667938203318711,
+      "grad_norm": 2.5992709311409077,
+      "learning_rate": 1.842079284545022e-05,
+      "loss": 0.7505,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6673024349926887,
+      "grad_norm": 2.1456763818123945,
+      "learning_rate": 1.841776737449369e-05,
+      "loss": 0.7783,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6678110496535062,
+      "grad_norm": 3.0122880409986292,
+      "learning_rate": 1.8414739257137646e-05,
+      "loss": 0.8466,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6683196643143239,
+      "grad_norm": 2.4234916951497816,
+      "learning_rate": 1.841170849433407e-05,
+      "loss": 0.775,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6688282789751414,
+      "grad_norm": 2.521725478342668,
+      "learning_rate": 1.840867508703578e-05,
+      "loss": 0.7632,
+      "step": 1315
+    },
+    {
+      "epoch": 0.669336893635959,
+      "grad_norm": 2.2076782358519096,
+      "learning_rate": 1.8405639036196427e-05,
+      "loss": 0.8183,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6698455082967767,
+      "grad_norm": 2.1342674574560836,
+      "learning_rate": 1.8402600342770495e-05,
+      "loss": 0.883,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6703541229575942,
+      "grad_norm": 2.2110576260017485,
+      "learning_rate": 1.8399559007713296e-05,
+      "loss": 0.8405,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6708627376184119,
+      "grad_norm": 2.140590361154791,
+      "learning_rate": 1.839651503198097e-05,
+      "loss": 0.9328,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6713713522792294,
+      "grad_norm": 1.9530572498014995,
+      "learning_rate": 1.839346841653049e-05,
+      "loss": 0.777,
+      "step": 1320
+    },
+    {
+      "epoch": 0.671879966940047,
+      "grad_norm": 2.3843568893684743,
+      "learning_rate": 1.8390419162319663e-05,
+      "loss": 0.7975,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6723885816008647,
+      "grad_norm": 2.1489036470507146,
+      "learning_rate": 1.8387367270307122e-05,
+      "loss": 0.7639,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6728971962616822,
+      "grad_norm": 2.3325127339398812,
+      "learning_rate": 1.8384312741452322e-05,
+      "loss": 0.8481,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6734058109224998,
+      "grad_norm": 2.2957962513148393,
+      "learning_rate": 1.838125557671556e-05,
+      "loss": 0.7147,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6739144255833175,
+      "grad_norm": 2.6548257936556636,
+      "learning_rate": 1.837819577705795e-05,
+      "loss": 0.821,
+      "step": 1325
+    },
+    {
+      "epoch": 0.674423040244135,
+      "grad_norm": 2.843998824751084,
+      "learning_rate": 1.8375133343441445e-05,
+      "loss": 0.7894,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6749316549049527,
+      "grad_norm": 2.0345889305690332,
+      "learning_rate": 1.8372068276828814e-05,
+      "loss": 0.7992,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6754402695657702,
+      "grad_norm": 2.140707306776177,
+      "learning_rate": 1.836900057818367e-05,
+      "loss": 0.8266,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6759488842265878,
+      "grad_norm": 2.660370626227948,
+      "learning_rate": 1.8365930248470436e-05,
+      "loss": 0.8113,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6764574988874055,
+      "grad_norm": 2.386252699581196,
+      "learning_rate": 1.8362857288654374e-05,
+      "loss": 0.9403,
+      "step": 1330
+    },
+    {
+      "epoch": 0.676966113548223,
+      "grad_norm": 2.7613716824660868,
+      "learning_rate": 1.8359781699701573e-05,
+      "loss": 0.727,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6774747282090406,
+      "grad_norm": 1.9599452452893857,
+      "learning_rate": 1.8356703482578934e-05,
+      "loss": 0.733,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6779833428698582,
+      "grad_norm": 2.4881538106960255,
+      "learning_rate": 1.83536226382542e-05,
+      "loss": 0.7639,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6784919575306758,
+      "grad_norm": 2.2341577384264744,
+      "learning_rate": 1.835053916769594e-05,
+      "loss": 0.7634,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6790005721914935,
+      "grad_norm": 3.67719743387711,
+      "learning_rate": 1.8347453071873536e-05,
+      "loss": 1.0182,
+      "step": 1335
+    },
+    {
+      "epoch": 0.679509186852311,
+      "grad_norm": 2.0280277222990146,
+      "learning_rate": 1.8344364351757206e-05,
+      "loss": 0.8374,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6800178015131286,
+      "grad_norm": 2.4584242809304415,
+      "learning_rate": 1.8341273008317992e-05,
+      "loss": 0.7392,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6805264161739463,
+      "grad_norm": 3.553745958323936,
+      "learning_rate": 1.833817904252775e-05,
+      "loss": 0.7307,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6810350308347638,
+      "grad_norm": 2.328018595676939,
+      "learning_rate": 1.8335082455359184e-05,
+      "loss": 0.6939,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6815436454955814,
+      "grad_norm": 2.3203917210473803,
+      "learning_rate": 1.8331983247785794e-05,
+      "loss": 0.7586,
+      "step": 1340
+    },
+    {
+      "epoch": 0.682052260156399,
+      "grad_norm": 2.8508473792943207,
+      "learning_rate": 1.832888142078192e-05,
+      "loss": 0.6707,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6825608748172166,
+      "grad_norm": 2.408446857280281,
+      "learning_rate": 1.8325776975322725e-05,
+      "loss": 0.8052,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6830694894780343,
+      "grad_norm": 2.3232606596050003,
+      "learning_rate": 1.8322669912384194e-05,
+      "loss": 0.8802,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6835781041388518,
+      "grad_norm": 2.068485591160505,
+      "learning_rate": 1.8319560232943128e-05,
+      "loss": 0.6699,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6840867187996694,
+      "grad_norm": 5.044898812495679,
+      "learning_rate": 1.831644793797716e-05,
+      "loss": 0.7444,
+      "step": 1345
+    },
+    {
+      "epoch": 0.684595333460487,
+      "grad_norm": 2.747496418972302,
+      "learning_rate": 1.831333302846474e-05,
+      "loss": 0.9403,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6851039481213046,
+      "grad_norm": 2.4697418152084363,
+      "learning_rate": 1.831021550538514e-05,
+      "loss": 0.8629,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6856125627821222,
+      "grad_norm": 2.219042476665246,
+      "learning_rate": 1.8307095369718456e-05,
+      "loss": 0.7447,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6861211774429398,
+      "grad_norm": 2.543355932806091,
+      "learning_rate": 1.830397262244561e-05,
+      "loss": 0.79,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6866297921037574,
+      "grad_norm": 3.057522991901827,
+      "learning_rate": 1.830084726454833e-05,
+      "loss": 0.8196,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6871384067645749,
+      "grad_norm": 2.139571327341698,
+      "learning_rate": 1.8297719297009177e-05,
+      "loss": 0.7025,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6876470214253926,
+      "grad_norm": 2.2643060462386075,
+      "learning_rate": 1.8294588720811536e-05,
+      "loss": 0.7788,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6881556360862102,
+      "grad_norm": 2.30238749938551,
+      "learning_rate": 1.8291455536939596e-05,
+      "loss": 0.7404,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6886642507470278,
+      "grad_norm": 2.2615558783041685,
+      "learning_rate": 1.8288319746378383e-05,
+      "loss": 0.772,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6891728654078454,
+      "grad_norm": 8.474508319646965,
+      "learning_rate": 1.8285181350113727e-05,
+      "loss": 0.8317,
+      "step": 1355
+    },
+    {
+      "epoch": 0.689681480068663,
+      "grad_norm": 2.0896691935715594,
+      "learning_rate": 1.8282040349132295e-05,
+      "loss": 0.8108,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6901900947294806,
+      "grad_norm": 2.106574226576509,
+      "learning_rate": 1.8278896744421558e-05,
+      "loss": 0.7472,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6906987093902982,
+      "grad_norm": 3.273928657775116,
+      "learning_rate": 1.8275750536969807e-05,
+      "loss": 0.77,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6912073240511157,
+      "grad_norm": 3.4434056270823032,
+      "learning_rate": 1.8272601727766163e-05,
+      "loss": 0.7087,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6917159387119334,
+      "grad_norm": 2.2248488942816937,
+      "learning_rate": 1.826945031780055e-05,
+      "loss": 0.7787,
+      "step": 1360
+    },
+    {
+      "epoch": 0.692224553372751,
+      "grad_norm": 2.337497458227071,
+      "learning_rate": 1.826629630806372e-05,
+      "loss": 0.7378,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6927331680335685,
+      "grad_norm": 2.1576851754049082,
+      "learning_rate": 1.8263139699547236e-05,
+      "loss": 0.859,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6932417826943862,
+      "grad_norm": 2.233313461708748,
+      "learning_rate": 1.8259980493243484e-05,
+      "loss": 0.8632,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6937503973552037,
+      "grad_norm": 2.1713361551939716,
+      "learning_rate": 1.825681869014566e-05,
+      "loss": 0.7839,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6942590120160214,
+      "grad_norm": 2.613625268573356,
+      "learning_rate": 1.825365429124778e-05,
+      "loss": 0.6702,
+      "step": 1365
+    },
+    {
+      "epoch": 0.694767626676839,
+      "grad_norm": 2.3727525818995305,
+      "learning_rate": 1.8250487297544682e-05,
+      "loss": 0.7672,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6952762413376565,
+      "grad_norm": 2.5886784804561964,
+      "learning_rate": 1.824731771003201e-05,
+      "loss": 0.8112,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6957848559984742,
+      "grad_norm": 2.4162408080816697,
+      "learning_rate": 1.8244145529706223e-05,
+      "loss": 0.791,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6962934706592917,
+      "grad_norm": 2.3109903305029693,
+      "learning_rate": 1.8240970757564606e-05,
+      "loss": 0.7939,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6968020853201093,
+      "grad_norm": 2.6030843596474478,
+      "learning_rate": 1.8237793394605245e-05,
+      "loss": 0.8294,
+      "step": 1370
+    },
+    {
+      "epoch": 0.697310699980927,
+      "grad_norm": 2.300811508316524,
+      "learning_rate": 1.823461344182705e-05,
+      "loss": 0.8593,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6978193146417445,
+      "grad_norm": 2.4603683540150985,
+      "learning_rate": 1.823143090022975e-05,
+      "loss": 0.8301,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6983279293025622,
+      "grad_norm": 2.101000515697621,
+      "learning_rate": 1.822824577081387e-05,
+      "loss": 0.7044,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6988365439633798,
+      "grad_norm": 2.381165582557243,
+      "learning_rate": 1.8225058054580765e-05,
+      "loss": 0.7841,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6993451586241973,
+      "grad_norm": 2.503683999801456,
+      "learning_rate": 1.8221867752532597e-05,
+      "loss": 0.7802,
+      "step": 1375
+    },
+    {
+      "epoch": 0.699853773285015,
+      "grad_norm": 2.463290297821008,
+      "learning_rate": 1.8218674865672342e-05,
+      "loss": 0.789,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7003623879458325,
+      "grad_norm": 2.1643246929106486,
+      "learning_rate": 1.8215479395003787e-05,
+      "loss": 0.7379,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7008710026066501,
+      "grad_norm": 2.4517364459693427,
+      "learning_rate": 1.821228134153153e-05,
+      "loss": 0.8162,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7013796172674678,
+      "grad_norm": 2.5645643153493882,
+      "learning_rate": 1.8209080706260988e-05,
+      "loss": 0.7974,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7018882319282853,
+      "grad_norm": 2.2285908344492844,
+      "learning_rate": 1.8205877490198382e-05,
+      "loss": 0.784,
+      "step": 1380
+    },
+    {
+      "epoch": 0.702396846589103,
+      "grad_norm": 2.222934913262436,
+      "learning_rate": 1.8202671694350747e-05,
+      "loss": 0.6794,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7029054612499205,
+      "grad_norm": 2.19360284991034,
+      "learning_rate": 1.8199463319725935e-05,
+      "loss": 0.7846,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7034140759107381,
+      "grad_norm": 2.1643208373626672,
+      "learning_rate": 1.8196252367332595e-05,
+      "loss": 0.7437,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7039226905715558,
+      "grad_norm": 2.122468000873956,
+      "learning_rate": 1.81930388381802e-05,
+      "loss": 0.7662,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7044313052323733,
+      "grad_norm": 2.0737223632884034,
+      "learning_rate": 1.8189822733279025e-05,
+      "loss": 0.7088,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7049399198931909,
+      "grad_norm": 2.312497422500411,
+      "learning_rate": 1.8186604053640154e-05,
+      "loss": 0.7804,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7054485345540085,
+      "grad_norm": 2.4703577335470173,
+      "learning_rate": 1.8183382800275492e-05,
+      "loss": 0.8075,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7059571492148261,
+      "grad_norm": 2.174235207279163,
+      "learning_rate": 1.818015897419774e-05,
+      "loss": 0.722,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7064657638756437,
+      "grad_norm": 2.102439612248265,
+      "learning_rate": 1.8176932576420415e-05,
+      "loss": 0.7406,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7069743785364613,
+      "grad_norm": 2.369638112441549,
+      "learning_rate": 1.8173703607957837e-05,
+      "loss": 0.7977,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7074829931972789,
+      "grad_norm": 2.066522317377728,
+      "learning_rate": 1.817047206982514e-05,
+      "loss": 0.8191,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7079916078580966,
+      "grad_norm": 2.299633627197154,
+      "learning_rate": 1.8167237963038266e-05,
+      "loss": 0.8003,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7085002225189141,
+      "grad_norm": 3.393158137802057,
+      "learning_rate": 1.8164001288613958e-05,
+      "loss": 0.7479,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7090088371797317,
+      "grad_norm": 2.0878782923220895,
+      "learning_rate": 1.8160762047569767e-05,
+      "loss": 0.774,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7095174518405493,
+      "grad_norm": 2.5886902692931906,
+      "learning_rate": 1.8157520240924063e-05,
+      "loss": 0.7127,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7100260665013669,
+      "grad_norm": 2.2740270105936884,
+      "learning_rate": 1.8154275869696003e-05,
+      "loss": 0.7113,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7105346811621845,
+      "grad_norm": 2.0692045882098395,
+      "learning_rate": 1.815102893490557e-05,
+      "loss": 0.7564,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7110432958230021,
+      "grad_norm": 2.1839251053582878,
+      "learning_rate": 1.814777943757354e-05,
+      "loss": 0.8068,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7115519104838197,
+      "grad_norm": 2.147080204247829,
+      "learning_rate": 1.8144527378721495e-05,
+      "loss": 0.778,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7120605251446372,
+      "grad_norm": 3.159760288852275,
+      "learning_rate": 1.814127275937183e-05,
+      "loss": 0.7734,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7120605251446372,
+      "eval_loss": 0.7360262870788574,
+      "eval_runtime": 170.2227,
+      "eval_samples_per_second": 19.457,
+      "eval_steps_per_second": 9.728,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7125691398054549,
+      "grad_norm": 2.2657548538353,
+      "learning_rate": 1.813801558054774e-05,
+      "loss": 0.7473,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7130777544662725,
+      "grad_norm": 2.0425415759483427,
+      "learning_rate": 1.813475584327323e-05,
+      "loss": 0.7656,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7135863691270901,
+      "grad_norm": 2.2811168475026546,
+      "learning_rate": 1.8131493548573097e-05,
+      "loss": 0.8346,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7140949837879077,
+      "grad_norm": 2.2955390003026306,
+      "learning_rate": 1.812822869747295e-05,
+      "loss": 0.7643,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7146035984487252,
+      "grad_norm": 1.978282677511567,
+      "learning_rate": 1.8124961290999206e-05,
+      "loss": 0.8298,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7151122131095429,
+      "grad_norm": 2.191520375154659,
+      "learning_rate": 1.812169133017908e-05,
+      "loss": 0.8228,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7156208277703605,
+      "grad_norm": 2.124333894336977,
+      "learning_rate": 1.8118418816040583e-05,
+      "loss": 0.7999,
+      "step": 1407
+    },
+    {
+      "epoch": 0.716129442431178,
+      "grad_norm": 2.133209750450364,
+      "learning_rate": 1.8115143749612547e-05,
+      "loss": 0.6951,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7166380570919957,
+      "grad_norm": 2.1046989321663694,
+      "learning_rate": 1.811186613192459e-05,
+      "loss": 0.8575,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7171466717528133,
+      "grad_norm": 2.4092018550105,
+      "learning_rate": 1.8108585964007144e-05,
+      "loss": 0.7914,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7176552864136309,
+      "grad_norm": 2.2900852873799526,
+      "learning_rate": 1.8105303246891428e-05,
+      "loss": 0.876,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7181639010744485,
+      "grad_norm": 2.168182527466751,
+      "learning_rate": 1.8102017981609474e-05,
+      "loss": 0.7431,
+      "step": 1412
+    },
+    {
+      "epoch": 0.718672515735266,
+      "grad_norm": 2.155254299975921,
+      "learning_rate": 1.8098730169194115e-05,
+      "loss": 0.7375,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7191811303960837,
+      "grad_norm": 3.2562798203990075,
+      "learning_rate": 1.8095439810678984e-05,
+      "loss": 0.8245,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7196897450569013,
+      "grad_norm": 4.045443837791666,
+      "learning_rate": 1.8092146907098505e-05,
+      "loss": 0.7497,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7201983597177188,
+      "grad_norm": 2.0206954227332425,
+      "learning_rate": 1.8088851459487915e-05,
+      "loss": 0.7839,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7207069743785365,
+      "grad_norm": 2.1347475321597194,
+      "learning_rate": 1.808555346888324e-05,
+      "loss": 0.7516,
+      "step": 1417
+    },
+    {
+      "epoch": 0.721215589039354,
+      "grad_norm": 2.07900030302529,
+      "learning_rate": 1.808225293632132e-05,
+      "loss": 0.7428,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7217242037001717,
+      "grad_norm": 2.6164176150166534,
+      "learning_rate": 1.807894986283978e-05,
+      "loss": 0.761,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7222328183609893,
+      "grad_norm": 2.233233073342186,
+      "learning_rate": 1.8075644249477045e-05,
+      "loss": 0.7709,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7227414330218068,
+      "grad_norm": 3.305725079062057,
+      "learning_rate": 1.8072336097272347e-05,
+      "loss": 0.77,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7232500476826245,
+      "grad_norm": 1.949093006183371,
+      "learning_rate": 1.806902540726571e-05,
+      "loss": 0.7142,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7237586623434421,
+      "grad_norm": 2.2981111482810217,
+      "learning_rate": 1.8065712180497956e-05,
+      "loss": 0.7751,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7242672770042596,
+      "grad_norm": 2.434003816838277,
+      "learning_rate": 1.806239641801071e-05,
+      "loss": 0.9103,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7247758916650773,
+      "grad_norm": 2.037827976719678,
+      "learning_rate": 1.8059078120846388e-05,
+      "loss": 0.5865,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7252845063258948,
+      "grad_norm": 2.299608640960633,
+      "learning_rate": 1.80557572900482e-05,
+      "loss": 0.7581,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7257931209867124,
+      "grad_norm": 2.2030228394975064,
+      "learning_rate": 1.8052433926660167e-05,
+      "loss": 0.8025,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7263017356475301,
+      "grad_norm": 2.3110247881165322,
+      "learning_rate": 1.8049108031727088e-05,
+      "loss": 0.8022,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7268103503083476,
+      "grad_norm": 2.3269709856808882,
+      "learning_rate": 1.8045779606294568e-05,
+      "loss": 0.7141,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7273189649691653,
+      "grad_norm": 2.2832978345612855,
+      "learning_rate": 1.804244865140901e-05,
+      "loss": 0.7789,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7278275796299828,
+      "grad_norm": 2.773392969767952,
+      "learning_rate": 1.8039115168117608e-05,
+      "loss": 0.9242,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7283361942908004,
+      "grad_norm": 2.3649355812251476,
+      "learning_rate": 1.8035779157468345e-05,
+      "loss": 0.8687,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7288448089516181,
+      "grad_norm": 2.0120718933062527,
+      "learning_rate": 1.803244062051001e-05,
+      "loss": 0.8236,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7293534236124356,
+      "grad_norm": 2.23764183404718,
+      "learning_rate": 1.802909955829218e-05,
+      "loss": 0.7552,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7298620382732532,
+      "grad_norm": 2.0085344612217346,
+      "learning_rate": 1.8025755971865227e-05,
+      "loss": 0.8021,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7303706529340708,
+      "grad_norm": 2.3063661830345503,
+      "learning_rate": 1.8022409862280313e-05,
+      "loss": 0.8459,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7308792675948884,
+      "grad_norm": 2.314492990657455,
+      "learning_rate": 1.8019061230589404e-05,
+      "loss": 0.7049,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7313878822557061,
+      "grad_norm": 2.1979248143146033,
+      "learning_rate": 1.8015710077845243e-05,
+      "loss": 0.7206,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7318964969165236,
+      "grad_norm": 2.200038181320401,
+      "learning_rate": 1.8012356405101378e-05,
+      "loss": 0.8209,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7324051115773412,
+      "grad_norm": 2.151181663655775,
+      "learning_rate": 1.8009000213412146e-05,
+      "loss": 0.7169,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7329137262381589,
+      "grad_norm": 1.9836268178460452,
+      "learning_rate": 1.8005641503832674e-05,
+      "loss": 0.7474,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7334223408989764,
+      "grad_norm": 2.3936098700161788,
+      "learning_rate": 1.8002280277418887e-05,
+      "loss": 0.7846,
+      "step": 1442
+    },
+    {
+      "epoch": 0.733930955559794,
+      "grad_norm": 2.362077188753552,
+      "learning_rate": 1.7998916535227487e-05,
+      "loss": 0.6761,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7344395702206116,
+      "grad_norm": 2.239413788757909,
+      "learning_rate": 1.799555027831598e-05,
+      "loss": 0.7567,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7349481848814292,
+      "grad_norm": 2.3320093712986463,
+      "learning_rate": 1.799218150774266e-05,
+      "loss": 0.7983,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7354567995422469,
+      "grad_norm": 2.1561858679038415,
+      "learning_rate": 1.798881022456661e-05,
+      "loss": 0.8632,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7359654142030644,
+      "grad_norm": 2.0478913790388877,
+      "learning_rate": 1.79854364298477e-05,
+      "loss": 0.6884,
+      "step": 1447
+    },
+    {
+      "epoch": 0.736474028863882,
+      "grad_norm": 1.9571910800502144,
+      "learning_rate": 1.7982060124646593e-05,
+      "loss": 0.8342,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7369826435246996,
+      "grad_norm": 2.446601406114034,
+      "learning_rate": 1.797868131002474e-05,
+      "loss": 0.8057,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7374912581855172,
+      "grad_norm": 2.0649306106916607,
+      "learning_rate": 1.7975299987044388e-05,
+      "loss": 0.7612,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7379998728463348,
+      "grad_norm": 2.494100476257712,
+      "learning_rate": 1.797191615676856e-05,
+      "loss": 0.8215,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7385084875071524,
+      "grad_norm": 2.2366078564131757,
+      "learning_rate": 1.796852982026107e-05,
+      "loss": 0.871,
+      "step": 1452
+    },
+    {
+      "epoch": 0.73901710216797,
+      "grad_norm": 1.975830787046742,
+      "learning_rate": 1.796514097858653e-05,
+      "loss": 0.6804,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7395257168287875,
+      "grad_norm": 2.088518810067873,
+      "learning_rate": 1.796174963281033e-05,
+      "loss": 0.6871,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7400343314896052,
+      "grad_norm": 2.908114911670022,
+      "learning_rate": 1.7958355783998654e-05,
+      "loss": 0.7494,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7405429461504228,
+      "grad_norm": 2.32281051416003,
+      "learning_rate": 1.795495943321846e-05,
+      "loss": 0.8809,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7410515608112404,
+      "grad_norm": 2.6640737762659996,
+      "learning_rate": 1.795156058153751e-05,
+      "loss": 0.7938,
+      "step": 1457
+    },
+    {
+      "epoch": 0.741560175472058,
+      "grad_norm": 2.668096179069293,
+      "learning_rate": 1.794815923002434e-05,
+      "loss": 0.7713,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7420687901328756,
+      "grad_norm": 2.036137262060486,
+      "learning_rate": 1.7944755379748277e-05,
+      "loss": 0.7631,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7425774047936932,
+      "grad_norm": 2.1327370431148944,
+      "learning_rate": 1.7941349031779434e-05,
+      "loss": 0.8776,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7430860194545108,
+      "grad_norm": 2.0782916927168498,
+      "learning_rate": 1.7937940187188704e-05,
+      "loss": 0.8359,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7435946341153283,
+      "grad_norm": 2.245706063688502,
+      "learning_rate": 1.7934528847047766e-05,
+      "loss": 0.8286,
+      "step": 1462
+    },
+    {
+      "epoch": 0.744103248776146,
+      "grad_norm": 2.350645982952433,
+      "learning_rate": 1.7931115012429093e-05,
+      "loss": 0.7943,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7446118634369636,
+      "grad_norm": 2.0934925205706616,
+      "learning_rate": 1.792769868440593e-05,
+      "loss": 0.8539,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7451204780977811,
+      "grad_norm": 2.1948311165944165,
+      "learning_rate": 1.7924279864052316e-05,
+      "loss": 0.7927,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7456290927585988,
+      "grad_norm": 2.797225674430622,
+      "learning_rate": 1.7920858552443062e-05,
+      "loss": 0.8026,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7461377074194163,
+      "grad_norm": 1.7974464502966312,
+      "learning_rate": 1.7917434750653773e-05,
+      "loss": 0.774,
+      "step": 1467
+    },
+    {
+      "epoch": 0.746646322080234,
+      "grad_norm": 2.0327568920367245,
+      "learning_rate": 1.791400845976083e-05,
+      "loss": 0.7681,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7471549367410516,
+      "grad_norm": 2.2717812918918265,
+      "learning_rate": 1.7910579680841407e-05,
+      "loss": 0.7174,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "grad_norm": 2.088976072065711,
+      "learning_rate": 1.7907148414973443e-05,
+      "loss": 0.8288,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7481721660626868,
+      "grad_norm": 1.9050571273888646,
+      "learning_rate": 1.790371466323567e-05,
+      "loss": 0.7795,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7486807807235043,
+      "grad_norm": 1.9491642579314816,
+      "learning_rate": 1.7900278426707604e-05,
+      "loss": 0.753,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7491893953843219,
+      "grad_norm": 2.2587875847784833,
+      "learning_rate": 1.789683970646954e-05,
+      "loss": 0.815,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7496980100451396,
+      "grad_norm": 2.423547179779504,
+      "learning_rate": 1.789339850360254e-05,
+      "loss": 0.806,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7502066247059571,
+      "grad_norm": 2.1541852048842363,
+      "learning_rate": 1.7889954819188468e-05,
+      "loss": 0.7524,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7507152393667748,
+      "grad_norm": 2.5422373480150466,
+      "learning_rate": 1.7886508654309954e-05,
+      "loss": 0.7928,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7512238540275924,
+      "grad_norm": 2.1214012639653785,
+      "learning_rate": 1.7883060010050414e-05,
+      "loss": 0.8186,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7517324686884099,
+      "grad_norm": 2.1109246107451956,
+      "learning_rate": 1.7879608887494047e-05,
+      "loss": 0.8642,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7522410833492276,
+      "grad_norm": 3.9186906387237923,
+      "learning_rate": 1.7876155287725817e-05,
+      "loss": 0.9008,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7527496980100451,
+      "grad_norm": 2.1853922907727785,
+      "learning_rate": 1.787269921183148e-05,
+      "loss": 0.786,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7532583126708627,
+      "grad_norm": 2.2773807058067628,
+      "learning_rate": 1.786924066089757e-05,
+      "loss": 0.8091,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7537669273316804,
+      "grad_norm": 2.0955233679023495,
+      "learning_rate": 1.7865779636011385e-05,
+      "loss": 0.8921,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7542755419924979,
+      "grad_norm": 2.3269547971393187,
+      "learning_rate": 1.786231613826102e-05,
+      "loss": 0.7327,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7547841566533156,
+      "grad_norm": 2.403231027394727,
+      "learning_rate": 1.7858850168735332e-05,
+      "loss": 0.694,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7552927713141331,
+      "grad_norm": 2.1059757676718402,
+      "learning_rate": 1.7855381728523964e-05,
+      "loss": 0.8328,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7558013859749507,
+      "grad_norm": 2.2718091029038114,
+      "learning_rate": 1.785191081871734e-05,
+      "loss": 0.7672,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7563100006357684,
+      "grad_norm": 2.164903057474052,
+      "learning_rate": 1.784843744040664e-05,
+      "loss": 0.907,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7568186152965859,
+      "grad_norm": 2.1943401749166362,
+      "learning_rate": 1.7844961594683844e-05,
+      "loss": 0.8043,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7573272299574035,
+      "grad_norm": 2.4692277083982415,
+      "learning_rate": 1.7841483282641696e-05,
+      "loss": 0.7085,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7578358446182211,
+      "grad_norm": 2.4178442114093603,
+      "learning_rate": 1.7838002505373714e-05,
+      "loss": 0.7425,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7583444592790387,
+      "grad_norm": 2.070586909695956,
+      "learning_rate": 1.7834519263974198e-05,
+      "loss": 0.7453,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7588530739398563,
+      "grad_norm": 2.760108575482078,
+      "learning_rate": 1.7831033559538214e-05,
+      "loss": 0.7612,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7593616886006739,
+      "grad_norm": 2.103190078975455,
+      "learning_rate": 1.7827545393161606e-05,
+      "loss": 0.7314,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7598703032614915,
+      "grad_norm": 2.0977308093127305,
+      "learning_rate": 1.7824054765940993e-05,
+      "loss": 0.6885,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7603789179223092,
+      "grad_norm": 2.256498066376209,
+      "learning_rate": 1.7820561678973774e-05,
+      "loss": 0.7224,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7608875325831267,
+      "grad_norm": 2.4861923384464846,
+      "learning_rate": 1.781706613335811e-05,
+      "loss": 0.7071,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7613961472439443,
+      "grad_norm": 2.293519144108479,
+      "learning_rate": 1.781356813019294e-05,
+      "loss": 0.7844,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 2.043420803610319,
+      "learning_rate": 1.781006767057797e-05,
+      "loss": 0.8708,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7624133765655795,
+      "grad_norm": 2.3902762164826816,
+      "learning_rate": 1.7806564755613696e-05,
+      "loss": 0.7493,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7629219912263971,
+      "grad_norm": 2.010433994146069,
+      "learning_rate": 1.7803059386401364e-05,
+      "loss": 0.7928,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7629219912263971,
+      "eval_loss": 0.7290509343147278,
+      "eval_runtime": 170.2193,
+      "eval_samples_per_second": 19.457,
+      "eval_steps_per_second": 9.729,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7634306058872147,
+      "grad_norm": 2.4241734580684335,
+      "learning_rate": 1.7799551564043e-05,
+      "loss": 0.8075,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7639392205480323,
+      "grad_norm": 1.9907461899018273,
+      "learning_rate": 1.7796041289641408e-05,
+      "loss": 0.7461,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7644478352088498,
+      "grad_norm": 2.031707946800255,
+      "learning_rate": 1.7792528564300156e-05,
+      "loss": 0.7185,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7649564498696675,
+      "grad_norm": 1.7781852993520633,
+      "learning_rate": 1.7789013389123583e-05,
+      "loss": 0.6429,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7654650645304851,
+      "grad_norm": 1.9750074217451656,
+      "learning_rate": 1.7785495765216792e-05,
+      "loss": 0.7475,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7659736791913027,
+      "grad_norm": 2.373839446155845,
+      "learning_rate": 1.7781975693685677e-05,
+      "loss": 0.7752,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7664822938521203,
+      "grad_norm": 2.513602824177846,
+      "learning_rate": 1.7778453175636877e-05,
+      "loss": 0.8313,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7669909085129378,
+      "grad_norm": 2.253673838991013,
+      "learning_rate": 1.777492821217781e-05,
+      "loss": 0.7811,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7674995231737555,
+      "grad_norm": 2.2788784293489535,
+      "learning_rate": 1.7771400804416667e-05,
+      "loss": 0.7622,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7680081378345731,
+      "grad_norm": 2.045722695315856,
+      "learning_rate": 1.7767870953462404e-05,
+      "loss": 0.7648,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7685167524953906,
+      "grad_norm": 2.311584703917161,
+      "learning_rate": 1.776433866042474e-05,
+      "loss": 0.7917,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7690253671562083,
+      "grad_norm": 2.21372632023162,
+      "learning_rate": 1.776080392641417e-05,
+      "loss": 0.7328,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7695339818170259,
+      "grad_norm": 1.9673780474741884,
+      "learning_rate": 1.7757266752541954e-05,
+      "loss": 0.7734,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7700425964778435,
+      "grad_norm": 1.968146700989281,
+      "learning_rate": 1.7753727139920115e-05,
+      "loss": 0.7153,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7705512111386611,
+      "grad_norm": 2.1065935799841573,
+      "learning_rate": 1.775018508966145e-05,
+      "loss": 0.8699,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7710598257994786,
+      "grad_norm": 2.085292073382296,
+      "learning_rate": 1.7746640602879512e-05,
+      "loss": 0.714,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7715684404602963,
+      "grad_norm": 2.575579131841979,
+      "learning_rate": 1.7743093680688626e-05,
+      "loss": 0.7719,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7720770551211139,
+      "grad_norm": 2.1884384322174673,
+      "learning_rate": 1.7739544324203892e-05,
+      "loss": 0.6914,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7725856697819314,
+      "grad_norm": 2.0977733160274994,
+      "learning_rate": 1.7735992534541157e-05,
+      "loss": 0.7265,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7730942844427491,
+      "grad_norm": 2.095749882072474,
+      "learning_rate": 1.773243831281704e-05,
+      "loss": 0.7259,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7736028991035666,
+      "grad_norm": 2.1938462977996296,
+      "learning_rate": 1.772888166014893e-05,
+      "loss": 0.6931,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7741115137643843,
+      "grad_norm": 2.1947070607113894,
+      "learning_rate": 1.772532257765498e-05,
+      "loss": 0.7035,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7746201284252019,
+      "grad_norm": 2.356364268155393,
+      "learning_rate": 1.7721761066454102e-05,
+      "loss": 0.9173,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7751287430860194,
+      "grad_norm": 2.675861797373433,
+      "learning_rate": 1.771819712766597e-05,
+      "loss": 0.8047,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7756373577468371,
+      "grad_norm": 1.885383085791317,
+      "learning_rate": 1.7714630762411025e-05,
+      "loss": 0.7134,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7761459724076547,
+      "grad_norm": 2.227693183647872,
+      "learning_rate": 1.7711061971810476e-05,
+      "loss": 0.7592,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7766545870684722,
+      "grad_norm": 2.2911364433256294,
+      "learning_rate": 1.770749075698628e-05,
+      "loss": 0.7094,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7771632017292899,
+      "grad_norm": 2.7469725850795657,
+      "learning_rate": 1.770391711906117e-05,
+      "loss": 0.8001,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7776718163901074,
+      "grad_norm": 2.1979786169593076,
+      "learning_rate": 1.7700341059158634e-05,
+      "loss": 0.797,
+      "step": 1529
+    },
+    {
+      "epoch": 0.778180431050925,
+      "grad_norm": 2.4083806302751283,
+      "learning_rate": 1.769676257840292e-05,
+      "loss": 0.841,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7786890457117427,
+      "grad_norm": 2.3568979358998607,
+      "learning_rate": 1.769318167791905e-05,
+      "loss": 0.6612,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7791976603725602,
+      "grad_norm": 2.0437442044517167,
+      "learning_rate": 1.768959835883278e-05,
+      "loss": 0.6983,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7797062750333779,
+      "grad_norm": 2.2793095317838925,
+      "learning_rate": 1.7686012622270663e-05,
+      "loss": 0.8951,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7802148896941954,
+      "grad_norm": 2.1585483188659804,
+      "learning_rate": 1.7682424469359975e-05,
+      "loss": 0.7164,
+      "step": 1534
+    },
+    {
+      "epoch": 0.780723504355013,
+      "grad_norm": 2.0806471355311533,
+      "learning_rate": 1.767883390122878e-05,
+      "loss": 0.6958,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7812321190158307,
+      "grad_norm": 2.2142518344168223,
+      "learning_rate": 1.7675240919005885e-05,
+      "loss": 0.8164,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7817407336766482,
+      "grad_norm": 2.4005251071588685,
+      "learning_rate": 1.7671645523820867e-05,
+      "loss": 0.8763,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7822493483374658,
+      "grad_norm": 2.359268186058715,
+      "learning_rate": 1.7668047716804043e-05,
+      "loss": 0.7727,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7827579629982834,
+      "grad_norm": 2.586558932866154,
+      "learning_rate": 1.7664447499086515e-05,
+      "loss": 0.7116,
+      "step": 1539
+    },
+    {
+      "epoch": 0.783266577659101,
+      "grad_norm": 2.089758869479529,
+      "learning_rate": 1.766084487180012e-05,
+      "loss": 0.7879,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7837751923199187,
+      "grad_norm": 1.9883332311284352,
+      "learning_rate": 1.7657239836077465e-05,
+      "loss": 0.8019,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7842838069807362,
+      "grad_norm": 2.124251177625497,
+      "learning_rate": 1.7653632393051908e-05,
+      "loss": 0.7689,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7847924216415538,
+      "grad_norm": 2.3265921647823897,
+      "learning_rate": 1.765002254385757e-05,
+      "loss": 0.7615,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7853010363023715,
+      "grad_norm": 2.562618718072173,
+      "learning_rate": 1.7646410289629318e-05,
+      "loss": 0.7288,
+      "step": 1544
+    },
+    {
+      "epoch": 0.785809650963189,
+      "grad_norm": 2.3072874759495368,
+      "learning_rate": 1.764279563150279e-05,
+      "loss": 0.7651,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7863182656240066,
+      "grad_norm": 2.550722737831481,
+      "learning_rate": 1.763917857061437e-05,
+      "loss": 0.9241,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7868268802848242,
+      "grad_norm": 2.0517341071334054,
+      "learning_rate": 1.7635559108101193e-05,
+      "loss": 0.8358,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7873354949456418,
+      "grad_norm": 2.454620100100101,
+      "learning_rate": 1.7631937245101158e-05,
+      "loss": 0.9347,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7878441096064595,
+      "grad_norm": 2.342169164951011,
+      "learning_rate": 1.762831298275292e-05,
+      "loss": 0.7065,
+      "step": 1549
+    },
+    {
+      "epoch": 0.788352724267277,
+      "grad_norm": 2.3187883810892305,
+      "learning_rate": 1.7624686322195876e-05,
+      "loss": 0.813,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7888613389280946,
+      "grad_norm": 2.001501949922106,
+      "learning_rate": 1.762105726457019e-05,
+      "loss": 0.7029,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7893699535889122,
+      "grad_norm": 2.403120011620869,
+      "learning_rate": 1.7617425811016778e-05,
+      "loss": 0.7463,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7898785682497298,
+      "grad_norm": 2.3364578739142563,
+      "learning_rate": 1.76137919626773e-05,
+      "loss": 0.6998,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7903871829105474,
+      "grad_norm": 2.624138463646993,
+      "learning_rate": 1.7610155720694175e-05,
+      "loss": 0.722,
+      "step": 1554
+    },
+    {
+      "epoch": 0.790895797571365,
+      "grad_norm": 2.0582752307099272,
+      "learning_rate": 1.7606517086210578e-05,
+      "loss": 0.8238,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7914044122321826,
+      "grad_norm": 2.2908074820197646,
+      "learning_rate": 1.7602876060370428e-05,
+      "loss": 0.8543,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7919130268930001,
+      "grad_norm": 2.5556859856183785,
+      "learning_rate": 1.7599232644318403e-05,
+      "loss": 0.8165,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7924216415538178,
+      "grad_norm": 2.4281697822673176,
+      "learning_rate": 1.759558683919993e-05,
+      "loss": 0.7807,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7929302562146354,
+      "grad_norm": 2.1469543886567912,
+      "learning_rate": 1.7591938646161182e-05,
+      "loss": 0.5886,
+      "step": 1559
+    },
+    {
+      "epoch": 0.793438870875453,
+      "grad_norm": 2.004903385374537,
+      "learning_rate": 1.758828806634909e-05,
+      "loss": 0.8383,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7939474855362706,
+      "grad_norm": 2.407117216710252,
+      "learning_rate": 1.7584635100911338e-05,
+      "loss": 0.702,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7944561001970882,
+      "grad_norm": 1.9729779576009456,
+      "learning_rate": 1.7580979750996347e-05,
+      "loss": 0.7447,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7949647148579058,
+      "grad_norm": 2.221035062730422,
+      "learning_rate": 1.7577322017753298e-05,
+      "loss": 0.7798,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7954733295187234,
+      "grad_norm": 1.7913244349366677,
+      "learning_rate": 1.757366190233212e-05,
+      "loss": 0.7791,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7959819441795409,
+      "grad_norm": 1.9863894592838176,
+      "learning_rate": 1.756999940588349e-05,
+      "loss": 0.7125,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7964905588403586,
+      "grad_norm": 2.4665001383405762,
+      "learning_rate": 1.7566334529558833e-05,
+      "loss": 0.7944,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7969991735011762,
+      "grad_norm": 2.7465578557963815,
+      "learning_rate": 1.756266727451032e-05,
+      "loss": 0.7805,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7975077881619937,
+      "grad_norm": 2.0056987872745644,
+      "learning_rate": 1.7558997641890872e-05,
+      "loss": 0.8148,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7980164028228114,
+      "grad_norm": 2.3821133948175186,
+      "learning_rate": 1.7555325632854157e-05,
+      "loss": 0.7823,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7985250174836289,
+      "grad_norm": 2.0928564869560207,
+      "learning_rate": 1.7551651248554598e-05,
+      "loss": 0.8576,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7990336321444466,
+      "grad_norm": 2.0755631567366337,
+      "learning_rate": 1.754797449014735e-05,
+      "loss": 0.7182,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7995422468052642,
+      "grad_norm": 2.408069170639564,
+      "learning_rate": 1.754429535878832e-05,
+      "loss": 0.7462,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8000508614660817,
+      "grad_norm": 2.156388762058814,
+      "learning_rate": 1.7540613855634172e-05,
+      "loss": 0.6813,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8005594761268994,
+      "grad_norm": 2.3420690993907076,
+      "learning_rate": 1.75369299818423e-05,
+      "loss": 0.7698,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8010680907877169,
+      "grad_norm": 2.279758122313076,
+      "learning_rate": 1.753324373857085e-05,
+      "loss": 0.7892,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8015767054485345,
+      "grad_norm": 2.164709442058305,
+      "learning_rate": 1.7529555126978715e-05,
+      "loss": 0.762,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8020853201093522,
+      "grad_norm": 2.0914928738962604,
+      "learning_rate": 1.752586414822553e-05,
+      "loss": 0.7835,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8025939347701697,
+      "grad_norm": 2.102788064389227,
+      "learning_rate": 1.7522170803471673e-05,
+      "loss": 0.8675,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8031025494309874,
+      "grad_norm": 2.421475186725137,
+      "learning_rate": 1.7518475093878264e-05,
+      "loss": 0.8047,
+      "step": 1579
+    },
+    {
+      "epoch": 0.803611164091805,
+      "grad_norm": 2.2784667084542676,
+      "learning_rate": 1.7514777020607177e-05,
+      "loss": 0.6765,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8041197787526225,
+      "grad_norm": 2.2049209092676065,
+      "learning_rate": 1.7511076584821017e-05,
+      "loss": 0.7244,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8046283934134402,
+      "grad_norm": 2.4034831996685284,
+      "learning_rate": 1.750737378768314e-05,
+      "loss": 0.7784,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8051370080742577,
+      "grad_norm": 2.20998402447342,
+      "learning_rate": 1.750366863035764e-05,
+      "loss": 0.7864,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8056456227350753,
+      "grad_norm": 2.3007272607139115,
+      "learning_rate": 1.749996111400935e-05,
+      "loss": 0.8043,
+      "step": 1584
+    },
+    {
+      "epoch": 0.806154237395893,
+      "grad_norm": 2.1504089188382274,
+      "learning_rate": 1.7496251239803853e-05,
+      "loss": 0.7725,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8066628520567105,
+      "grad_norm": 2.412569936420521,
+      "learning_rate": 1.7492539008907465e-05,
+      "loss": 0.7463,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8071714667175282,
+      "grad_norm": 2.318916697306708,
+      "learning_rate": 1.7488824422487254e-05,
+      "loss": 0.8033,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8076800813783457,
+      "grad_norm": 2.0964146070277527,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.6739,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8081886960391633,
+      "grad_norm": 1.8911836917848561,
+      "learning_rate": 1.7481388187747288e-05,
+      "loss": 0.7152,
+      "step": 1589
+    },
+    {
+      "epoch": 0.808697310699981,
+      "grad_norm": 2.30019175517949,
+      "learning_rate": 1.7477666541765357e-05,
+      "loss": 0.7055,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8092059253607985,
+      "grad_norm": 2.3948293379024364,
+      "learning_rate": 1.7473942544935245e-05,
+      "loss": 0.9221,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8097145400216161,
+      "grad_norm": 2.1035830629495647,
+      "learning_rate": 1.747021619842771e-05,
+      "loss": 0.7082,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8102231546824337,
+      "grad_norm": 1.9946110961567167,
+      "learning_rate": 1.746648750341425e-05,
+      "loss": 0.7808,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8107317693432513,
+      "grad_norm": 2.576186219841364,
+      "learning_rate": 1.7462756461067102e-05,
+      "loss": 0.7664,
+      "step": 1594
+    },
+    {
+      "epoch": 0.811240384004069,
+      "grad_norm": 2.0371484652139937,
+      "learning_rate": 1.745902307255924e-05,
+      "loss": 0.8108,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8117489986648865,
+      "grad_norm": 2.0578234279055656,
+      "learning_rate": 1.7455287339064378e-05,
+      "loss": 0.7664,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8122576133257041,
+      "grad_norm": 1.8785301672772592,
+      "learning_rate": 1.7451549261756967e-05,
+      "loss": 0.6599,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8127662279865218,
+      "grad_norm": 2.4197571385110117,
+      "learning_rate": 1.7447808841812186e-05,
+      "loss": 0.7252,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8132748426473393,
+      "grad_norm": 2.0830931970164115,
+      "learning_rate": 1.7444066080405966e-05,
+      "loss": 0.7699,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8137834573081569,
+      "grad_norm": 2.1529420550708456,
+      "learning_rate": 1.744032097871496e-05,
+      "loss": 0.82,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8137834573081569,
+      "eval_loss": 0.7192549109458923,
+      "eval_runtime": 170.3864,
+      "eval_samples_per_second": 19.438,
+      "eval_steps_per_second": 9.719,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8142920719689745,
+      "grad_norm": 2.0958109718580875,
+      "learning_rate": 1.7436573537916572e-05,
+      "loss": 0.7529,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8148006866297921,
+      "grad_norm": 2.3231861545496963,
+      "learning_rate": 1.743282375918892e-05,
+      "loss": 0.7154,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8153093012906097,
+      "grad_norm": 2.3361137604317275,
+      "learning_rate": 1.7429071643710877e-05,
+      "loss": 0.8019,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8158179159514273,
+      "grad_norm": 2.272021294964136,
+      "learning_rate": 1.7425317192662042e-05,
+      "loss": 0.8978,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 2.2494875006327395,
+      "learning_rate": 1.7421560407222748e-05,
+      "loss": 0.8497,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8168351452730624,
+      "grad_norm": 2.1223339859701986,
+      "learning_rate": 1.741780128857406e-05,
+      "loss": 0.8373,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8173437599338801,
+      "grad_norm": 1.8995758536893645,
+      "learning_rate": 1.7414039837897788e-05,
+      "loss": 0.7361,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8178523745946977,
+      "grad_norm": 2.059390874636511,
+      "learning_rate": 1.7410276056376456e-05,
+      "loss": 0.6445,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8183609892555153,
+      "grad_norm": 1.8667640365454998,
+      "learning_rate": 1.7406509945193337e-05,
+      "loss": 0.6861,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8188696039163329,
+      "grad_norm": 2.795255565116794,
+      "learning_rate": 1.7402741505532432e-05,
+      "loss": 0.7932,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8193782185771505,
+      "grad_norm": 2.2154389511785078,
+      "learning_rate": 1.739897073857847e-05,
+      "loss": 0.8046,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8198868332379681,
+      "grad_norm": 2.4394965187043223,
+      "learning_rate": 1.739519764551692e-05,
+      "loss": 0.8521,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8203954478987857,
+      "grad_norm": 1.943919590852951,
+      "learning_rate": 1.7391422227533966e-05,
+      "loss": 0.7095,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8209040625596032,
+      "grad_norm": 2.0294914262491353,
+      "learning_rate": 1.738764448581654e-05,
+      "loss": 0.8001,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8214126772204209,
+      "grad_norm": 2.1489074193113207,
+      "learning_rate": 1.7383864421552302e-05,
+      "loss": 0.805,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8219212918812385,
+      "grad_norm": 2.304046955643442,
+      "learning_rate": 1.7380082035929627e-05,
+      "loss": 0.7972,
+      "step": 1616
+    },
+    {
+      "epoch": 0.822429906542056,
+      "grad_norm": 2.245225503041794,
+      "learning_rate": 1.7376297330137647e-05,
+      "loss": 0.8437,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8229385212028737,
+      "grad_norm": 2.0673223819108406,
+      "learning_rate": 1.7372510305366197e-05,
+      "loss": 0.715,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8234471358636912,
+      "grad_norm": 2.253797505365207,
+      "learning_rate": 1.7368720962805854e-05,
+      "loss": 0.7954,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8239557505245089,
+      "grad_norm": 2.0586756304502933,
+      "learning_rate": 1.736492930364792e-05,
+      "loss": 0.7962,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8244643651853265,
+      "grad_norm": 2.0426411411634415,
+      "learning_rate": 1.7361135329084427e-05,
+      "loss": 0.7822,
+      "step": 1621
+    },
+    {
+      "epoch": 0.824972979846144,
+      "grad_norm": 2.2283183352361067,
+      "learning_rate": 1.735733904030814e-05,
+      "loss": 0.6982,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8254815945069617,
+      "grad_norm": 2.142658576393798,
+      "learning_rate": 1.735354043851254e-05,
+      "loss": 0.8049,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8259902091677792,
+      "grad_norm": 2.2703033552097205,
+      "learning_rate": 1.734973952489184e-05,
+      "loss": 0.6986,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8264988238285969,
+      "grad_norm": 2.0828830486692222,
+      "learning_rate": 1.7345936300640986e-05,
+      "loss": 0.8181,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8270074384894145,
+      "grad_norm": 2.1995493774144017,
+      "learning_rate": 1.7342130766955645e-05,
+      "loss": 0.7848,
+      "step": 1626
+    },
+    {
+      "epoch": 0.827516053150232,
+      "grad_norm": 2.813888461393387,
+      "learning_rate": 1.7338322925032206e-05,
+      "loss": 0.8731,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8280246678110497,
+      "grad_norm": 2.412323853382085,
+      "learning_rate": 1.7334512776067796e-05,
+      "loss": 0.8716,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8285332824718673,
+      "grad_norm": 2.1315339556500983,
+      "learning_rate": 1.733070032126025e-05,
+      "loss": 0.722,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8290418971326848,
+      "grad_norm": 1.9237483748738327,
+      "learning_rate": 1.7326885561808145e-05,
+      "loss": 0.8672,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8295505117935025,
+      "grad_norm": 2.188609250704773,
+      "learning_rate": 1.732306849891077e-05,
+      "loss": 0.9421,
+      "step": 1631
+    },
+    {
+      "epoch": 0.83005912645432,
+      "grad_norm": 3.2732023823863337,
+      "learning_rate": 1.7319249133768145e-05,
+      "loss": 0.8453,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8305677411151376,
+      "grad_norm": 2.53759385109213,
+      "learning_rate": 1.7315427467581016e-05,
+      "loss": 0.8515,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8310763557759553,
+      "grad_norm": 1.9263656379502412,
+      "learning_rate": 1.7311603501550838e-05,
+      "loss": 0.8137,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8315849704367728,
+      "grad_norm": 2.0045288784338076,
+      "learning_rate": 1.730777723687981e-05,
+      "loss": 0.6463,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8320935850975905,
+      "grad_norm": 2.0955914043642694,
+      "learning_rate": 1.730394867477084e-05,
+      "loss": 0.7292,
+      "step": 1636
+    },
+    {
+      "epoch": 0.832602199758408,
+      "grad_norm": 2.0996884069428017,
+      "learning_rate": 1.7300117816427557e-05,
+      "loss": 0.6621,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8331108144192256,
+      "grad_norm": 2.4025235736813753,
+      "learning_rate": 1.729628466305432e-05,
+      "loss": 0.7322,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8336194290800433,
+      "grad_norm": 2.1945447563497242,
+      "learning_rate": 1.7292449215856207e-05,
+      "loss": 0.7799,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8341280437408608,
+      "grad_norm": 2.3742587287953656,
+      "learning_rate": 1.728861147603901e-05,
+      "loss": 0.795,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8346366584016784,
+      "grad_norm": 1.8841471393534435,
+      "learning_rate": 1.7284771444809257e-05,
+      "loss": 0.7257,
+      "step": 1641
+    },
+    {
+      "epoch": 0.835145273062496,
+      "grad_norm": 2.19657725762427,
+      "learning_rate": 1.728092912337418e-05,
+      "loss": 0.7013,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8356538877233136,
+      "grad_norm": 2.464588569212055,
+      "learning_rate": 1.7277084512941734e-05,
+      "loss": 0.7767,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8361625023841313,
+      "grad_norm": 2.0625544454150813,
+      "learning_rate": 1.727323761472061e-05,
+      "loss": 0.6508,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8366711170449488,
+      "grad_norm": 2.0227729560361296,
+      "learning_rate": 1.7269388429920193e-05,
+      "loss": 0.7563,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8371797317057664,
+      "grad_norm": 2.09959824670016,
+      "learning_rate": 1.7265536959750608e-05,
+      "loss": 0.7618,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8376883463665841,
+      "grad_norm": 2.610034425296515,
+      "learning_rate": 1.726168320542269e-05,
+      "loss": 0.9379,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8381969610274016,
+      "grad_norm": 1.9956821084385896,
+      "learning_rate": 1.7257827168147986e-05,
+      "loss": 0.7337,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8387055756882192,
+      "grad_norm": 2.2607433475631473,
+      "learning_rate": 1.725396884913877e-05,
+      "loss": 0.6676,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8392141903490368,
+      "grad_norm": 2.0812366107131957,
+      "learning_rate": 1.7250108249608034e-05,
+      "loss": 0.6852,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8397228050098544,
+      "grad_norm": 1.9620167471339116,
+      "learning_rate": 1.724624537076948e-05,
+      "loss": 0.6882,
+      "step": 1651
+    },
+    {
+      "epoch": 0.840231419670672,
+      "grad_norm": 2.0560342414617,
+      "learning_rate": 1.7242380213837528e-05,
+      "loss": 0.7637,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8407400343314896,
+      "grad_norm": 1.7892382006595267,
+      "learning_rate": 1.7238512780027322e-05,
+      "loss": 0.7159,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8412486489923072,
+      "grad_norm": 2.427019721467783,
+      "learning_rate": 1.7234643070554707e-05,
+      "loss": 0.7346,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8417572636531248,
+      "grad_norm": 2.2857165570758706,
+      "learning_rate": 1.7230771086636263e-05,
+      "loss": 0.8155,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8422658783139424,
+      "grad_norm": 2.8245677178399484,
+      "learning_rate": 1.7226896829489267e-05,
+      "loss": 0.7895,
+      "step": 1656
+    },
+    {
+      "epoch": 0.84277449297476,
+      "grad_norm": 3.581464649844521,
+      "learning_rate": 1.7223020300331722e-05,
+      "loss": 0.8587,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8432831076355776,
+      "grad_norm": 2.317888322906395,
+      "learning_rate": 1.7219141500382333e-05,
+      "loss": 0.8299,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8437917222963952,
+      "grad_norm": 2.2916684468580324,
+      "learning_rate": 1.721526043086054e-05,
+      "loss": 0.8156,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8443003369572127,
+      "grad_norm": 2.648437770067049,
+      "learning_rate": 1.7211377092986477e-05,
+      "loss": 0.7303,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8448089516180304,
+      "grad_norm": 2.310629216820499,
+      "learning_rate": 1.7207491487980993e-05,
+      "loss": 0.7811,
+      "step": 1661
+    },
+    {
+      "epoch": 0.845317566278848,
+      "grad_norm": 2.287068719282652,
+      "learning_rate": 1.720360361706566e-05,
+      "loss": 0.8651,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8458261809396656,
+      "grad_norm": 2.0993863435321236,
+      "learning_rate": 1.719971348146276e-05,
+      "loss": 0.6777,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8463347956004832,
+      "grad_norm": 2.191472615207766,
+      "learning_rate": 1.7195821082395275e-05,
+      "loss": 0.7704,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8468434102613008,
+      "grad_norm": 1.6164393514246387,
+      "learning_rate": 1.719192642108692e-05,
+      "loss": 0.7596,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8473520249221184,
+      "grad_norm": 2.1626699215495635,
+      "learning_rate": 1.7188029498762095e-05,
+      "loss": 0.7299,
+      "step": 1666
+    },
+    {
+      "epoch": 0.847860639582936,
+      "grad_norm": 3.0949946510379647,
+      "learning_rate": 1.7184130316645936e-05,
+      "loss": 0.846,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8483692542437535,
+      "grad_norm": 1.833198419577483,
+      "learning_rate": 1.718022887596427e-05,
+      "loss": 0.6712,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8488778689045712,
+      "grad_norm": 2.332664314361866,
+      "learning_rate": 1.7176325177943643e-05,
+      "loss": 0.6934,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8493864835653888,
+      "grad_norm": 2.0997113438175155,
+      "learning_rate": 1.7172419223811316e-05,
+      "loss": 0.791,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8498950982262063,
+      "grad_norm": 1.9223689243194098,
+      "learning_rate": 1.716851101479525e-05,
+      "loss": 0.7209,
+      "step": 1671
+    },
+    {
+      "epoch": 0.850403712887024,
+      "grad_norm": 2.331100712718546,
+      "learning_rate": 1.7164600552124113e-05,
+      "loss": 0.6994,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8509123275478415,
+      "grad_norm": 2.103559148226617,
+      "learning_rate": 1.716068783702729e-05,
+      "loss": 0.6855,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8514209422086592,
+      "grad_norm": 2.4857606682263134,
+      "learning_rate": 1.715677287073487e-05,
+      "loss": 0.7538,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8519295568694768,
+      "grad_norm": 2.185048064307731,
+      "learning_rate": 1.7152855654477652e-05,
+      "loss": 0.721,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8524381715302943,
+      "grad_norm": 2.4238696344526818,
+      "learning_rate": 1.7148936189487138e-05,
+      "loss": 0.7938,
+      "step": 1676
+    },
+    {
+      "epoch": 0.852946786191112,
+      "grad_norm": 2.160864053555146,
+      "learning_rate": 1.7145014476995543e-05,
+      "loss": 0.8043,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8534554008519295,
+      "grad_norm": 2.3263864874629694,
+      "learning_rate": 1.714109051823578e-05,
+      "loss": 0.7188,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8539640155127471,
+      "grad_norm": 2.4520447897873003,
+      "learning_rate": 1.713716431444148e-05,
+      "loss": 0.8212,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8544726301735648,
+      "grad_norm": 2.2362449663850823,
+      "learning_rate": 1.713323586684697e-05,
+      "loss": 0.7342,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8549812448343823,
+      "grad_norm": 1.9436328177991915,
+      "learning_rate": 1.712930517668728e-05,
+      "loss": 0.7635,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8554898594952,
+      "grad_norm": 2.0714447955731714,
+      "learning_rate": 1.712537224519816e-05,
+      "loss": 0.672,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8559984741560176,
+      "grad_norm": 2.2659093218013964,
+      "learning_rate": 1.7121437073616046e-05,
+      "loss": 0.7839,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8565070888168351,
+      "grad_norm": 1.9849447799035513,
+      "learning_rate": 1.7117499663178097e-05,
+      "loss": 0.728,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8570157034776528,
+      "grad_norm": 5.313204729796549,
+      "learning_rate": 1.7113560015122163e-05,
+      "loss": 0.8311,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8575243181384703,
+      "grad_norm": 2.1570143519716822,
+      "learning_rate": 1.7109618130686795e-05,
+      "loss": 0.8521,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8580329327992879,
+      "grad_norm": 2.076198631537666,
+      "learning_rate": 1.7105674011111256e-05,
+      "loss": 0.7297,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8585415474601056,
+      "grad_norm": 2.0209473122246018,
+      "learning_rate": 1.7101727657635513e-05,
+      "loss": 0.7721,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8590501621209231,
+      "grad_norm": 2.121025070286865,
+      "learning_rate": 1.7097779071500226e-05,
+      "loss": 0.8068,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8595587767817408,
+      "grad_norm": 2.5813210761604184,
+      "learning_rate": 1.7093828253946765e-05,
+      "loss": 0.7914,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8600673914425583,
+      "grad_norm": 2.0646919393859777,
+      "learning_rate": 1.7089875206217193e-05,
+      "loss": 0.6579,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8605760061033759,
+      "grad_norm": 2.41330516997215,
+      "learning_rate": 1.7085919929554286e-05,
+      "loss": 0.9031,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8610846207641936,
+      "grad_norm": 2.3427514555815336,
+      "learning_rate": 1.708196242520151e-05,
+      "loss": 0.8541,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8615932354250111,
+      "grad_norm": 2.266503209877392,
+      "learning_rate": 1.707800269440303e-05,
+      "loss": 0.737,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8621018500858287,
+      "grad_norm": 2.087836954481488,
+      "learning_rate": 1.7074040738403735e-05,
+      "loss": 0.7846,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8626104647466464,
+      "grad_norm": 2.029558504511778,
+      "learning_rate": 1.7070076558449177e-05,
+      "loss": 0.7404,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8631190794074639,
+      "grad_norm": 2.043603277043485,
+      "learning_rate": 1.706611015578563e-05,
+      "loss": 0.8123,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8636276940682815,
+      "grad_norm": 2.3330726023731745,
+      "learning_rate": 1.7062141531660063e-05,
+      "loss": 0.7294,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8641363087290991,
+      "grad_norm": 1.940762879515268,
+      "learning_rate": 1.7058170687320145e-05,
+      "loss": 0.7549,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8646449233899167,
+      "grad_norm": 1.9901486603881997,
+      "learning_rate": 1.7054197624014236e-05,
+      "loss": 0.8099,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8646449233899167,
+      "eval_loss": 0.7182188034057617,
+      "eval_runtime": 169.9471,
+      "eval_samples_per_second": 19.488,
+      "eval_steps_per_second": 9.744,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8651535380507344,
+      "grad_norm": 2.074573436232561,
+      "learning_rate": 1.7050222342991398e-05,
+      "loss": 0.7028,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8656621527115519,
+      "grad_norm": 2.075280558910857,
+      "learning_rate": 1.7046244845501393e-05,
+      "loss": 0.6742,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8661707673723695,
+      "grad_norm": 2.0750415843795267,
+      "learning_rate": 1.7042265132794677e-05,
+      "loss": 0.8077,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8666793820331871,
+      "grad_norm": 1.9353363507718488,
+      "learning_rate": 1.7038283206122397e-05,
+      "loss": 0.6254,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8671879966940047,
+      "grad_norm": 2.0801067798055626,
+      "learning_rate": 1.703429906673641e-05,
+      "loss": 0.7594,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8676966113548223,
+      "grad_norm": 3.2526035151036843,
+      "learning_rate": 1.7030312715889253e-05,
+      "loss": 0.6858,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8682052260156399,
+      "grad_norm": 2.108339634862034,
+      "learning_rate": 1.702632415483417e-05,
+      "loss": 0.7913,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8687138406764575,
+      "grad_norm": 1.933310624789464,
+      "learning_rate": 1.702233338482509e-05,
+      "loss": 0.7768,
+      "step": 1708
+    },
+    {
+      "epoch": 0.869222455337275,
+      "grad_norm": 2.4469330021985756,
+      "learning_rate": 1.7018340407116647e-05,
+      "loss": 0.8466,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8697310699980927,
+      "grad_norm": 2.553967949094832,
+      "learning_rate": 1.701434522296416e-05,
+      "loss": 0.9164,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8702396846589103,
+      "grad_norm": 1.9766567645004496,
+      "learning_rate": 1.7010347833623647e-05,
+      "loss": 0.8348,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8707482993197279,
+      "grad_norm": 2.0647440175779126,
+      "learning_rate": 1.700634824035182e-05,
+      "loss": 0.6338,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8712569139805455,
+      "grad_norm": 1.94910964201772,
+      "learning_rate": 1.7002346444406075e-05,
+      "loss": 0.8099,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8717655286413631,
+      "grad_norm": 1.9953059304651235,
+      "learning_rate": 1.6998342447044516e-05,
+      "loss": 0.8208,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8722741433021807,
+      "grad_norm": 2.0428361722039394,
+      "learning_rate": 1.699433624952592e-05,
+      "loss": 0.7708,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8727827579629983,
+      "grad_norm": 1.8562748172012968,
+      "learning_rate": 1.699032785310977e-05,
+      "loss": 0.7837,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8732913726238158,
+      "grad_norm": 2.1159581067241433,
+      "learning_rate": 1.6986317259056238e-05,
+      "loss": 0.7117,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8737999872846335,
+      "grad_norm": 2.0766723833367706,
+      "learning_rate": 1.6982304468626188e-05,
+      "loss": 0.7287,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8743086019454511,
+      "grad_norm": 2.321334973219472,
+      "learning_rate": 1.697828948308116e-05,
+      "loss": 0.8238,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8748172166062687,
+      "grad_norm": 1.9303224576866462,
+      "learning_rate": 1.6974272303683407e-05,
+      "loss": 0.7314,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8753258312670863,
+      "grad_norm": 1.9245605933307826,
+      "learning_rate": 1.6970252931695857e-05,
+      "loss": 0.6292,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8758344459279038,
+      "grad_norm": 2.15221015529415,
+      "learning_rate": 1.6966231368382125e-05,
+      "loss": 0.8518,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8763430605887215,
+      "grad_norm": 2.1029673138888563,
+      "learning_rate": 1.6962207615006527e-05,
+      "loss": 0.6702,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8768516752495391,
+      "grad_norm": 2.2227672689966447,
+      "learning_rate": 1.695818167283406e-05,
+      "loss": 0.8594,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8773602899103566,
+      "grad_norm": 2.080353424050231,
+      "learning_rate": 1.6954153543130407e-05,
+      "loss": 0.6479,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8778689045711743,
+      "grad_norm": 2.2573841573030005,
+      "learning_rate": 1.6950123227161946e-05,
+      "loss": 0.6887,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8783775192319918,
+      "grad_norm": 2.2771059831456646,
+      "learning_rate": 1.6946090726195735e-05,
+      "loss": 0.6929,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8788861338928095,
+      "grad_norm": 1.987874707101847,
+      "learning_rate": 1.6942056041499523e-05,
+      "loss": 0.8426,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8793947485536271,
+      "grad_norm": 2.1120678702344207,
+      "learning_rate": 1.6938019174341747e-05,
+      "loss": 0.8333,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8799033632144446,
+      "grad_norm": 1.8873634263705794,
+      "learning_rate": 1.6933980125991522e-05,
+      "loss": 0.7279,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8804119778752623,
+      "grad_norm": 1.735585245811134,
+      "learning_rate": 1.692993889771866e-05,
+      "loss": 0.7141,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8809205925360799,
+      "grad_norm": 1.99532439154395,
+      "learning_rate": 1.692589549079365e-05,
+      "loss": 0.7589,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8814292071968974,
+      "grad_norm": 1.999125289372185,
+      "learning_rate": 1.6921849906487672e-05,
+      "loss": 0.7981,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8819378218577151,
+      "grad_norm": 2.38466104813958,
+      "learning_rate": 1.6917802146072585e-05,
+      "loss": 0.7337,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8824464365185326,
+      "grad_norm": 2.2457485298560447,
+      "learning_rate": 1.6913752210820933e-05,
+      "loss": 0.7789,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8829550511793502,
+      "grad_norm": 2.4250215824385912,
+      "learning_rate": 1.6909700102005947e-05,
+      "loss": 0.7266,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8834636658401679,
+      "grad_norm": 2.059091815377958,
+      "learning_rate": 1.6905645820901536e-05,
+      "loss": 0.6617,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8839722805009854,
+      "grad_norm": 2.12350501036676,
+      "learning_rate": 1.6901589368782303e-05,
+      "loss": 0.8163,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8844808951618031,
+      "grad_norm": 2.293697001081393,
+      "learning_rate": 1.689753074692352e-05,
+      "loss": 0.6627,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8849895098226206,
+      "grad_norm": 2.116364266228038,
+      "learning_rate": 1.6893469956601144e-05,
+      "loss": 0.7703,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8854981244834382,
+      "grad_norm": 1.9142946686362423,
+      "learning_rate": 1.6889406999091825e-05,
+      "loss": 0.6459,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8860067391442559,
+      "grad_norm": 1.9204876862488462,
+      "learning_rate": 1.6885341875672884e-05,
+      "loss": 0.702,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8865153538050734,
+      "grad_norm": 2.1953715794986306,
+      "learning_rate": 1.688127458762232e-05,
+      "loss": 0.7534,
+      "step": 1743
+    },
+    {
+      "epoch": 0.887023968465891,
+      "grad_norm": 2.1796120251129762,
+      "learning_rate": 1.687720513621882e-05,
+      "loss": 0.8958,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8875325831267086,
+      "grad_norm": 2.2720730282034296,
+      "learning_rate": 1.6873133522741755e-05,
+      "loss": 0.7897,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8880411977875262,
+      "grad_norm": 2.0918424656563044,
+      "learning_rate": 1.686905974847116e-05,
+      "loss": 0.7849,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8885498124483439,
+      "grad_norm": 2.1669143999741167,
+      "learning_rate": 1.686498381468776e-05,
+      "loss": 0.7976,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8890584271091614,
+      "grad_norm": 2.297422278981425,
+      "learning_rate": 1.6860905722672966e-05,
+      "loss": 0.7324,
+      "step": 1748
+    },
+    {
+      "epoch": 0.889567041769979,
+      "grad_norm": 2.0110318390741044,
+      "learning_rate": 1.685682547370885e-05,
+      "loss": 0.681,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8900756564307967,
+      "grad_norm": 1.8785179196869433,
+      "learning_rate": 1.6852743069078176e-05,
+      "loss": 0.7745,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8905842710916142,
+      "grad_norm": 1.7044089141258074,
+      "learning_rate": 1.6848658510064378e-05,
+      "loss": 0.7002,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8910928857524318,
+      "grad_norm": 1.9740525544426157,
+      "learning_rate": 1.684457179795157e-05,
+      "loss": 0.6881,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8916015004132494,
+      "grad_norm": 2.2776146755387514,
+      "learning_rate": 1.6840482934024552e-05,
+      "loss": 0.9007,
+      "step": 1753
+    },
+    {
+      "epoch": 0.892110115074067,
+      "grad_norm": 2.1660574521153078,
+      "learning_rate": 1.683639191956878e-05,
+      "loss": 0.8321,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8926187297348847,
+      "grad_norm": 2.048497612461806,
+      "learning_rate": 1.68322987558704e-05,
+      "loss": 0.854,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8931273443957022,
+      "grad_norm": 3.110951393054268,
+      "learning_rate": 1.682820344421624e-05,
+      "loss": 0.8105,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8936359590565198,
+      "grad_norm": 1.8564605953063074,
+      "learning_rate": 1.6824105985893778e-05,
+      "loss": 0.7814,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8941445737173374,
+      "grad_norm": 2.0326634111204234,
+      "learning_rate": 1.6820006382191205e-05,
+      "loss": 0.7661,
+      "step": 1758
+    },
+    {
+      "epoch": 0.894653188378155,
+      "grad_norm": 2.1397912845259017,
+      "learning_rate": 1.6815904634397347e-05,
+      "loss": 0.6849,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8951618030389726,
+      "grad_norm": 2.1113746489728427,
+      "learning_rate": 1.6811800743801732e-05,
+      "loss": 0.8124,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8956704176997902,
+      "grad_norm": 2.256553536494897,
+      "learning_rate": 1.680769471169455e-05,
+      "loss": 0.6765,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8961790323606078,
+      "grad_norm": 1.989720211679826,
+      "learning_rate": 1.6803586539366662e-05,
+      "loss": 0.7438,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8966876470214253,
+      "grad_norm": 2.172065122776711,
+      "learning_rate": 1.679947622810961e-05,
+      "loss": 0.6734,
+      "step": 1763
+    },
+    {
+      "epoch": 0.897196261682243,
+      "grad_norm": 1.9978656108592376,
+      "learning_rate": 1.67953637792156e-05,
+      "loss": 0.6302,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8977048763430606,
+      "grad_norm": 2.2890022972965336,
+      "learning_rate": 1.679124919397752e-05,
+      "loss": 0.7177,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8982134910038782,
+      "grad_norm": 1.997872293700179,
+      "learning_rate": 1.6787132473688923e-05,
+      "loss": 0.6412,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8987221056646958,
+      "grad_norm": 2.188511423891235,
+      "learning_rate": 1.6783013619644033e-05,
+      "loss": 0.7276,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8992307203255134,
+      "grad_norm": 2.2219080411490375,
+      "learning_rate": 1.6778892633137746e-05,
+      "loss": 0.9048,
+      "step": 1768
+    },
+    {
+      "epoch": 0.899739334986331,
+      "grad_norm": 2.088882253627294,
+      "learning_rate": 1.6774769515465626e-05,
+      "loss": 0.8288,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9002479496471486,
+      "grad_norm": 2.173908297284792,
+      "learning_rate": 1.677064426792391e-05,
+      "loss": 0.7914,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9007565643079661,
+      "grad_norm": 1.925013576806481,
+      "learning_rate": 1.6766516891809508e-05,
+      "loss": 0.7259,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9012651789687838,
+      "grad_norm": 3.799711949322889,
+      "learning_rate": 1.676238738841999e-05,
+      "loss": 0.7831,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9017737936296014,
+      "grad_norm": 1.8729521694644427,
+      "learning_rate": 1.6758255759053602e-05,
+      "loss": 0.7278,
+      "step": 1773
+    },
+    {
+      "epoch": 0.902282408290419,
+      "grad_norm": 2.1181631620272405,
+      "learning_rate": 1.6754122005009258e-05,
+      "loss": 0.8649,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9027910229512366,
+      "grad_norm": 2.5194209111619474,
+      "learning_rate": 1.674998612758654e-05,
+      "loss": 0.7875,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9032996376120541,
+      "grad_norm": 2.13831934701573,
+      "learning_rate": 1.674584812808569e-05,
+      "loss": 0.8281,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9038082522728718,
+      "grad_norm": 1.9361770789524368,
+      "learning_rate": 1.6741708007807626e-05,
+      "loss": 0.7465,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9043168669336894,
+      "grad_norm": 2.105919954153529,
+      "learning_rate": 1.6737565768053933e-05,
+      "loss": 0.7579,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9048254815945069,
+      "grad_norm": 2.25562166885464,
+      "learning_rate": 1.673342141012685e-05,
+      "loss": 0.7429,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9053340962553246,
+      "grad_norm": 2.284092947563459,
+      "learning_rate": 1.6729274935329303e-05,
+      "loss": 0.737,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9058427109161422,
+      "grad_norm": 1.9948285716123202,
+      "learning_rate": 1.6725126344964865e-05,
+      "loss": 0.7725,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9063513255769597,
+      "grad_norm": 1.9213047577221836,
+      "learning_rate": 1.672097564033778e-05,
+      "loss": 0.8226,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9068599402377774,
+      "grad_norm": 2.0957220099772362,
+      "learning_rate": 1.6716822822752954e-05,
+      "loss": 0.7136,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9073685548985949,
+      "grad_norm": 1.8304725637164176,
+      "learning_rate": 1.6712667893515967e-05,
+      "loss": 0.7019,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9078771695594126,
+      "grad_norm": 1.8817143700395147,
+      "learning_rate": 1.670851085393306e-05,
+      "loss": 0.7561,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9083857842202302,
+      "grad_norm": 2.136774316933435,
+      "learning_rate": 1.670435170531112e-05,
+      "loss": 0.7766,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9088943988810477,
+      "grad_norm": 2.289559892899469,
+      "learning_rate": 1.6700190448957724e-05,
+      "loss": 0.7888,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9094030135418654,
+      "grad_norm": 2.181953373607504,
+      "learning_rate": 1.669602708618109e-05,
+      "loss": 0.7818,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9099116282026829,
+      "grad_norm": 1.9110729303661393,
+      "learning_rate": 1.6691861618290113e-05,
+      "loss": 0.7616,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9104202428635005,
+      "grad_norm": 1.9302993607471433,
+      "learning_rate": 1.668769404659434e-05,
+      "loss": 0.7547,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9109288575243182,
+      "grad_norm": 2.2591780918366675,
+      "learning_rate": 1.6683524372403984e-05,
+      "loss": 0.883,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9114374721851357,
+      "grad_norm": 1.899259663952816,
+      "learning_rate": 1.6679352597029916e-05,
+      "loss": 0.7149,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9119460868459534,
+      "grad_norm": 2.097418233484868,
+      "learning_rate": 1.6675178721783675e-05,
+      "loss": 0.7896,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9124547015067709,
+      "grad_norm": 1.9319390509016907,
+      "learning_rate": 1.6671002747977447e-05,
+      "loss": 0.7403,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9129633161675885,
+      "grad_norm": 2.394264393011469,
+      "learning_rate": 1.6666824676924093e-05,
+      "loss": 0.8005,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9134719308284062,
+      "grad_norm": 1.8500711736648763,
+      "learning_rate": 1.6662644509937123e-05,
+      "loss": 0.7855,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9139805454892237,
+      "grad_norm": 2.131471093307729,
+      "learning_rate": 1.6658462248330706e-05,
+      "loss": 0.739,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9144891601500413,
+      "grad_norm": 2.025893205108161,
+      "learning_rate": 1.6654277893419678e-05,
+      "loss": 0.784,
+      "step": 1798
+    },
+    {
+      "epoch": 0.914997774810859,
+      "grad_norm": 2.3023866709632843,
+      "learning_rate": 1.6650091446519517e-05,
+      "loss": 0.859,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9155063894716765,
+      "grad_norm": 2.1048439245285517,
+      "learning_rate": 1.664590290894638e-05,
+      "loss": 0.8449,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9155063894716765,
+      "eval_loss": 0.708531379699707,
+      "eval_runtime": 170.1374,
+      "eval_samples_per_second": 19.467,
+      "eval_steps_per_second": 9.733,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9160150041324941,
+      "grad_norm": 1.9883343103070137,
+      "learning_rate": 1.6641712282017065e-05,
+      "loss": 0.7951,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9165236187933117,
+      "grad_norm": 1.9858732799888275,
+      "learning_rate": 1.6637519567049035e-05,
+      "loss": 0.7422,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9170322334541293,
+      "grad_norm": 1.94820682044183,
+      "learning_rate": 1.6633324765360405e-05,
+      "loss": 0.7261,
+      "step": 1803
+    },
+    {
+      "epoch": 0.917540848114947,
+      "grad_norm": 2.2394556290050103,
+      "learning_rate": 1.6629127878269944e-05,
+      "loss": 0.8143,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9180494627757645,
+      "grad_norm": 1.9382445227738594,
+      "learning_rate": 1.6624928907097088e-05,
+      "loss": 0.7503,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9185580774365821,
+      "grad_norm": 2.09107048180775,
+      "learning_rate": 1.662072785316191e-05,
+      "loss": 0.8043,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9190666920973997,
+      "grad_norm": 1.9893753249263708,
+      "learning_rate": 1.661652471778516e-05,
+      "loss": 0.6812,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9195753067582173,
+      "grad_norm": 2.26670874622031,
+      "learning_rate": 1.6612319502288222e-05,
+      "loss": 0.6905,
+      "step": 1808
+    },
+    {
+      "epoch": 0.920083921419035,
+      "grad_norm": 2.703145109774366,
+      "learning_rate": 1.660811220799315e-05,
+      "loss": 0.8272,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9205925360798525,
+      "grad_norm": 2.0259791138185275,
+      "learning_rate": 1.6603902836222633e-05,
+      "loss": 0.7043,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9211011507406701,
+      "grad_norm": 2.0700561364712047,
+      "learning_rate": 1.6599691388300034e-05,
+      "loss": 0.7158,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9216097654014876,
+      "grad_norm": 1.997080877970982,
+      "learning_rate": 1.6595477865549353e-05,
+      "loss": 0.8153,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9221183800623053,
+      "grad_norm": 2.1818231217733004,
+      "learning_rate": 1.6591262269295255e-05,
+      "loss": 0.8129,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9226269947231229,
+      "grad_norm": 2.151655526736867,
+      "learning_rate": 1.6587044600863044e-05,
+      "loss": 0.7154,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9231356093839405,
+      "grad_norm": 1.806556295540302,
+      "learning_rate": 1.6582824861578687e-05,
+      "loss": 0.7351,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9236442240447581,
+      "grad_norm": 6.863992097329879,
+      "learning_rate": 1.6578603052768786e-05,
+      "loss": 0.9036,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9241528387055757,
+      "grad_norm": 1.753395736923074,
+      "learning_rate": 1.6574379175760618e-05,
+      "loss": 0.702,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9246614533663933,
+      "grad_norm": 2.050916336696696,
+      "learning_rate": 1.6570153231882084e-05,
+      "loss": 0.759,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9251700680272109,
+      "grad_norm": 2.0752798696008234,
+      "learning_rate": 1.6565925222461757e-05,
+      "loss": 0.7535,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9256786826880284,
+      "grad_norm": 2.132360277685579,
+      "learning_rate": 1.6561695148828843e-05,
+      "loss": 0.7876,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9261872973488461,
+      "grad_norm": 2.773689881782521,
+      "learning_rate": 1.6557463012313207e-05,
+      "loss": 0.758,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9266959120096637,
+      "grad_norm": 2.0336752864569725,
+      "learning_rate": 1.6553228814245358e-05,
+      "loss": 0.8122,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9272045266704813,
+      "grad_norm": 2.094689229708115,
+      "learning_rate": 1.654899255595645e-05,
+      "loss": 0.6652,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9277131413312989,
+      "grad_norm": 2.4823909490476264,
+      "learning_rate": 1.65447542387783e-05,
+      "loss": 0.8237,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9282217559921164,
+      "grad_norm": 2.2973915609145465,
+      "learning_rate": 1.654051386404335e-05,
+      "loss": 0.6993,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9287303706529341,
+      "grad_norm": 2.5754130578449095,
+      "learning_rate": 1.6536271433084704e-05,
+      "loss": 0.7855,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9292389853137517,
+      "grad_norm": 1.9420259588060387,
+      "learning_rate": 1.653202694723611e-05,
+      "loss": 0.8171,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9297475999745692,
+      "grad_norm": 3.063607151992285,
+      "learning_rate": 1.6527780407831958e-05,
+      "loss": 0.7538,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9302562146353869,
+      "grad_norm": 2.147786322658535,
+      "learning_rate": 1.6523531816207284e-05,
+      "loss": 0.7064,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9307648292962044,
+      "grad_norm": 2.0051085079825928,
+      "learning_rate": 1.6519281173697778e-05,
+      "loss": 0.6624,
+      "step": 1830
+    },
+    {
+      "epoch": 0.931273443957022,
+      "grad_norm": 1.9959682119051372,
+      "learning_rate": 1.6515028481639763e-05,
+      "loss": 0.6776,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9317820586178397,
+      "grad_norm": 2.2348231153022944,
+      "learning_rate": 1.6510773741370213e-05,
+      "loss": 0.6161,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9322906732786572,
+      "grad_norm": 2.017250765658621,
+      "learning_rate": 1.6506516954226745e-05,
+      "loss": 0.7009,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9327992879394749,
+      "grad_norm": 2.0931884240251524,
+      "learning_rate": 1.6502258121547617e-05,
+      "loss": 0.7369,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9333079026002925,
+      "grad_norm": 2.0169354104710955,
+      "learning_rate": 1.6497997244671734e-05,
+      "loss": 0.7654,
+      "step": 1835
+    },
+    {
+      "epoch": 0.93381651726111,
+      "grad_norm": 2.2784721497237244,
+      "learning_rate": 1.6493734324938638e-05,
+      "loss": 0.8158,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9343251319219277,
+      "grad_norm": 2.2036139845133764,
+      "learning_rate": 1.648946936368852e-05,
+      "loss": 0.8926,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9348337465827452,
+      "grad_norm": 2.150171308567986,
+      "learning_rate": 1.6485202362262203e-05,
+      "loss": 0.8371,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9353423612435628,
+      "grad_norm": 2.23807139852595,
+      "learning_rate": 1.648093332200117e-05,
+      "loss": 0.714,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9358509759043805,
+      "grad_norm": 2.7410481142638616,
+      "learning_rate": 1.6476662244247524e-05,
+      "loss": 0.7885,
+      "step": 1840
+    },
+    {
+      "epoch": 0.936359590565198,
+      "grad_norm": 2.257522405975778,
+      "learning_rate": 1.647238913034402e-05,
+      "loss": 0.8093,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9368682052260157,
+      "grad_norm": 2.376002551844861,
+      "learning_rate": 1.646811398163405e-05,
+      "loss": 0.8407,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9373768198868332,
+      "grad_norm": 2.1157007381343624,
+      "learning_rate": 1.6463836799461644e-05,
+      "loss": 0.7888,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9378854345476508,
+      "grad_norm": 2.4583949927232163,
+      "learning_rate": 1.645955758517148e-05,
+      "loss": 0.6776,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9383940492084685,
+      "grad_norm": 2.065680161748098,
+      "learning_rate": 1.6455276340108864e-05,
+      "loss": 0.7492,
+      "step": 1845
+    },
+    {
+      "epoch": 0.938902663869286,
+      "grad_norm": 2.2148992945250794,
+      "learning_rate": 1.6450993065619747e-05,
+      "loss": 0.7119,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9394112785301036,
+      "grad_norm": 1.9010911769603411,
+      "learning_rate": 1.644670776305071e-05,
+      "loss": 0.8086,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9399198931909212,
+      "grad_norm": 2.4518065584555093,
+      "learning_rate": 1.6442420433748986e-05,
+      "loss": 0.7807,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9404285078517388,
+      "grad_norm": 1.9970534792612693,
+      "learning_rate": 1.6438131079062434e-05,
+      "loss": 0.6831,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9409371225125565,
+      "grad_norm": 2.0099328863981274,
+      "learning_rate": 1.643383970033955e-05,
+      "loss": 0.7048,
+      "step": 1850
+    },
+    {
+      "epoch": 0.941445737173374,
+      "grad_norm": 2.2205135160505955,
+      "learning_rate": 1.642954629892947e-05,
+      "loss": 0.8016,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9419543518341916,
+      "grad_norm": 1.7871777464081378,
+      "learning_rate": 1.6425250876181963e-05,
+      "loss": 0.7101,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9424629664950093,
+      "grad_norm": 2.160891637087878,
+      "learning_rate": 1.6420953433447436e-05,
+      "loss": 0.7054,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9429715811558268,
+      "grad_norm": 2.431786516707788,
+      "learning_rate": 1.6416653972076938e-05,
+      "loss": 0.8107,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9434801958166444,
+      "grad_norm": 1.9083796627412237,
+      "learning_rate": 1.641235249342213e-05,
+      "loss": 0.838,
+      "step": 1855
+    },
+    {
+      "epoch": 0.943988810477462,
+      "grad_norm": 2.019725797568393,
+      "learning_rate": 1.6408048998835333e-05,
+      "loss": 0.7851,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9444974251382796,
+      "grad_norm": 1.8288956672398127,
+      "learning_rate": 1.6403743489669487e-05,
+      "loss": 0.7496,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9450060397990973,
+      "grad_norm": 2.15392886189023,
+      "learning_rate": 1.639943596727817e-05,
+      "loss": 0.8131,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9455146544599148,
+      "grad_norm": 2.201207835082262,
+      "learning_rate": 1.6395126433015595e-05,
+      "loss": 0.7042,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9460232691207324,
+      "grad_norm": 2.1536461408461904,
+      "learning_rate": 1.6390814888236598e-05,
+      "loss": 0.8019,
+      "step": 1860
+    },
+    {
+      "epoch": 0.94653188378155,
+      "grad_norm": 1.9784647717582384,
+      "learning_rate": 1.6386501334296653e-05,
+      "loss": 0.6505,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9470404984423676,
+      "grad_norm": 2.0100211378665676,
+      "learning_rate": 1.6382185772551876e-05,
+      "loss": 0.7568,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9475491131031852,
+      "grad_norm": 1.8453273168577853,
+      "learning_rate": 1.6377868204358998e-05,
+      "loss": 0.8177,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9480577277640028,
+      "grad_norm": 2.447766130296252,
+      "learning_rate": 1.6373548631075385e-05,
+      "loss": 0.596,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9485663424248204,
+      "grad_norm": 2.098386044414622,
+      "learning_rate": 1.636922705405904e-05,
+      "loss": 0.8082,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9490749570856379,
+      "grad_norm": 2.0591868753135008,
+      "learning_rate": 1.6364903474668595e-05,
+      "loss": 0.8311,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9495835717464556,
+      "grad_norm": 1.9073121910581223,
+      "learning_rate": 1.63605778942633e-05,
+      "loss": 0.7392,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9500921864072732,
+      "grad_norm": 1.948089945492759,
+      "learning_rate": 1.6356250314203044e-05,
+      "loss": 0.8908,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9506008010680908,
+      "grad_norm": 1.8678822521698715,
+      "learning_rate": 1.6351920735848346e-05,
+      "loss": 0.7062,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9511094157289084,
+      "grad_norm": 1.934605620753794,
+      "learning_rate": 1.6347589160560346e-05,
+      "loss": 0.7128,
+      "step": 1870
+    },
+    {
+      "epoch": 0.951618030389726,
+      "grad_norm": 2.5307750491792707,
+      "learning_rate": 1.6343255589700818e-05,
+      "loss": 0.8578,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9521266450505436,
+      "grad_norm": 2.675322154958632,
+      "learning_rate": 1.6338920024632167e-05,
+      "loss": 0.775,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9526352597113612,
+      "grad_norm": 2.2678611506457793,
+      "learning_rate": 1.633458246671741e-05,
+      "loss": 0.8085,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9531438743721787,
+      "grad_norm": 2.0673424487527554,
+      "learning_rate": 1.6330242917320202e-05,
+      "loss": 0.8251,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9536524890329964,
+      "grad_norm": 2.443334394856318,
+      "learning_rate": 1.6325901377804825e-05,
+      "loss": 0.6469,
+      "step": 1875
+    },
+    {
+      "epoch": 0.954161103693814,
+      "grad_norm": 2.110268071759197,
+      "learning_rate": 1.6321557849536182e-05,
+      "loss": 0.767,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9546697183546315,
+      "grad_norm": 1.9934874837122216,
+      "learning_rate": 1.6317212333879798e-05,
+      "loss": 0.8246,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9551783330154492,
+      "grad_norm": 1.8322075797588895,
+      "learning_rate": 1.6312864832201837e-05,
+      "loss": 0.6564,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9556869476762667,
+      "grad_norm": 2.00072251620584,
+      "learning_rate": 1.630851534586907e-05,
+      "loss": 0.875,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9561955623370844,
+      "grad_norm": 2.106193573419447,
+      "learning_rate": 1.6304163876248904e-05,
+      "loss": 0.8537,
+      "step": 1880
+    },
+    {
+      "epoch": 0.956704176997902,
+      "grad_norm": 2.344301591179969,
+      "learning_rate": 1.629981042470936e-05,
+      "loss": 0.7082,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9572127916587195,
+      "grad_norm": 2.112946489088915,
+      "learning_rate": 1.6295454992619094e-05,
+      "loss": 0.6322,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9577214063195372,
+      "grad_norm": 1.8462806403652916,
+      "learning_rate": 1.6291097581347374e-05,
+      "loss": 0.6748,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9582300209803548,
+      "grad_norm": 2.702762910262492,
+      "learning_rate": 1.6286738192264097e-05,
+      "loss": 0.7898,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9587386356411723,
+      "grad_norm": 2.309598494311411,
+      "learning_rate": 1.6282376826739775e-05,
+      "loss": 0.7445,
+      "step": 1885
+    },
+    {
+      "epoch": 0.95924725030199,
+      "grad_norm": 1.9927793334287474,
+      "learning_rate": 1.627801348614555e-05,
+      "loss": 0.7661,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9597558649628075,
+      "grad_norm": 1.9943480500516637,
+      "learning_rate": 1.6273648171853177e-05,
+      "loss": 0.6858,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9602644796236252,
+      "grad_norm": 1.9995335988767755,
+      "learning_rate": 1.6269280885235037e-05,
+      "loss": 0.8511,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9607730942844428,
+      "grad_norm": 2.354204183183907,
+      "learning_rate": 1.6264911627664127e-05,
+      "loss": 0.7705,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9612817089452603,
+      "grad_norm": 2.1443124461466008,
+      "learning_rate": 1.6260540400514065e-05,
+      "loss": 0.7378,
+      "step": 1890
+    },
+    {
+      "epoch": 0.961790323606078,
+      "grad_norm": 2.228150668003066,
+      "learning_rate": 1.625616720515909e-05,
+      "loss": 0.6691,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9622989382668955,
+      "grad_norm": 2.167226523424122,
+      "learning_rate": 1.6251792042974058e-05,
+      "loss": 0.8196,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9628075529277131,
+      "grad_norm": 7.174761289021633,
+      "learning_rate": 1.6247414915334446e-05,
+      "loss": 0.7156,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9633161675885308,
+      "grad_norm": 2.3403193479242432,
+      "learning_rate": 1.6243035823616348e-05,
+      "loss": 0.7177,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9638247822493483,
+      "grad_norm": 2.0388147196270725,
+      "learning_rate": 1.6238654769196466e-05,
+      "loss": 0.7265,
+      "step": 1895
+    },
+    {
+      "epoch": 0.964333396910166,
+      "grad_norm": 1.8253575432014575,
+      "learning_rate": 1.6234271753452133e-05,
+      "loss": 0.7174,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9648420115709835,
+      "grad_norm": 1.9377377887134852,
+      "learning_rate": 1.6229886777761295e-05,
+      "loss": 0.7072,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9653506262318011,
+      "grad_norm": 1.922867929399618,
+      "learning_rate": 1.6225499843502507e-05,
+      "loss": 0.706,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9658592408926188,
+      "grad_norm": 1.8898732606082635,
+      "learning_rate": 1.6221110952054947e-05,
+      "loss": 0.6622,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9663678555534363,
+      "grad_norm": 2.405443824373793,
+      "learning_rate": 1.6216720104798408e-05,
+      "loss": 0.8034,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9663678555534363,
+      "eval_loss": 0.710275411605835,
+      "eval_runtime": 170.0388,
+      "eval_samples_per_second": 19.478,
+      "eval_steps_per_second": 9.739,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9668764702142539,
+      "grad_norm": 2.7995331136547508,
+      "learning_rate": 1.6212327303113293e-05,
+      "loss": 0.7253,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9673850848750716,
+      "grad_norm": 2.0780396408103714,
+      "learning_rate": 1.620793254838062e-05,
+      "loss": 0.7201,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9678936995358891,
+      "grad_norm": 2.4626271094033423,
+      "learning_rate": 1.6203535841982033e-05,
+      "loss": 0.7424,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9684023141967067,
+      "grad_norm": 2.212574490947304,
+      "learning_rate": 1.6199137185299774e-05,
+      "loss": 0.7065,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9689109288575243,
+      "grad_norm": 2.0615666907619032,
+      "learning_rate": 1.61947365797167e-05,
+      "loss": 0.7766,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9694195435183419,
+      "grad_norm": 1.8433300768574559,
+      "learning_rate": 1.619033402661629e-05,
+      "loss": 0.7938,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9699281581791596,
+      "grad_norm": 2.235012150332474,
+      "learning_rate": 1.6185929527382628e-05,
+      "loss": 0.8163,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9704367728399771,
+      "grad_norm": 1.9356411045822237,
+      "learning_rate": 1.6181523083400413e-05,
+      "loss": 0.6967,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9709453875007947,
+      "grad_norm": 2.144851855161755,
+      "learning_rate": 1.6177114696054955e-05,
+      "loss": 0.7705,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9714540021616123,
+      "grad_norm": 1.9847009346083366,
+      "learning_rate": 1.6172704366732174e-05,
+      "loss": 0.7157,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9719626168224299,
+      "grad_norm": 2.643165855714357,
+      "learning_rate": 1.6168292096818597e-05,
+      "loss": 0.6501,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9724712314832475,
+      "grad_norm": 2.2145696668899983,
+      "learning_rate": 1.6163877887701367e-05,
+      "loss": 0.7628,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9729798461440651,
+      "grad_norm": 1.923998768011004,
+      "learning_rate": 1.6159461740768236e-05,
+      "loss": 0.7445,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9734884608048827,
+      "grad_norm": 2.069959387265653,
+      "learning_rate": 1.6155043657407564e-05,
+      "loss": 0.7413,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9739970754657002,
+      "grad_norm": 2.0161692749437115,
+      "learning_rate": 1.6150623639008317e-05,
+      "loss": 0.6376,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9745056901265179,
+      "grad_norm": 2.1609026704023826,
+      "learning_rate": 1.6146201686960075e-05,
+      "loss": 0.7689,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9750143047873355,
+      "grad_norm": 2.2301414236205166,
+      "learning_rate": 1.6141777802653024e-05,
+      "loss": 0.7769,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9755229194481531,
+      "grad_norm": 2.051140337795845,
+      "learning_rate": 1.613735198747795e-05,
+      "loss": 0.7439,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9760315341089707,
+      "grad_norm": 2.13995506774221,
+      "learning_rate": 1.6132924242826255e-05,
+      "loss": 0.7225,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9765401487697883,
+      "grad_norm": 2.3529749078997444,
+      "learning_rate": 1.6128494570089945e-05,
+      "loss": 0.7215,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9770487634306059,
+      "grad_norm": 2.3510295479433547,
+      "learning_rate": 1.6124062970661634e-05,
+      "loss": 0.7644,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9775573780914235,
+      "grad_norm": 2.074857279453828,
+      "learning_rate": 1.6119629445934535e-05,
+      "loss": 0.8318,
+      "step": 1922
+    },
+    {
+      "epoch": 0.978065992752241,
+      "grad_norm": 2.0650065825128823,
+      "learning_rate": 1.6115193997302476e-05,
+      "loss": 0.7489,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9785746074130587,
+      "grad_norm": 1.9528127802205963,
+      "learning_rate": 1.611075662615988e-05,
+      "loss": 0.681,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9790832220738763,
+      "grad_norm": 2.195562273796878,
+      "learning_rate": 1.6106317333901786e-05,
+      "loss": 0.8023,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 2.336617361284085,
+      "learning_rate": 1.6101876121923827e-05,
+      "loss": 0.9204,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9801004513955115,
+      "grad_norm": 2.3426746190145464,
+      "learning_rate": 1.609743299162224e-05,
+      "loss": 0.7869,
+      "step": 1927
+    },
+    {
+      "epoch": 0.980609066056329,
+      "grad_norm": 1.7395377086511623,
+      "learning_rate": 1.609298794439387e-05,
+      "loss": 0.7356,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9811176807171467,
+      "grad_norm": 2.2182532815606932,
+      "learning_rate": 1.6088540981636162e-05,
+      "loss": 0.8463,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9816262953779643,
+      "grad_norm": 2.052200610945406,
+      "learning_rate": 1.6084092104747163e-05,
+      "loss": 0.8071,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9821349100387818,
+      "grad_norm": 1.9232403627047807,
+      "learning_rate": 1.6079641315125526e-05,
+      "loss": 0.8175,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9826435246995995,
+      "grad_norm": 2.0702212907384934,
+      "learning_rate": 1.6075188614170497e-05,
+      "loss": 0.667,
+      "step": 1932
+    },
+    {
+      "epoch": 0.983152139360417,
+      "grad_norm": 2.179787254101769,
+      "learning_rate": 1.607073400328193e-05,
+      "loss": 0.7868,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9836607540212347,
+      "grad_norm": 2.091144249813736,
+      "learning_rate": 1.6066277483860276e-05,
+      "loss": 0.746,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9841693686820523,
+      "grad_norm": 1.9844283449707227,
+      "learning_rate": 1.606181905730659e-05,
+      "loss": 0.6912,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9846779833428698,
+      "grad_norm": 2.2060630092320355,
+      "learning_rate": 1.605735872502252e-05,
+      "loss": 0.7914,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9851865980036875,
+      "grad_norm": 2.451131508787393,
+      "learning_rate": 1.605289648841032e-05,
+      "loss": 0.8452,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9856952126645051,
+      "grad_norm": 1.9613528476684787,
+      "learning_rate": 1.604843234887284e-05,
+      "loss": 0.761,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9862038273253226,
+      "grad_norm": 1.9454133647216088,
+      "learning_rate": 1.6043966307813526e-05,
+      "loss": 0.8429,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9867124419861403,
+      "grad_norm": 1.9608736673733453,
+      "learning_rate": 1.603949836663642e-05,
+      "loss": 0.7197,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9872210566469578,
+      "grad_norm": 1.7295181729105527,
+      "learning_rate": 1.603502852674617e-05,
+      "loss": 0.6538,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9877296713077754,
+      "grad_norm": 2.6278610986032436,
+      "learning_rate": 1.6030556789548014e-05,
+      "loss": 0.8581,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9882382859685931,
+      "grad_norm": 2.114612096832031,
+      "learning_rate": 1.6026083156447787e-05,
+      "loss": 0.7652,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9887469006294106,
+      "grad_norm": 2.242099348189571,
+      "learning_rate": 1.6021607628851926e-05,
+      "loss": 0.7453,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9892555152902283,
+      "grad_norm": 2.2773485658048354,
+      "learning_rate": 1.6017130208167452e-05,
+      "loss": 0.8197,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9897641299510458,
+      "grad_norm": 2.318656150378698,
+      "learning_rate": 1.6012650895801994e-05,
+      "loss": 0.7305,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9902727446118634,
+      "grad_norm": 2.5640244950122746,
+      "learning_rate": 1.6008169693163767e-05,
+      "loss": 0.8184,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9907813592726811,
+      "grad_norm": 1.9327906318235413,
+      "learning_rate": 1.6003686601661586e-05,
+      "loss": 0.7884,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9912899739334986,
+      "grad_norm": 2.2605584681694686,
+      "learning_rate": 1.599920162270485e-05,
+      "loss": 0.6986,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9917985885943162,
+      "grad_norm": 1.906600544701077,
+      "learning_rate": 1.5994714757703568e-05,
+      "loss": 0.7071,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9923072032551338,
+      "grad_norm": 1.7935254930099371,
+      "learning_rate": 1.5990226008068327e-05,
+      "loss": 0.7418,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9928158179159514,
+      "grad_norm": 2.4044368898355697,
+      "learning_rate": 1.598573537521031e-05,
+      "loss": 0.7482,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9933244325767691,
+      "grad_norm": 1.9060125359386153,
+      "learning_rate": 1.5981242860541297e-05,
+      "loss": 0.7179,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9938330472375866,
+      "grad_norm": 1.806007794043366,
+      "learning_rate": 1.5976748465473658e-05,
+      "loss": 0.737,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9943416618984042,
+      "grad_norm": 1.8932546187993904,
+      "learning_rate": 1.5972252191420347e-05,
+      "loss": 0.6973,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9948502765592219,
+      "grad_norm": 1.8905324361541325,
+      "learning_rate": 1.5967754039794916e-05,
+      "loss": 0.8049,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9953588912200394,
+      "grad_norm": 1.9424513032325128,
+      "learning_rate": 1.5963254012011508e-05,
+      "loss": 0.8983,
+      "step": 1957
+    },
+    {
+      "epoch": 0.995867505880857,
+      "grad_norm": 2.3438165273761222,
+      "learning_rate": 1.5958752109484852e-05,
+      "loss": 0.7276,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9963761205416746,
+      "grad_norm": 2.180881842993367,
+      "learning_rate": 1.5954248333630267e-05,
+      "loss": 0.7318,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9968847352024922,
+      "grad_norm": 2.0436170435596632,
+      "learning_rate": 1.5949742685863664e-05,
+      "loss": 0.7343,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9973933498633099,
+      "grad_norm": 2.208275584299518,
+      "learning_rate": 1.5945235167601542e-05,
+      "loss": 0.759,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9979019645241274,
+      "grad_norm": 1.8931750290243214,
+      "learning_rate": 1.5940725780260977e-05,
+      "loss": 0.7534,
+      "step": 1962
+    },
+    {
+      "epoch": 0.998410579184945,
+      "grad_norm": 1.9085699254039004,
+      "learning_rate": 1.593621452525965e-05,
+      "loss": 0.7375,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9989191938457626,
+      "grad_norm": 2.1641226529461375,
+      "learning_rate": 1.5931701404015818e-05,
+      "loss": 0.7333,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9994278085065802,
+      "grad_norm": 2.1392297583225655,
+      "learning_rate": 1.592718641794833e-05,
+      "loss": 0.773,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9999364231673978,
+      "grad_norm": 2.094855628672408,
+      "learning_rate": 1.592266956847662e-05,
+      "loss": 0.7768,
+      "step": 1966
+    },
+    {
+      "epoch": 1.0004450378282155,
+      "grad_norm": 1.8517201030630657,
+      "learning_rate": 1.59181508570207e-05,
+      "loss": 0.5666,
+      "step": 1967
+    },
+    {
+      "epoch": 1.000953652489033,
+      "grad_norm": 1.926871093884622,
+      "learning_rate": 1.5913630285001184e-05,
+      "loss": 0.6184,
+      "step": 1968
+    },
+    {
+      "epoch": 1.0014622671498505,
+      "grad_norm": 1.7699366225194606,
+      "learning_rate": 1.5909107853839252e-05,
+      "loss": 0.5082,
+      "step": 1969
+    },
+    {
+      "epoch": 1.0019708818106683,
+      "grad_norm": 1.8837412209613156,
+      "learning_rate": 1.590458356495668e-05,
+      "loss": 0.5489,
+      "step": 1970
+    },
+    {
+      "epoch": 1.0024794964714858,
+      "grad_norm": 2.051470773219551,
+      "learning_rate": 1.5900057419775833e-05,
+      "loss": 0.5934,
+      "step": 1971
+    },
+    {
+      "epoch": 1.0029881111323034,
+      "grad_norm": 1.7497905878383966,
+      "learning_rate": 1.5895529419719645e-05,
+      "loss": 0.3822,
+      "step": 1972
+    },
+    {
+      "epoch": 1.0034967257931209,
+      "grad_norm": 1.9473917316707468,
+      "learning_rate": 1.5890999566211643e-05,
+      "loss": 0.5524,
+      "step": 1973
+    },
+    {
+      "epoch": 1.0040053404539386,
+      "grad_norm": 2.2773121329414385,
+      "learning_rate": 1.5886467860675928e-05,
+      "loss": 0.556,
+      "step": 1974
+    },
+    {
+      "epoch": 1.0045139551147562,
+      "grad_norm": 1.9730276230329926,
+      "learning_rate": 1.58819343045372e-05,
+      "loss": 0.5102,
+      "step": 1975
+    },
+    {
+      "epoch": 1.0050225697755737,
+      "grad_norm": 2.036536511988097,
+      "learning_rate": 1.5877398899220717e-05,
+      "loss": 0.4918,
+      "step": 1976
+    },
+    {
+      "epoch": 1.0055311844363914,
+      "grad_norm": 2.1398321757955854,
+      "learning_rate": 1.5872861646152347e-05,
+      "loss": 0.5165,
+      "step": 1977
+    },
+    {
+      "epoch": 1.006039799097209,
+      "grad_norm": 1.711673386711343,
+      "learning_rate": 1.586832254675851e-05,
+      "loss": 0.356,
+      "step": 1978
+    },
+    {
+      "epoch": 1.0065484137580265,
+      "grad_norm": 2.3291815082256897,
+      "learning_rate": 1.5863781602466228e-05,
+      "loss": 0.602,
+      "step": 1979
+    },
+    {
+      "epoch": 1.0070570284188443,
+      "grad_norm": 2.2622354342223256,
+      "learning_rate": 1.585923881470309e-05,
+      "loss": 0.4925,
+      "step": 1980
+    },
+    {
+      "epoch": 1.0075656430796618,
+      "grad_norm": 2.203070132607174,
+      "learning_rate": 1.5854694184897273e-05,
+      "loss": 0.4955,
+      "step": 1981
+    },
+    {
+      "epoch": 1.0080742577404793,
+      "grad_norm": 2.171766758698321,
+      "learning_rate": 1.5850147714477527e-05,
+      "loss": 0.4776,
+      "step": 1982
+    },
+    {
+      "epoch": 1.008582872401297,
+      "grad_norm": 2.1256034218720044,
+      "learning_rate": 1.5845599404873184e-05,
+      "loss": 0.4798,
+      "step": 1983
+    },
+    {
+      "epoch": 1.0090914870621146,
+      "grad_norm": 2.2474433360525667,
+      "learning_rate": 1.584104925751415e-05,
+      "loss": 0.5655,
+      "step": 1984
+    },
+    {
+      "epoch": 1.0096001017229321,
+      "grad_norm": 1.691273179667235,
+      "learning_rate": 1.583649727383092e-05,
+      "loss": 0.446,
+      "step": 1985
+    },
+    {
+      "epoch": 1.0101087163837497,
+      "grad_norm": 2.3165842023301453,
+      "learning_rate": 1.5831943455254552e-05,
+      "loss": 0.4949,
+      "step": 1986
+    },
+    {
+      "epoch": 1.0106173310445674,
+      "grad_norm": 1.9089633828873172,
+      "learning_rate": 1.5827387803216693e-05,
+      "loss": 0.4811,
+      "step": 1987
+    },
+    {
+      "epoch": 1.011125945705385,
+      "grad_norm": 2.3282253576576384,
+      "learning_rate": 1.5822830319149552e-05,
+      "loss": 0.5206,
+      "step": 1988
+    },
+    {
+      "epoch": 1.0116345603662025,
+      "grad_norm": 2.2287840351409267,
+      "learning_rate": 1.581827100448593e-05,
+      "loss": 0.5657,
+      "step": 1989
+    },
+    {
+      "epoch": 1.0121431750270202,
+      "grad_norm": 2.124102234529773,
+      "learning_rate": 1.5813709860659192e-05,
+      "loss": 0.6387,
+      "step": 1990
+    },
+    {
+      "epoch": 1.0126517896878378,
+      "grad_norm": 2.2841412759970385,
+      "learning_rate": 1.5809146889103285e-05,
+      "loss": 0.53,
+      "step": 1991
+    },
+    {
+      "epoch": 1.0131604043486553,
+      "grad_norm": 1.8906716585313101,
+      "learning_rate": 1.5804582091252728e-05,
+      "loss": 0.5466,
+      "step": 1992
+    },
+    {
+      "epoch": 1.013669019009473,
+      "grad_norm": 2.224236059305727,
+      "learning_rate": 1.580001546854261e-05,
+      "loss": 0.5582,
+      "step": 1993
+    },
+    {
+      "epoch": 1.0141776336702906,
+      "grad_norm": 2.176075338489957,
+      "learning_rate": 1.5795447022408602e-05,
+      "loss": 0.5159,
+      "step": 1994
+    },
+    {
+      "epoch": 1.014686248331108,
+      "grad_norm": 2.132406119172936,
+      "learning_rate": 1.5790876754286942e-05,
+      "loss": 0.5419,
+      "step": 1995
+    },
+    {
+      "epoch": 1.0151948629919259,
+      "grad_norm": 2.4239095694163195,
+      "learning_rate": 1.5786304665614444e-05,
+      "loss": 0.459,
+      "step": 1996
+    },
+    {
+      "epoch": 1.0157034776527434,
+      "grad_norm": 1.9701605693616404,
+      "learning_rate": 1.578173075782849e-05,
+      "loss": 0.4107,
+      "step": 1997
+    },
+    {
+      "epoch": 1.016212092313561,
+      "grad_norm": 2.2248712909632635,
+      "learning_rate": 1.577715503236704e-05,
+      "loss": 0.532,
+      "step": 1998
+    },
+    {
+      "epoch": 1.0167207069743784,
+      "grad_norm": 2.420181962415465,
+      "learning_rate": 1.5772577490668618e-05,
+      "loss": 0.5978,
+      "step": 1999
+    },
+    {
+      "epoch": 1.0172293216351962,
+      "grad_norm": 1.8909527202971443,
+      "learning_rate": 1.5767998134172336e-05,
+      "loss": 0.4853,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0172293216351962,
+      "eval_loss": 0.7371433973312378,
+      "eval_runtime": 171.3334,
+      "eval_samples_per_second": 19.331,
+      "eval_steps_per_second": 9.665,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0177379362960137,
+      "grad_norm": 2.0157771090104304,
+      "learning_rate": 1.5763416964317845e-05,
+      "loss": 0.4864,
+      "step": 2001
+    },
+    {
+      "epoch": 1.0182465509568313,
+      "grad_norm": 2.135478718173966,
+      "learning_rate": 1.5758833982545404e-05,
+      "loss": 0.4554,
+      "step": 2002
+    },
+    {
+      "epoch": 1.018755165617649,
+      "grad_norm": 1.9813347903618947,
+      "learning_rate": 1.5754249190295807e-05,
+      "loss": 0.5171,
+      "step": 2003
+    },
+    {
+      "epoch": 1.0192637802784665,
+      "grad_norm": 2.5299558745746777,
+      "learning_rate": 1.574966258901044e-05,
+      "loss": 0.4866,
+      "step": 2004
+    },
+    {
+      "epoch": 1.019772394939284,
+      "grad_norm": 2.3820285195309663,
+      "learning_rate": 1.5745074180131255e-05,
+      "loss": 0.5406,
+      "step": 2005
+    },
+    {
+      "epoch": 1.0202810096001018,
+      "grad_norm": 2.154584600856066,
+      "learning_rate": 1.5740483965100758e-05,
+      "loss": 0.577,
+      "step": 2006
+    },
+    {
+      "epoch": 1.0207896242609193,
+      "grad_norm": 2.2452274144456115,
+      "learning_rate": 1.573589194536204e-05,
+      "loss": 0.4418,
+      "step": 2007
+    },
+    {
+      "epoch": 1.0212982389217369,
+      "grad_norm": 1.9213478792135417,
+      "learning_rate": 1.573129812235875e-05,
+      "loss": 0.5636,
+      "step": 2008
+    },
+    {
+      "epoch": 1.0218068535825544,
+      "grad_norm": 2.0274219534606672,
+      "learning_rate": 1.5726702497535104e-05,
+      "loss": 0.5819,
+      "step": 2009
+    },
+    {
+      "epoch": 1.0223154682433722,
+      "grad_norm": 2.14569782705745,
+      "learning_rate": 1.5722105072335892e-05,
+      "loss": 0.4554,
+      "step": 2010
+    },
+    {
+      "epoch": 1.0228240829041897,
+      "grad_norm": 2.2226937932764,
+      "learning_rate": 1.5717505848206457e-05,
+      "loss": 0.4921,
+      "step": 2011
+    },
+    {
+      "epoch": 1.0233326975650072,
+      "grad_norm": 2.190716585647347,
+      "learning_rate": 1.571290482659272e-05,
+      "loss": 0.5662,
+      "step": 2012
+    },
+    {
+      "epoch": 1.023841312225825,
+      "grad_norm": 2.0517259728758703,
+      "learning_rate": 1.570830200894116e-05,
+      "loss": 0.5401,
+      "step": 2013
+    },
+    {
+      "epoch": 1.0243499268866425,
+      "grad_norm": 2.002758388922187,
+      "learning_rate": 1.5703697396698823e-05,
+      "loss": 0.4912,
+      "step": 2014
+    },
+    {
+      "epoch": 1.02485854154746,
+      "grad_norm": 1.8798369008192282,
+      "learning_rate": 1.569909099131332e-05,
+      "loss": 0.5284,
+      "step": 2015
+    },
+    {
+      "epoch": 1.0253671562082778,
+      "grad_norm": 1.8707801380821651,
+      "learning_rate": 1.569448279423282e-05,
+      "loss": 0.4241,
+      "step": 2016
+    },
+    {
+      "epoch": 1.0258757708690953,
+      "grad_norm": 2.1133319051880126,
+      "learning_rate": 1.568987280690607e-05,
+      "loss": 0.5276,
+      "step": 2017
+    },
+    {
+      "epoch": 1.0263843855299128,
+      "grad_norm": 2.116009035851957,
+      "learning_rate": 1.568526103078236e-05,
+      "loss": 0.4933,
+      "step": 2018
+    },
+    {
+      "epoch": 1.0268930001907306,
+      "grad_norm": 1.7644935966412263,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.4052,
+      "step": 2019
+    },
+    {
+      "epoch": 1.0274016148515481,
+      "grad_norm": 2.216579719023738,
+      "learning_rate": 1.5676032117944088e-05,
+      "loss": 0.4635,
+      "step": 2020
+    },
+    {
+      "epoch": 1.0279102295123657,
+      "grad_norm": 2.4103069041941536,
+      "learning_rate": 1.567141498413093e-05,
+      "loss": 0.4952,
+      "step": 2021
+    },
+    {
+      "epoch": 1.0284188441731832,
+      "grad_norm": 1.8772075373477346,
+      "learning_rate": 1.5666796067323632e-05,
+      "loss": 0.4649,
+      "step": 2022
+    },
+    {
+      "epoch": 1.028927458834001,
+      "grad_norm": 2.3660968748302644,
+      "learning_rate": 1.566217536897431e-05,
+      "loss": 0.5699,
+      "step": 2023
+    },
+    {
+      "epoch": 1.0294360734948185,
+      "grad_norm": 2.5704353094473533,
+      "learning_rate": 1.565755289053562e-05,
+      "loss": 0.6107,
+      "step": 2024
+    },
+    {
+      "epoch": 1.029944688155636,
+      "grad_norm": 2.207149752245894,
+      "learning_rate": 1.565292863346079e-05,
+      "loss": 0.4996,
+      "step": 2025
+    },
+    {
+      "epoch": 1.0304533028164538,
+      "grad_norm": 2.1884055306968744,
+      "learning_rate": 1.5648302599203612e-05,
+      "loss": 0.5243,
+      "step": 2026
+    },
+    {
+      "epoch": 1.0309619174772713,
+      "grad_norm": 2.2447111700956315,
+      "learning_rate": 1.5643674789218425e-05,
+      "loss": 0.519,
+      "step": 2027
+    },
+    {
+      "epoch": 1.0314705321380888,
+      "grad_norm": 2.3806646954348367,
+      "learning_rate": 1.5639045204960137e-05,
+      "loss": 0.512,
+      "step": 2028
+    },
+    {
+      "epoch": 1.0319791467989066,
+      "grad_norm": 2.2588670637134762,
+      "learning_rate": 1.5634413847884205e-05,
+      "loss": 0.4875,
+      "step": 2029
+    },
+    {
+      "epoch": 1.032487761459724,
+      "grad_norm": 1.80310451555434,
+      "learning_rate": 1.5629780719446647e-05,
+      "loss": 0.3999,
+      "step": 2030
+    },
+    {
+      "epoch": 1.0329963761205416,
+      "grad_norm": 2.0782352504310078,
+      "learning_rate": 1.562514582110404e-05,
+      "loss": 0.5285,
+      "step": 2031
+    },
+    {
+      "epoch": 1.0335049907813594,
+      "grad_norm": 2.08502405861712,
+      "learning_rate": 1.5620509154313515e-05,
+      "loss": 0.4869,
+      "step": 2032
+    },
+    {
+      "epoch": 1.034013605442177,
+      "grad_norm": 2.0425031725930047,
+      "learning_rate": 1.5615870720532753e-05,
+      "loss": 0.4812,
+      "step": 2033
+    },
+    {
+      "epoch": 1.0345222201029944,
+      "grad_norm": 1.963968798845994,
+      "learning_rate": 1.561123052122001e-05,
+      "loss": 0.4715,
+      "step": 2034
+    },
+    {
+      "epoch": 1.035030834763812,
+      "grad_norm": 2.1555402389235634,
+      "learning_rate": 1.560658855783407e-05,
+      "loss": 0.4394,
+      "step": 2035
+    },
+    {
+      "epoch": 1.0355394494246297,
+      "grad_norm": 1.879198904106337,
+      "learning_rate": 1.5601944831834294e-05,
+      "loss": 0.4519,
+      "step": 2036
+    },
+    {
+      "epoch": 1.0360480640854473,
+      "grad_norm": 1.894396653260838,
+      "learning_rate": 1.559729934468059e-05,
+      "loss": 0.4688,
+      "step": 2037
+    },
+    {
+      "epoch": 1.0365566787462648,
+      "grad_norm": 2.0451704309604137,
+      "learning_rate": 1.559265209783341e-05,
+      "loss": 0.4406,
+      "step": 2038
+    },
+    {
+      "epoch": 1.0370652934070825,
+      "grad_norm": 2.3213811911474247,
+      "learning_rate": 1.5588003092753776e-05,
+      "loss": 0.5017,
+      "step": 2039
+    },
+    {
+      "epoch": 1.0375739080679,
+      "grad_norm": 2.311177055730385,
+      "learning_rate": 1.5583352330903246e-05,
+      "loss": 0.5115,
+      "step": 2040
+    },
+    {
+      "epoch": 1.0380825227287176,
+      "grad_norm": 1.986583111955958,
+      "learning_rate": 1.5578699813743947e-05,
+      "loss": 0.4856,
+      "step": 2041
+    },
+    {
+      "epoch": 1.0385911373895353,
+      "grad_norm": 1.9898482407637905,
+      "learning_rate": 1.5574045542738543e-05,
+      "loss": 0.5246,
+      "step": 2042
+    },
+    {
+      "epoch": 1.0390997520503529,
+      "grad_norm": 2.0697267099473855,
+      "learning_rate": 1.556938951935026e-05,
+      "loss": 0.4789,
+      "step": 2043
+    },
+    {
+      "epoch": 1.0396083667111704,
+      "grad_norm": 2.1535584610017886,
+      "learning_rate": 1.5564731745042867e-05,
+      "loss": 0.5903,
+      "step": 2044
+    },
+    {
+      "epoch": 1.040116981371988,
+      "grad_norm": 2.031122878571768,
+      "learning_rate": 1.556007222128069e-05,
+      "loss": 0.4722,
+      "step": 2045
+    },
+    {
+      "epoch": 1.0406255960328057,
+      "grad_norm": 2.1606731658589844,
+      "learning_rate": 1.5555410949528604e-05,
+      "loss": 0.421,
+      "step": 2046
+    },
+    {
+      "epoch": 1.0411342106936232,
+      "grad_norm": 1.7894634480523157,
+      "learning_rate": 1.5550747931252022e-05,
+      "loss": 0.5028,
+      "step": 2047
+    },
+    {
+      "epoch": 1.0416428253544407,
+      "grad_norm": 2.4251985448984863,
+      "learning_rate": 1.5546083167916926e-05,
+      "loss": 0.463,
+      "step": 2048
+    },
+    {
+      "epoch": 1.0421514400152585,
+      "grad_norm": 1.9324961620060757,
+      "learning_rate": 1.5541416660989835e-05,
+      "loss": 0.5661,
+      "step": 2049
+    },
+    {
+      "epoch": 1.042660054676076,
+      "grad_norm": 1.9905685728210991,
+      "learning_rate": 1.5536748411937816e-05,
+      "loss": 0.4323,
+      "step": 2050
+    },
+    {
+      "epoch": 1.0431686693368936,
+      "grad_norm": 2.518633068768695,
+      "learning_rate": 1.5532078422228482e-05,
+      "loss": 0.5947,
+      "step": 2051
+    },
+    {
+      "epoch": 1.0436772839977113,
+      "grad_norm": 1.995504213830011,
+      "learning_rate": 1.552740669333e-05,
+      "loss": 0.5504,
+      "step": 2052
+    },
+    {
+      "epoch": 1.0441858986585288,
+      "grad_norm": 2.205094127073248,
+      "learning_rate": 1.552273322671108e-05,
+      "loss": 0.5365,
+      "step": 2053
+    },
+    {
+      "epoch": 1.0446945133193464,
+      "grad_norm": 1.9752115450206866,
+      "learning_rate": 1.5518058023840974e-05,
+      "loss": 0.5281,
+      "step": 2054
+    },
+    {
+      "epoch": 1.0452031279801641,
+      "grad_norm": 1.9299924663535022,
+      "learning_rate": 1.5513381086189493e-05,
+      "loss": 0.5284,
+      "step": 2055
+    },
+    {
+      "epoch": 1.0457117426409817,
+      "grad_norm": 2.4132641703704243,
+      "learning_rate": 1.550870241522698e-05,
+      "loss": 0.5647,
+      "step": 2056
+    },
+    {
+      "epoch": 1.0462203573017992,
+      "grad_norm": 2.1228257443989045,
+      "learning_rate": 1.5504022012424326e-05,
+      "loss": 0.5095,
+      "step": 2057
+    },
+    {
+      "epoch": 1.0467289719626167,
+      "grad_norm": 2.436379590881572,
+      "learning_rate": 1.5499339879252972e-05,
+      "loss": 0.5423,
+      "step": 2058
+    },
+    {
+      "epoch": 1.0472375866234345,
+      "grad_norm": 2.08067326158291,
+      "learning_rate": 1.5494656017184898e-05,
+      "loss": 0.5062,
+      "step": 2059
+    },
+    {
+      "epoch": 1.047746201284252,
+      "grad_norm": 1.9844927414934121,
+      "learning_rate": 1.5489970427692624e-05,
+      "loss": 0.5163,
+      "step": 2060
+    },
+    {
+      "epoch": 1.0482548159450695,
+      "grad_norm": 2.1692974795704267,
+      "learning_rate": 1.5485283112249225e-05,
+      "loss": 0.58,
+      "step": 2061
+    },
+    {
+      "epoch": 1.0487634306058873,
+      "grad_norm": 2.443047104845533,
+      "learning_rate": 1.5480594072328305e-05,
+      "loss": 0.5158,
+      "step": 2062
+    },
+    {
+      "epoch": 1.0492720452667048,
+      "grad_norm": 2.0064482451322503,
+      "learning_rate": 1.5475903309404022e-05,
+      "loss": 0.4959,
+      "step": 2063
+    },
+    {
+      "epoch": 1.0497806599275223,
+      "grad_norm": 2.040777778456057,
+      "learning_rate": 1.5471210824951072e-05,
+      "loss": 0.4923,
+      "step": 2064
+    },
+    {
+      "epoch": 1.05028927458834,
+      "grad_norm": 2.089414976823306,
+      "learning_rate": 1.5466516620444677e-05,
+      "loss": 0.5734,
+      "step": 2065
+    },
+    {
+      "epoch": 1.0507978892491576,
+      "grad_norm": 2.23862272257537,
+      "learning_rate": 1.546182069736063e-05,
+      "loss": 0.5134,
+      "step": 2066
+    },
+    {
+      "epoch": 1.0513065039099752,
+      "grad_norm": 2.1617173250370825,
+      "learning_rate": 1.5457123057175237e-05,
+      "loss": 0.5724,
+      "step": 2067
+    },
+    {
+      "epoch": 1.051815118570793,
+      "grad_norm": 1.9562846756564476,
+      "learning_rate": 1.5452423701365355e-05,
+      "loss": 0.5201,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0523237332316104,
+      "grad_norm": 2.180627982306197,
+      "learning_rate": 1.5447722631408382e-05,
+      "loss": 0.5123,
+      "step": 2069
+    },
+    {
+      "epoch": 1.052832347892428,
+      "grad_norm": 1.9898179267133278,
+      "learning_rate": 1.5443019848782256e-05,
+      "loss": 0.4548,
+      "step": 2070
+    },
+    {
+      "epoch": 1.0533409625532455,
+      "grad_norm": 1.8887481200677794,
+      "learning_rate": 1.5438315354965445e-05,
+      "loss": 0.5038,
+      "step": 2071
+    },
+    {
+      "epoch": 1.0538495772140632,
+      "grad_norm": 2.002177364537458,
+      "learning_rate": 1.543360915143696e-05,
+      "loss": 0.5014,
+      "step": 2072
+    },
+    {
+      "epoch": 1.0543581918748808,
+      "grad_norm": 2.1628260195843794,
+      "learning_rate": 1.542890123967635e-05,
+      "loss": 0.4383,
+      "step": 2073
+    },
+    {
+      "epoch": 1.0548668065356983,
+      "grad_norm": 2.526093965493695,
+      "learning_rate": 1.5424191621163703e-05,
+      "loss": 0.5045,
+      "step": 2074
+    },
+    {
+      "epoch": 1.055375421196516,
+      "grad_norm": 2.1534890452598545,
+      "learning_rate": 1.541948029737964e-05,
+      "loss": 0.4929,
+      "step": 2075
+    },
+    {
+      "epoch": 1.0558840358573336,
+      "grad_norm": 1.9797317367464717,
+      "learning_rate": 1.5414767269805317e-05,
+      "loss": 0.4485,
+      "step": 2076
+    },
+    {
+      "epoch": 1.0563926505181511,
+      "grad_norm": 2.0710099107255586,
+      "learning_rate": 1.541005253992243e-05,
+      "loss": 0.5931,
+      "step": 2077
+    },
+    {
+      "epoch": 1.0569012651789689,
+      "grad_norm": 2.3508550143625335,
+      "learning_rate": 1.5405336109213208e-05,
+      "loss": 0.5796,
+      "step": 2078
+    },
+    {
+      "epoch": 1.0574098798397864,
+      "grad_norm": 2.586026377461521,
+      "learning_rate": 1.5400617979160413e-05,
+      "loss": 0.5377,
+      "step": 2079
+    },
+    {
+      "epoch": 1.057918494500604,
+      "grad_norm": 1.98811336625457,
+      "learning_rate": 1.5395898151247342e-05,
+      "loss": 0.4692,
+      "step": 2080
+    },
+    {
+      "epoch": 1.0584271091614217,
+      "grad_norm": 2.13089047976553,
+      "learning_rate": 1.5391176626957833e-05,
+      "loss": 0.5186,
+      "step": 2081
+    },
+    {
+      "epoch": 1.0589357238222392,
+      "grad_norm": 1.8954697005200583,
+      "learning_rate": 1.5386453407776247e-05,
+      "loss": 0.5048,
+      "step": 2082
+    },
+    {
+      "epoch": 1.0594443384830567,
+      "grad_norm": 1.921294085779242,
+      "learning_rate": 1.5381728495187477e-05,
+      "loss": 0.4052,
+      "step": 2083
+    },
+    {
+      "epoch": 1.0599529531438743,
+      "grad_norm": 2.1274612982666827,
+      "learning_rate": 1.537700189067696e-05,
+      "loss": 0.5352,
+      "step": 2084
+    },
+    {
+      "epoch": 1.060461567804692,
+      "grad_norm": 1.8594309694217088,
+      "learning_rate": 1.5372273595730657e-05,
+      "loss": 0.5607,
+      "step": 2085
+    },
+    {
+      "epoch": 1.0609701824655096,
+      "grad_norm": 2.226409799793976,
+      "learning_rate": 1.536754361183506e-05,
+      "loss": 0.5522,
+      "step": 2086
+    },
+    {
+      "epoch": 1.061478797126327,
+      "grad_norm": 2.185343305909865,
+      "learning_rate": 1.5362811940477193e-05,
+      "loss": 0.5931,
+      "step": 2087
+    },
+    {
+      "epoch": 1.0619874117871448,
+      "grad_norm": 1.9180951669211972,
+      "learning_rate": 1.5358078583144614e-05,
+      "loss": 0.5324,
+      "step": 2088
+    },
+    {
+      "epoch": 1.0624960264479624,
+      "grad_norm": 2.3445877612433534,
+      "learning_rate": 1.5353343541325406e-05,
+      "loss": 0.4919,
+      "step": 2089
+    },
+    {
+      "epoch": 1.06300464110878,
+      "grad_norm": 2.014131569399407,
+      "learning_rate": 1.5348606816508188e-05,
+      "loss": 0.5507,
+      "step": 2090
+    },
+    {
+      "epoch": 1.0635132557695977,
+      "grad_norm": 2.3840973938690864,
+      "learning_rate": 1.5343868410182094e-05,
+      "loss": 0.4808,
+      "step": 2091
+    },
+    {
+      "epoch": 1.0640218704304152,
+      "grad_norm": 1.9888899495330956,
+      "learning_rate": 1.5339128323836806e-05,
+      "loss": 0.5266,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0645304850912327,
+      "grad_norm": 2.024991061343963,
+      "learning_rate": 1.533438655896252e-05,
+      "loss": 0.476,
+      "step": 2093
+    },
+    {
+      "epoch": 1.0650390997520502,
+      "grad_norm": 2.3771848166019223,
+      "learning_rate": 1.5329643117049966e-05,
+      "loss": 0.5538,
+      "step": 2094
+    },
+    {
+      "epoch": 1.065547714412868,
+      "grad_norm": 2.0724123112772106,
+      "learning_rate": 1.53248979995904e-05,
+      "loss": 0.4835,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0660563290736855,
+      "grad_norm": 1.9739198773210451,
+      "learning_rate": 1.53201512080756e-05,
+      "loss": 0.4862,
+      "step": 2096
+    },
+    {
+      "epoch": 1.066564943734503,
+      "grad_norm": 1.82460615794397,
+      "learning_rate": 1.531540274399788e-05,
+      "loss": 0.47,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0670735583953208,
+      "grad_norm": 2.140897858764541,
+      "learning_rate": 1.531065260885007e-05,
+      "loss": 0.5118,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0675821730561383,
+      "grad_norm": 2.09288029613735,
+      "learning_rate": 1.5305900804125536e-05,
+      "loss": 0.5586,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0680907877169559,
+      "grad_norm": 2.0074597950634994,
+      "learning_rate": 1.5301147331318157e-05,
+      "loss": 0.5117,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0680907877169559,
+      "eval_loss": 0.7263098359107971,
+      "eval_runtime": 169.9932,
+      "eval_samples_per_second": 19.483,
+      "eval_steps_per_second": 9.742,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0685994023777736,
+      "grad_norm": 2.0865847837004643,
+      "learning_rate": 1.5296392191922348e-05,
+      "loss": 0.6245,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0691080170385912,
+      "grad_norm": 1.9670601510420018,
+      "learning_rate": 1.529163538743303e-05,
+      "loss": 0.4579,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0696166316994087,
+      "grad_norm": 2.213652652127253,
+      "learning_rate": 1.5286876919345673e-05,
+      "loss": 0.5449,
+      "step": 2103
+    },
+    {
+      "epoch": 1.0701252463602264,
+      "grad_norm": 2.0269851514425805,
+      "learning_rate": 1.528211678915625e-05,
+      "loss": 0.4597,
+      "step": 2104
+    },
+    {
+      "epoch": 1.070633861021044,
+      "grad_norm": 1.9288038941206795,
+      "learning_rate": 1.5277354998361263e-05,
+      "loss": 0.5024,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0711424756818615,
+      "grad_norm": 2.432080136453282,
+      "learning_rate": 1.527259154845774e-05,
+      "loss": 0.5253,
+      "step": 2106
+    },
+    {
+      "epoch": 1.071651090342679,
+      "grad_norm": 1.9561610923770654,
+      "learning_rate": 1.5267826440943222e-05,
+      "loss": 0.5473,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0721597050034968,
+      "grad_norm": 1.9420926820499573,
+      "learning_rate": 1.526305967731578e-05,
+      "loss": 0.5063,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0726683196643143,
+      "grad_norm": 2.191789399785083,
+      "learning_rate": 1.5258291259073997e-05,
+      "loss": 0.5071,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0731769343251318,
+      "grad_norm": 2.0718097383050984,
+      "learning_rate": 1.5253521187716982e-05,
+      "loss": 0.5776,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0736855489859496,
+      "grad_norm": 2.2396885573633614,
+      "learning_rate": 1.5248749464744365e-05,
+      "loss": 0.453,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0741941636467671,
+      "grad_norm": 2.1430858194858637,
+      "learning_rate": 1.5243976091656291e-05,
+      "loss": 0.4473,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0747027783075846,
+      "grad_norm": 2.288212328622096,
+      "learning_rate": 1.5239201069953428e-05,
+      "loss": 0.6087,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0752113929684024,
+      "grad_norm": 1.926149963613236,
+      "learning_rate": 1.5234424401136955e-05,
+      "loss": 0.4581,
+      "step": 2114
+    },
+    {
+      "epoch": 1.07572000762922,
+      "grad_norm": 2.113987830986864,
+      "learning_rate": 1.5229646086708574e-05,
+      "loss": 0.5213,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0762286222900375,
+      "grad_norm": 2.005688445507401,
+      "learning_rate": 1.5224866128170507e-05,
+      "loss": 0.5421,
+      "step": 2116
+    },
+    {
+      "epoch": 1.076737236950855,
+      "grad_norm": 2.0449167450090857,
+      "learning_rate": 1.5220084527025489e-05,
+      "loss": 0.4909,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0772458516116727,
+      "grad_norm": 2.137677021392738,
+      "learning_rate": 1.5215301284776774e-05,
+      "loss": 0.5408,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0777544662724903,
+      "grad_norm": 2.3980769280891403,
+      "learning_rate": 1.5210516402928121e-05,
+      "loss": 0.6444,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0782630809333078,
+      "grad_norm": 2.2587446248803835,
+      "learning_rate": 1.5205729882983823e-05,
+      "loss": 0.5169,
+      "step": 2120
+    },
+    {
+      "epoch": 1.0787716955941256,
+      "grad_norm": 2.2617908525586032,
+      "learning_rate": 1.5200941726448677e-05,
+      "loss": 0.5186,
+      "step": 2121
+    },
+    {
+      "epoch": 1.079280310254943,
+      "grad_norm": 2.8696685783080618,
+      "learning_rate": 1.5196151934827991e-05,
+      "loss": 0.5685,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0797889249157606,
+      "grad_norm": 2.0982016310846485,
+      "learning_rate": 1.51913605096276e-05,
+      "loss": 0.4899,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0802975395765784,
+      "grad_norm": 1.8605112402765303,
+      "learning_rate": 1.5186567452353838e-05,
+      "loss": 0.49,
+      "step": 2124
+    },
+    {
+      "epoch": 1.080806154237396,
+      "grad_norm": 2.1717222969656915,
+      "learning_rate": 1.5181772764513562e-05,
+      "loss": 0.5852,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0813147688982134,
+      "grad_norm": 2.2785099243322953,
+      "learning_rate": 1.517697644761414e-05,
+      "loss": 0.4867,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0818233835590312,
+      "grad_norm": 2.0429798366617797,
+      "learning_rate": 1.5172178503163446e-05,
+      "loss": 0.5417,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0823319982198487,
+      "grad_norm": 2.1272423077344595,
+      "learning_rate": 1.5167378932669872e-05,
+      "loss": 0.4427,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0828406128806662,
+      "grad_norm": 1.9907943940961907,
+      "learning_rate": 1.5162577737642318e-05,
+      "loss": 0.3854,
+      "step": 2129
+    },
+    {
+      "epoch": 1.083349227541484,
+      "grad_norm": 2.366891168861977,
+      "learning_rate": 1.5157774919590203e-05,
+      "loss": 0.5136,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0838578422023015,
+      "grad_norm": 1.9925111993836295,
+      "learning_rate": 1.5152970480023441e-05,
+      "loss": 0.5659,
+      "step": 2131
+    },
+    {
+      "epoch": 1.084366456863119,
+      "grad_norm": 2.228700167130393,
+      "learning_rate": 1.5148164420452469e-05,
+      "loss": 0.5278,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0848750715239366,
+      "grad_norm": 2.0890004961260638,
+      "learning_rate": 1.5143356742388225e-05,
+      "loss": 0.5356,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0853836861847543,
+      "grad_norm": 1.9583086729187906,
+      "learning_rate": 1.513854744734216e-05,
+      "loss": 0.466,
+      "step": 2134
+    },
+    {
+      "epoch": 1.0858923008455719,
+      "grad_norm": 1.9975349016446682,
+      "learning_rate": 1.5133736536826235e-05,
+      "loss": 0.474,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0864009155063894,
+      "grad_norm": 2.159151861252175,
+      "learning_rate": 1.5128924012352918e-05,
+      "loss": 0.5965,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0869095301672071,
+      "grad_norm": 2.4309996959038016,
+      "learning_rate": 1.5124109875435178e-05,
+      "loss": 0.4859,
+      "step": 2137
+    },
+    {
+      "epoch": 1.0874181448280247,
+      "grad_norm": 2.0032332035542604,
+      "learning_rate": 1.5119294127586497e-05,
+      "loss": 0.6324,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0879267594888422,
+      "grad_norm": 1.887246004913706,
+      "learning_rate": 1.5114476770320865e-05,
+      "loss": 0.4278,
+      "step": 2139
+    },
+    {
+      "epoch": 1.08843537414966,
+      "grad_norm": 2.0884624158579372,
+      "learning_rate": 1.5109657805152775e-05,
+      "loss": 0.4578,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0889439888104775,
+      "grad_norm": 1.9866461785217955,
+      "learning_rate": 1.5104837233597222e-05,
+      "loss": 0.5279,
+      "step": 2141
+    },
+    {
+      "epoch": 1.089452603471295,
+      "grad_norm": 2.104682959740718,
+      "learning_rate": 1.510001505716971e-05,
+      "loss": 0.5123,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0899612181321126,
+      "grad_norm": 1.7777327897750637,
+      "learning_rate": 1.5095191277386254e-05,
+      "loss": 0.4462,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0904698327929303,
+      "grad_norm": 2.1980912077483423,
+      "learning_rate": 1.5090365895763362e-05,
+      "loss": 0.5558,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0909784474537478,
+      "grad_norm": 2.068024951385849,
+      "learning_rate": 1.5085538913818048e-05,
+      "loss": 0.4579,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0914870621145654,
+      "grad_norm": 2.0859203552227297,
+      "learning_rate": 1.5080710333067834e-05,
+      "loss": 0.5584,
+      "step": 2146
+    },
+    {
+      "epoch": 1.0919956767753831,
+      "grad_norm": 2.45520485343789,
+      "learning_rate": 1.5075880155030737e-05,
+      "loss": 0.6154,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0925042914362006,
+      "grad_norm": 2.2439389715072737,
+      "learning_rate": 1.5071048381225289e-05,
+      "loss": 0.5349,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0930129060970182,
+      "grad_norm": 2.0190383043892615,
+      "learning_rate": 1.5066215013170511e-05,
+      "loss": 0.4556,
+      "step": 2149
+    },
+    {
+      "epoch": 1.093521520757836,
+      "grad_norm": 1.9657464503867867,
+      "learning_rate": 1.5061380052385928e-05,
+      "loss": 0.4479,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0940301354186535,
+      "grad_norm": 1.7687423328462633,
+      "learning_rate": 1.505654350039157e-05,
+      "loss": 0.5405,
+      "step": 2151
+    },
+    {
+      "epoch": 1.094538750079471,
+      "grad_norm": 1.8633275696057394,
+      "learning_rate": 1.5051705358707964e-05,
+      "loss": 0.4263,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0950473647402887,
+      "grad_norm": 1.9288003712453212,
+      "learning_rate": 1.5046865628856136e-05,
+      "loss": 0.5031,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0955559794011063,
+      "grad_norm": 2.0393786422785816,
+      "learning_rate": 1.5042024312357618e-05,
+      "loss": 0.5554,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0960645940619238,
+      "grad_norm": 2.1096302019378235,
+      "learning_rate": 1.503718141073443e-05,
+      "loss": 0.5214,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0965732087227413,
+      "grad_norm": 2.043081481040682,
+      "learning_rate": 1.5032336925509095e-05,
+      "loss": 0.4126,
+      "step": 2156
+    },
+    {
+      "epoch": 1.097081823383559,
+      "grad_norm": 1.8688938532037733,
+      "learning_rate": 1.5027490858204643e-05,
+      "loss": 0.5278,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0975904380443766,
+      "grad_norm": 2.2593059946077485,
+      "learning_rate": 1.5022643210344585e-05,
+      "loss": 0.5334,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0980990527051941,
+      "grad_norm": 1.8866891799410266,
+      "learning_rate": 1.5017793983452941e-05,
+      "loss": 0.4601,
+      "step": 2159
+    },
+    {
+      "epoch": 1.098607667366012,
+      "grad_norm": 1.9082093721910276,
+      "learning_rate": 1.5012943179054224e-05,
+      "loss": 0.5361,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0991162820268294,
+      "grad_norm": 2.0253912135742276,
+      "learning_rate": 1.500809079867344e-05,
+      "loss": 0.5355,
+      "step": 2161
+    },
+    {
+      "epoch": 1.099624896687647,
+      "grad_norm": 1.886682482384664,
+      "learning_rate": 1.5003236843836098e-05,
+      "loss": 0.4741,
+      "step": 2162
+    },
+    {
+      "epoch": 1.1001335113484647,
+      "grad_norm": 2.133313354259556,
+      "learning_rate": 1.499838131606819e-05,
+      "loss": 0.5153,
+      "step": 2163
+    },
+    {
+      "epoch": 1.1006421260092822,
+      "grad_norm": 2.0660944741317127,
+      "learning_rate": 1.4993524216896209e-05,
+      "loss": 0.4703,
+      "step": 2164
+    },
+    {
+      "epoch": 1.1011507406700998,
+      "grad_norm": 2.1258494979182023,
+      "learning_rate": 1.4988665547847148e-05,
+      "loss": 0.4674,
+      "step": 2165
+    },
+    {
+      "epoch": 1.1016593553309173,
+      "grad_norm": 2.298447155235233,
+      "learning_rate": 1.4983805310448487e-05,
+      "loss": 0.589,
+      "step": 2166
+    },
+    {
+      "epoch": 1.102167969991735,
+      "grad_norm": 2.1156597185314796,
+      "learning_rate": 1.4978943506228198e-05,
+      "loss": 0.5718,
+      "step": 2167
+    },
+    {
+      "epoch": 1.1026765846525526,
+      "grad_norm": 1.6934051331494204,
+      "learning_rate": 1.4974080136714748e-05,
+      "loss": 0.4017,
+      "step": 2168
+    },
+    {
+      "epoch": 1.1031851993133701,
+      "grad_norm": 2.2187079842041273,
+      "learning_rate": 1.4969215203437097e-05,
+      "loss": 0.4303,
+      "step": 2169
+    },
+    {
+      "epoch": 1.1036938139741879,
+      "grad_norm": 2.0165651479763094,
+      "learning_rate": 1.4964348707924694e-05,
+      "loss": 0.4712,
+      "step": 2170
+    },
+    {
+      "epoch": 1.1042024286350054,
+      "grad_norm": 1.8817913404484612,
+      "learning_rate": 1.4959480651707482e-05,
+      "loss": 0.4345,
+      "step": 2171
+    },
+    {
+      "epoch": 1.104711043295823,
+      "grad_norm": 2.035610741298024,
+      "learning_rate": 1.495461103631589e-05,
+      "loss": 0.4877,
+      "step": 2172
+    },
+    {
+      "epoch": 1.1052196579566407,
+      "grad_norm": 2.0888876180554754,
+      "learning_rate": 1.4949739863280844e-05,
+      "loss": 0.4765,
+      "step": 2173
+    },
+    {
+      "epoch": 1.1057282726174582,
+      "grad_norm": 2.133143137360973,
+      "learning_rate": 1.4944867134133752e-05,
+      "loss": 0.4925,
+      "step": 2174
+    },
+    {
+      "epoch": 1.1062368872782757,
+      "grad_norm": 2.257656422008545,
+      "learning_rate": 1.493999285040652e-05,
+      "loss": 0.514,
+      "step": 2175
+    },
+    {
+      "epoch": 1.1067455019390935,
+      "grad_norm": 2.08562475710311,
+      "learning_rate": 1.4935117013631535e-05,
+      "loss": 0.5403,
+      "step": 2176
+    },
+    {
+      "epoch": 1.107254116599911,
+      "grad_norm": 2.1087994956023173,
+      "learning_rate": 1.4930239625341675e-05,
+      "loss": 0.5244,
+      "step": 2177
+    },
+    {
+      "epoch": 1.1077627312607286,
+      "grad_norm": 2.0624653206423953,
+      "learning_rate": 1.4925360687070308e-05,
+      "loss": 0.5537,
+      "step": 2178
+    },
+    {
+      "epoch": 1.1082713459215463,
+      "grad_norm": 1.8704384311229232,
+      "learning_rate": 1.4920480200351284e-05,
+      "loss": 0.4948,
+      "step": 2179
+    },
+    {
+      "epoch": 1.1087799605823638,
+      "grad_norm": 1.8193605086530176,
+      "learning_rate": 1.4915598166718945e-05,
+      "loss": 0.451,
+      "step": 2180
+    },
+    {
+      "epoch": 1.1092885752431814,
+      "grad_norm": 1.8916810924411303,
+      "learning_rate": 1.491071458770812e-05,
+      "loss": 0.5426,
+      "step": 2181
+    },
+    {
+      "epoch": 1.109797189903999,
+      "grad_norm": 2.1920637208628864,
+      "learning_rate": 1.4905829464854117e-05,
+      "loss": 0.4896,
+      "step": 2182
+    },
+    {
+      "epoch": 1.1103058045648166,
+      "grad_norm": 2.0236465169403726,
+      "learning_rate": 1.4900942799692735e-05,
+      "loss": 0.5748,
+      "step": 2183
+    },
+    {
+      "epoch": 1.1108144192256342,
+      "grad_norm": 2.021657978656631,
+      "learning_rate": 1.4896054593760256e-05,
+      "loss": 0.5218,
+      "step": 2184
+    },
+    {
+      "epoch": 1.1113230338864517,
+      "grad_norm": 2.478016615971237,
+      "learning_rate": 1.4891164848593448e-05,
+      "loss": 0.4603,
+      "step": 2185
+    },
+    {
+      "epoch": 1.1118316485472695,
+      "grad_norm": 1.9574839055165805,
+      "learning_rate": 1.4886273565729563e-05,
+      "loss": 0.5219,
+      "step": 2186
+    },
+    {
+      "epoch": 1.112340263208087,
+      "grad_norm": 1.9272042794979367,
+      "learning_rate": 1.4881380746706331e-05,
+      "loss": 0.5322,
+      "step": 2187
+    },
+    {
+      "epoch": 1.1128488778689045,
+      "grad_norm": 1.934077778353851,
+      "learning_rate": 1.4876486393061972e-05,
+      "loss": 0.4883,
+      "step": 2188
+    },
+    {
+      "epoch": 1.1133574925297223,
+      "grad_norm": 1.9939680333038252,
+      "learning_rate": 1.4871590506335185e-05,
+      "loss": 0.5445,
+      "step": 2189
+    },
+    {
+      "epoch": 1.1138661071905398,
+      "grad_norm": 2.7933130865645466,
+      "learning_rate": 1.4866693088065152e-05,
+      "loss": 0.4527,
+      "step": 2190
+    },
+    {
+      "epoch": 1.1143747218513573,
+      "grad_norm": 1.8889007271913043,
+      "learning_rate": 1.4861794139791537e-05,
+      "loss": 0.4669,
+      "step": 2191
+    },
+    {
+      "epoch": 1.1148833365121749,
+      "grad_norm": 1.8102510245533248,
+      "learning_rate": 1.485689366305448e-05,
+      "loss": 0.4687,
+      "step": 2192
+    },
+    {
+      "epoch": 1.1153919511729926,
+      "grad_norm": 2.2261364981040397,
+      "learning_rate": 1.4851991659394609e-05,
+      "loss": 0.492,
+      "step": 2193
+    },
+    {
+      "epoch": 1.1159005658338101,
+      "grad_norm": 2.051446367217725,
+      "learning_rate": 1.4847088130353032e-05,
+      "loss": 0.4857,
+      "step": 2194
+    },
+    {
+      "epoch": 1.1164091804946277,
+      "grad_norm": 1.9384558381094625,
+      "learning_rate": 1.4842183077471324e-05,
+      "loss": 0.5231,
+      "step": 2195
+    },
+    {
+      "epoch": 1.1169177951554454,
+      "grad_norm": 2.5632347356173777,
+      "learning_rate": 1.4837276502291556e-05,
+      "loss": 0.4665,
+      "step": 2196
+    },
+    {
+      "epoch": 1.117426409816263,
+      "grad_norm": 2.200122166622841,
+      "learning_rate": 1.4832368406356265e-05,
+      "loss": 0.4794,
+      "step": 2197
+    },
+    {
+      "epoch": 1.1179350244770805,
+      "grad_norm": 2.1039503634729617,
+      "learning_rate": 1.4827458791208475e-05,
+      "loss": 0.5043,
+      "step": 2198
+    },
+    {
+      "epoch": 1.1184436391378982,
+      "grad_norm": 2.4163712507316224,
+      "learning_rate": 1.4822547658391684e-05,
+      "loss": 0.4385,
+      "step": 2199
+    },
+    {
+      "epoch": 1.1189522537987158,
+      "grad_norm": 2.1525790160382234,
+      "learning_rate": 1.481763500944986e-05,
+      "loss": 0.4997,
+      "step": 2200
+    },
+    {
+      "epoch": 1.1189522537987158,
+      "eval_loss": 0.7296145558357239,
+      "eval_runtime": 169.4634,
+      "eval_samples_per_second": 19.544,
+      "eval_steps_per_second": 9.772,
+      "step": 2200
+    },
+    {
+      "epoch": 1.1194608684595333,
+      "grad_norm": 2.4066034032726105,
+      "learning_rate": 1.481272084592746e-05,
+      "loss": 0.5293,
+      "step": 2201
+    },
+    {
+      "epoch": 1.119969483120351,
+      "grad_norm": 2.077529477714799,
+      "learning_rate": 1.480780516936941e-05,
+      "loss": 0.5106,
+      "step": 2202
+    },
+    {
+      "epoch": 1.1204780977811686,
+      "grad_norm": 2.2130389790918135,
+      "learning_rate": 1.4802887981321112e-05,
+      "loss": 0.4924,
+      "step": 2203
+    },
+    {
+      "epoch": 1.120986712441986,
+      "grad_norm": 1.774192414115588,
+      "learning_rate": 1.4797969283328448e-05,
+      "loss": 0.4892,
+      "step": 2204
+    },
+    {
+      "epoch": 1.1214953271028036,
+      "grad_norm": 2.2017198775756017,
+      "learning_rate": 1.4793049076937769e-05,
+      "loss": 0.446,
+      "step": 2205
+    },
+    {
+      "epoch": 1.1220039417636214,
+      "grad_norm": 2.245877728232799,
+      "learning_rate": 1.4788127363695897e-05,
+      "loss": 0.4748,
+      "step": 2206
+    },
+    {
+      "epoch": 1.122512556424439,
+      "grad_norm": 1.9892600897062935,
+      "learning_rate": 1.4783204145150143e-05,
+      "loss": 0.4942,
+      "step": 2207
+    },
+    {
+      "epoch": 1.1230211710852565,
+      "grad_norm": 2.438831552523309,
+      "learning_rate": 1.4778279422848271e-05,
+      "loss": 0.569,
+      "step": 2208
+    },
+    {
+      "epoch": 1.1235297857460742,
+      "grad_norm": 2.2642725623953504,
+      "learning_rate": 1.4773353198338538e-05,
+      "loss": 0.4447,
+      "step": 2209
+    },
+    {
+      "epoch": 1.1240384004068917,
+      "grad_norm": 1.8618091710167899,
+      "learning_rate": 1.4768425473169654e-05,
+      "loss": 0.5401,
+      "step": 2210
+    },
+    {
+      "epoch": 1.1245470150677093,
+      "grad_norm": 2.130508351980838,
+      "learning_rate": 1.4763496248890817e-05,
+      "loss": 0.5248,
+      "step": 2211
+    },
+    {
+      "epoch": 1.125055629728527,
+      "grad_norm": 2.1669719432076007,
+      "learning_rate": 1.4758565527051685e-05,
+      "loss": 0.5191,
+      "step": 2212
+    },
+    {
+      "epoch": 1.1255642443893445,
+      "grad_norm": 1.8679436465386976,
+      "learning_rate": 1.4753633309202394e-05,
+      "loss": 0.4344,
+      "step": 2213
+    },
+    {
+      "epoch": 1.126072859050162,
+      "grad_norm": 2.1233599448619955,
+      "learning_rate": 1.4748699596893542e-05,
+      "loss": 0.5299,
+      "step": 2214
+    },
+    {
+      "epoch": 1.1265814737109796,
+      "grad_norm": 1.8957025112526051,
+      "learning_rate": 1.474376439167621e-05,
+      "loss": 0.564,
+      "step": 2215
+    },
+    {
+      "epoch": 1.1270900883717974,
+      "grad_norm": 2.0313969485607855,
+      "learning_rate": 1.4738827695101936e-05,
+      "loss": 0.5317,
+      "step": 2216
+    },
+    {
+      "epoch": 1.127598703032615,
+      "grad_norm": 3.022360124277912,
+      "learning_rate": 1.4733889508722734e-05,
+      "loss": 0.5797,
+      "step": 2217
+    },
+    {
+      "epoch": 1.1281073176934324,
+      "grad_norm": 1.8245681983451858,
+      "learning_rate": 1.4728949834091078e-05,
+      "loss": 0.429,
+      "step": 2218
+    },
+    {
+      "epoch": 1.1286159323542502,
+      "grad_norm": 2.0519960129905472,
+      "learning_rate": 1.4724008672759922e-05,
+      "loss": 0.5406,
+      "step": 2219
+    },
+    {
+      "epoch": 1.1291245470150677,
+      "grad_norm": 2.048460251880406,
+      "learning_rate": 1.4719066026282678e-05,
+      "loss": 0.561,
+      "step": 2220
+    },
+    {
+      "epoch": 1.1296331616758852,
+      "grad_norm": 1.8622228147356057,
+      "learning_rate": 1.471412189621323e-05,
+      "loss": 0.4869,
+      "step": 2221
+    },
+    {
+      "epoch": 1.130141776336703,
+      "grad_norm": 2.6331829410374166,
+      "learning_rate": 1.4709176284105925e-05,
+      "loss": 0.5143,
+      "step": 2222
+    },
+    {
+      "epoch": 1.1306503909975205,
+      "grad_norm": 2.3005339873368724,
+      "learning_rate": 1.4704229191515577e-05,
+      "loss": 0.5445,
+      "step": 2223
+    },
+    {
+      "epoch": 1.131159005658338,
+      "grad_norm": 1.8802498119847666,
+      "learning_rate": 1.4699280619997467e-05,
+      "loss": 0.5128,
+      "step": 2224
+    },
+    {
+      "epoch": 1.1316676203191558,
+      "grad_norm": 1.7628005400551008,
+      "learning_rate": 1.4694330571107338e-05,
+      "loss": 0.4158,
+      "step": 2225
+    },
+    {
+      "epoch": 1.1321762349799733,
+      "grad_norm": 1.9141643886306265,
+      "learning_rate": 1.4689379046401405e-05,
+      "loss": 0.5136,
+      "step": 2226
+    },
+    {
+      "epoch": 1.1326848496407909,
+      "grad_norm": 2.0923949454311956,
+      "learning_rate": 1.4684426047436331e-05,
+      "loss": 0.4953,
+      "step": 2227
+    },
+    {
+      "epoch": 1.1331934643016086,
+      "grad_norm": 2.3545970917111134,
+      "learning_rate": 1.4679471575769261e-05,
+      "loss": 0.5883,
+      "step": 2228
+    },
+    {
+      "epoch": 1.1337020789624261,
+      "grad_norm": 2.0232020649861346,
+      "learning_rate": 1.4674515632957792e-05,
+      "loss": 0.483,
+      "step": 2229
+    },
+    {
+      "epoch": 1.1342106936232437,
+      "grad_norm": 1.8706246506741315,
+      "learning_rate": 1.4669558220559988e-05,
+      "loss": 0.5198,
+      "step": 2230
+    },
+    {
+      "epoch": 1.1347193082840612,
+      "grad_norm": 1.931605200833231,
+      "learning_rate": 1.4664599340134366e-05,
+      "loss": 0.5315,
+      "step": 2231
+    },
+    {
+      "epoch": 1.135227922944879,
+      "grad_norm": 2.0499356515600944,
+      "learning_rate": 1.465963899323992e-05,
+      "loss": 0.4723,
+      "step": 2232
+    },
+    {
+      "epoch": 1.1357365376056965,
+      "grad_norm": 1.9510427743474448,
+      "learning_rate": 1.4654677181436093e-05,
+      "loss": 0.4752,
+      "step": 2233
+    },
+    {
+      "epoch": 1.136245152266514,
+      "grad_norm": 2.302526048809076,
+      "learning_rate": 1.4649713906282792e-05,
+      "loss": 0.5215,
+      "step": 2234
+    },
+    {
+      "epoch": 1.1367537669273318,
+      "grad_norm": 2.147009690680841,
+      "learning_rate": 1.4644749169340387e-05,
+      "loss": 0.5199,
+      "step": 2235
+    },
+    {
+      "epoch": 1.1372623815881493,
+      "grad_norm": 2.0645447045750114,
+      "learning_rate": 1.4639782972169703e-05,
+      "loss": 0.5316,
+      "step": 2236
+    },
+    {
+      "epoch": 1.1377709962489668,
+      "grad_norm": 2.0195337291827746,
+      "learning_rate": 1.4634815316332023e-05,
+      "loss": 0.4128,
+      "step": 2237
+    },
+    {
+      "epoch": 1.1382796109097844,
+      "grad_norm": 2.451517343682674,
+      "learning_rate": 1.4629846203389098e-05,
+      "loss": 0.4797,
+      "step": 2238
+    },
+    {
+      "epoch": 1.138788225570602,
+      "grad_norm": 2.399387034977001,
+      "learning_rate": 1.4624875634903123e-05,
+      "loss": 0.5746,
+      "step": 2239
+    },
+    {
+      "epoch": 1.1392968402314196,
+      "grad_norm": 2.205040930543262,
+      "learning_rate": 1.4619903612436766e-05,
+      "loss": 0.4721,
+      "step": 2240
+    },
+    {
+      "epoch": 1.1398054548922372,
+      "grad_norm": 2.142416879294569,
+      "learning_rate": 1.4614930137553138e-05,
+      "loss": 0.4866,
+      "step": 2241
+    },
+    {
+      "epoch": 1.140314069553055,
+      "grad_norm": 2.0300397348508334,
+      "learning_rate": 1.4609955211815816e-05,
+      "loss": 0.4088,
+      "step": 2242
+    },
+    {
+      "epoch": 1.1408226842138725,
+      "grad_norm": 2.529242484785609,
+      "learning_rate": 1.4604978836788827e-05,
+      "loss": 0.5167,
+      "step": 2243
+    },
+    {
+      "epoch": 1.14133129887469,
+      "grad_norm": 2.2238504590591996,
+      "learning_rate": 1.4600001014036664e-05,
+      "loss": 0.5387,
+      "step": 2244
+    },
+    {
+      "epoch": 1.1418399135355077,
+      "grad_norm": 1.9613893147982406,
+      "learning_rate": 1.4595021745124263e-05,
+      "loss": 0.4905,
+      "step": 2245
+    },
+    {
+      "epoch": 1.1423485281963253,
+      "grad_norm": 2.169970658626922,
+      "learning_rate": 1.4590041031617017e-05,
+      "loss": 0.4774,
+      "step": 2246
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.9412510155492755,
+      "learning_rate": 1.4585058875080777e-05,
+      "loss": 0.5183,
+      "step": 2247
+    },
+    {
+      "epoch": 1.1433657575179605,
+      "grad_norm": 2.0062130743413578,
+      "learning_rate": 1.458007527708185e-05,
+      "loss": 0.4797,
+      "step": 2248
+    },
+    {
+      "epoch": 1.143874372178778,
+      "grad_norm": 2.1155994272869947,
+      "learning_rate": 1.4575090239186991e-05,
+      "loss": 0.5276,
+      "step": 2249
+    },
+    {
+      "epoch": 1.1443829868395956,
+      "grad_norm": 2.2990018378702977,
+      "learning_rate": 1.4570103762963404e-05,
+      "loss": 0.5707,
+      "step": 2250
+    },
+    {
+      "epoch": 1.1448916015004134,
+      "grad_norm": 1.8358189321337794,
+      "learning_rate": 1.4565115849978755e-05,
+      "loss": 0.4304,
+      "step": 2251
+    },
+    {
+      "epoch": 1.1454002161612309,
+      "grad_norm": 1.8943777120451863,
+      "learning_rate": 1.4560126501801157e-05,
+      "loss": 0.531,
+      "step": 2252
+    },
+    {
+      "epoch": 1.1459088308220484,
+      "grad_norm": 1.8515446658019998,
+      "learning_rate": 1.4555135719999176e-05,
+      "loss": 0.5355,
+      "step": 2253
+    },
+    {
+      "epoch": 1.146417445482866,
+      "grad_norm": 2.525945025820637,
+      "learning_rate": 1.4550143506141821e-05,
+      "loss": 0.5417,
+      "step": 2254
+    },
+    {
+      "epoch": 1.1469260601436837,
+      "grad_norm": 2.1399527280702357,
+      "learning_rate": 1.4545149861798563e-05,
+      "loss": 0.4932,
+      "step": 2255
+    },
+    {
+      "epoch": 1.1474346748045012,
+      "grad_norm": 2.0648701083362453,
+      "learning_rate": 1.4540154788539311e-05,
+      "loss": 0.4971,
+      "step": 2256
+    },
+    {
+      "epoch": 1.1479432894653188,
+      "grad_norm": 1.9854963843668878,
+      "learning_rate": 1.4535158287934432e-05,
+      "loss": 0.4554,
+      "step": 2257
+    },
+    {
+      "epoch": 1.1484519041261365,
+      "grad_norm": 2.3513446085917034,
+      "learning_rate": 1.4530160361554742e-05,
+      "loss": 0.5308,
+      "step": 2258
+    },
+    {
+      "epoch": 1.148960518786954,
+      "grad_norm": 2.260285603697998,
+      "learning_rate": 1.45251610109715e-05,
+      "loss": 0.5018,
+      "step": 2259
+    },
+    {
+      "epoch": 1.1494691334477716,
+      "grad_norm": 1.778983028800515,
+      "learning_rate": 1.4520160237756411e-05,
+      "loss": 0.4918,
+      "step": 2260
+    },
+    {
+      "epoch": 1.1499777481085893,
+      "grad_norm": 2.2216700291685667,
+      "learning_rate": 1.4515158043481636e-05,
+      "loss": 0.5087,
+      "step": 2261
+    },
+    {
+      "epoch": 1.1504863627694069,
+      "grad_norm": 2.2149696919798156,
+      "learning_rate": 1.4510154429719778e-05,
+      "loss": 0.566,
+      "step": 2262
+    },
+    {
+      "epoch": 1.1509949774302244,
+      "grad_norm": 2.603935923208421,
+      "learning_rate": 1.450514939804388e-05,
+      "loss": 0.5414,
+      "step": 2263
+    },
+    {
+      "epoch": 1.151503592091042,
+      "grad_norm": 1.9787315559612022,
+      "learning_rate": 1.4500142950027443e-05,
+      "loss": 0.4541,
+      "step": 2264
+    },
+    {
+      "epoch": 1.1520122067518597,
+      "grad_norm": 1.856886215869436,
+      "learning_rate": 1.4495135087244406e-05,
+      "loss": 0.5108,
+      "step": 2265
+    },
+    {
+      "epoch": 1.1525208214126772,
+      "grad_norm": 2.529854849168491,
+      "learning_rate": 1.4490125811269148e-05,
+      "loss": 0.5338,
+      "step": 2266
+    },
+    {
+      "epoch": 1.1530294360734947,
+      "grad_norm": 2.273999331552211,
+      "learning_rate": 1.448511512367651e-05,
+      "loss": 0.5401,
+      "step": 2267
+    },
+    {
+      "epoch": 1.1535380507343125,
+      "grad_norm": 1.8204708439797979,
+      "learning_rate": 1.4480103026041752e-05,
+      "loss": 0.4968,
+      "step": 2268
+    },
+    {
+      "epoch": 1.15404666539513,
+      "grad_norm": 1.9641653903422405,
+      "learning_rate": 1.4475089519940597e-05,
+      "loss": 0.4702,
+      "step": 2269
+    },
+    {
+      "epoch": 1.1545552800559475,
+      "grad_norm": 2.196397506458741,
+      "learning_rate": 1.4470074606949205e-05,
+      "loss": 0.5043,
+      "step": 2270
+    },
+    {
+      "epoch": 1.1550638947167653,
+      "grad_norm": 2.1060546739152484,
+      "learning_rate": 1.4465058288644173e-05,
+      "loss": 0.5952,
+      "step": 2271
+    },
+    {
+      "epoch": 1.1555725093775828,
+      "grad_norm": 1.7828886291671928,
+      "learning_rate": 1.4460040566602551e-05,
+      "loss": 0.4834,
+      "step": 2272
+    },
+    {
+      "epoch": 1.1560811240384004,
+      "grad_norm": 2.266755552045048,
+      "learning_rate": 1.4455021442401816e-05,
+      "loss": 0.6343,
+      "step": 2273
+    },
+    {
+      "epoch": 1.156589738699218,
+      "grad_norm": 1.9110122977938515,
+      "learning_rate": 1.4450000917619898e-05,
+      "loss": 0.4914,
+      "step": 2274
+    },
+    {
+      "epoch": 1.1570983533600356,
+      "grad_norm": 1.9131959783229076,
+      "learning_rate": 1.444497899383516e-05,
+      "loss": 0.4255,
+      "step": 2275
+    },
+    {
+      "epoch": 1.1576069680208532,
+      "grad_norm": 2.081896464438459,
+      "learning_rate": 1.4439955672626406e-05,
+      "loss": 0.4998,
+      "step": 2276
+    },
+    {
+      "epoch": 1.158115582681671,
+      "grad_norm": 2.235915499938173,
+      "learning_rate": 1.4434930955572886e-05,
+      "loss": 0.5659,
+      "step": 2277
+    },
+    {
+      "epoch": 1.1586241973424884,
+      "grad_norm": 2.673759821819797,
+      "learning_rate": 1.442990484425428e-05,
+      "loss": 0.5827,
+      "step": 2278
+    },
+    {
+      "epoch": 1.159132812003306,
+      "grad_norm": 2.1972303599600624,
+      "learning_rate": 1.442487734025071e-05,
+      "loss": 0.4638,
+      "step": 2279
+    },
+    {
+      "epoch": 1.1596414266641235,
+      "grad_norm": 1.9939981049095339,
+      "learning_rate": 1.441984844514274e-05,
+      "loss": 0.499,
+      "step": 2280
+    },
+    {
+      "epoch": 1.1601500413249413,
+      "grad_norm": 1.891629984439773,
+      "learning_rate": 1.4414818160511357e-05,
+      "loss": 0.46,
+      "step": 2281
+    },
+    {
+      "epoch": 1.1606586559857588,
+      "grad_norm": 2.117035694943992,
+      "learning_rate": 1.4409786487938004e-05,
+      "loss": 0.4196,
+      "step": 2282
+    },
+    {
+      "epoch": 1.1611672706465763,
+      "grad_norm": 2.306185167843499,
+      "learning_rate": 1.440475342900455e-05,
+      "loss": 0.4992,
+      "step": 2283
+    },
+    {
+      "epoch": 1.161675885307394,
+      "grad_norm": 2.111265564449525,
+      "learning_rate": 1.4399718985293298e-05,
+      "loss": 0.4833,
+      "step": 2284
+    },
+    {
+      "epoch": 1.1621844999682116,
+      "grad_norm": 1.9451611318258109,
+      "learning_rate": 1.4394683158386995e-05,
+      "loss": 0.4488,
+      "step": 2285
+    },
+    {
+      "epoch": 1.1626931146290291,
+      "grad_norm": 2.0250800693835664,
+      "learning_rate": 1.4389645949868808e-05,
+      "loss": 0.426,
+      "step": 2286
+    },
+    {
+      "epoch": 1.1632017292898467,
+      "grad_norm": 2.14348917450428,
+      "learning_rate": 1.4384607361322356e-05,
+      "loss": 0.5572,
+      "step": 2287
+    },
+    {
+      "epoch": 1.1637103439506644,
+      "grad_norm": 1.9640117119752625,
+      "learning_rate": 1.4379567394331679e-05,
+      "loss": 0.5012,
+      "step": 2288
+    },
+    {
+      "epoch": 1.164218958611482,
+      "grad_norm": 2.2271931073142417,
+      "learning_rate": 1.4374526050481258e-05,
+      "loss": 0.5338,
+      "step": 2289
+    },
+    {
+      "epoch": 1.1647275732722995,
+      "grad_norm": 2.5745182586692197,
+      "learning_rate": 1.4369483331356005e-05,
+      "loss": 0.5729,
+      "step": 2290
+    },
+    {
+      "epoch": 1.1652361879331172,
+      "grad_norm": 2.338884109085446,
+      "learning_rate": 1.4364439238541257e-05,
+      "loss": 0.5823,
+      "step": 2291
+    },
+    {
+      "epoch": 1.1657448025939348,
+      "grad_norm": 2.150768666034858,
+      "learning_rate": 1.4359393773622791e-05,
+      "loss": 0.4204,
+      "step": 2292
+    },
+    {
+      "epoch": 1.1662534172547523,
+      "grad_norm": 1.8425780628908723,
+      "learning_rate": 1.4354346938186815e-05,
+      "loss": 0.5109,
+      "step": 2293
+    },
+    {
+      "epoch": 1.16676203191557,
+      "grad_norm": 1.8602022526614908,
+      "learning_rate": 1.434929873381997e-05,
+      "loss": 0.4799,
+      "step": 2294
+    },
+    {
+      "epoch": 1.1672706465763876,
+      "grad_norm": 2.71683452259254,
+      "learning_rate": 1.4344249162109316e-05,
+      "loss": 0.5361,
+      "step": 2295
+    },
+    {
+      "epoch": 1.167779261237205,
+      "grad_norm": 2.146350637752523,
+      "learning_rate": 1.4339198224642353e-05,
+      "loss": 0.5468,
+      "step": 2296
+    },
+    {
+      "epoch": 1.1682878758980229,
+      "grad_norm": 1.9681674761532928,
+      "learning_rate": 1.433414592300701e-05,
+      "loss": 0.5177,
+      "step": 2297
+    },
+    {
+      "epoch": 1.1687964905588404,
+      "grad_norm": 2.391369273695267,
+      "learning_rate": 1.4329092258791643e-05,
+      "loss": 0.5275,
+      "step": 2298
+    },
+    {
+      "epoch": 1.169305105219658,
+      "grad_norm": 1.8435061665311043,
+      "learning_rate": 1.4324037233585032e-05,
+      "loss": 0.4739,
+      "step": 2299
+    },
+    {
+      "epoch": 1.1698137198804757,
+      "grad_norm": 2.3854344685208138,
+      "learning_rate": 1.4318980848976391e-05,
+      "loss": 0.4189,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1698137198804757,
+      "eval_loss": 0.7337888479232788,
+      "eval_runtime": 170.4489,
+      "eval_samples_per_second": 19.431,
+      "eval_steps_per_second": 9.716,
+      "step": 2300
+    },
+    {
+      "epoch": 1.1703223345412932,
+      "grad_norm": 2.4091497016590027,
+      "learning_rate": 1.4313923106555362e-05,
+      "loss": 0.5113,
+      "step": 2301
+    },
+    {
+      "epoch": 1.1708309492021107,
+      "grad_norm": 2.1129299019822967,
+      "learning_rate": 1.4308864007912009e-05,
+      "loss": 0.462,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1713395638629283,
+      "grad_norm": 2.088971620921145,
+      "learning_rate": 1.4303803554636823e-05,
+      "loss": 0.4805,
+      "step": 2303
+    },
+    {
+      "epoch": 1.171848178523746,
+      "grad_norm": 2.032328744690524,
+      "learning_rate": 1.4298741748320723e-05,
+      "loss": 0.5008,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1723567931845635,
+      "grad_norm": 2.216950022242957,
+      "learning_rate": 1.4293678590555055e-05,
+      "loss": 0.4343,
+      "step": 2305
+    },
+    {
+      "epoch": 1.172865407845381,
+      "grad_norm": 2.4701837167100176,
+      "learning_rate": 1.4288614082931584e-05,
+      "loss": 0.6193,
+      "step": 2306
+    },
+    {
+      "epoch": 1.1733740225061988,
+      "grad_norm": 2.088195493045441,
+      "learning_rate": 1.4283548227042507e-05,
+      "loss": 0.5754,
+      "step": 2307
+    },
+    {
+      "epoch": 1.1738826371670164,
+      "grad_norm": 2.002147713485139,
+      "learning_rate": 1.4278481024480442e-05,
+      "loss": 0.5848,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1743912518278339,
+      "grad_norm": 2.3582323351509524,
+      "learning_rate": 1.4273412476838422e-05,
+      "loss": 0.6184,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1748998664886514,
+      "grad_norm": 1.8431551992193778,
+      "learning_rate": 1.4268342585709914e-05,
+      "loss": 0.4293,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1754084811494692,
+      "grad_norm": 1.948106467507959,
+      "learning_rate": 1.4263271352688805e-05,
+      "loss": 0.4257,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1759170958102867,
+      "grad_norm": 1.9834390670844704,
+      "learning_rate": 1.4258198779369404e-05,
+      "loss": 0.4898,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1764257104711042,
+      "grad_norm": 2.4889268741187824,
+      "learning_rate": 1.4253124867346435e-05,
+      "loss": 0.5686,
+      "step": 2313
+    },
+    {
+      "epoch": 1.176934325131922,
+      "grad_norm": 2.2049041490200807,
+      "learning_rate": 1.4248049618215049e-05,
+      "loss": 0.5917,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1774429397927395,
+      "grad_norm": 2.2658901717125817,
+      "learning_rate": 1.424297303357082e-05,
+      "loss": 0.4983,
+      "step": 2315
+    },
+    {
+      "epoch": 1.177951554453557,
+      "grad_norm": 1.9171313537981858,
+      "learning_rate": 1.4237895115009731e-05,
+      "loss": 0.463,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1784601691143748,
+      "grad_norm": 2.333383650464258,
+      "learning_rate": 1.4232815864128198e-05,
+      "loss": 0.5305,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1789687837751923,
+      "grad_norm": 2.0500069455284993,
+      "learning_rate": 1.4227735282523045e-05,
+      "loss": 0.5032,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1794773984360098,
+      "grad_norm": 2.231994534605438,
+      "learning_rate": 1.4222653371791517e-05,
+      "loss": 0.4668,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1799860130968276,
+      "grad_norm": 1.964153494657885,
+      "learning_rate": 1.4217570133531282e-05,
+      "loss": 0.4653,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1804946277576451,
+      "grad_norm": 2.0317601663647973,
+      "learning_rate": 1.4212485569340421e-05,
+      "loss": 0.6006,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1810032424184627,
+      "grad_norm": 2.1209370535898815,
+      "learning_rate": 1.420739968081743e-05,
+      "loss": 0.4482,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1815118570792804,
+      "grad_norm": 2.1439227020720026,
+      "learning_rate": 1.4202312469561228e-05,
+      "loss": 0.4375,
+      "step": 2323
+    },
+    {
+      "epoch": 1.182020471740098,
+      "grad_norm": 2.236114678337248,
+      "learning_rate": 1.419722393717114e-05,
+      "loss": 0.5775,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1825290864009155,
+      "grad_norm": 2.0565910079781733,
+      "learning_rate": 1.4192134085246916e-05,
+      "loss": 0.5349,
+      "step": 2325
+    },
+    {
+      "epoch": 1.1830377010617332,
+      "grad_norm": 2.3958019973253597,
+      "learning_rate": 1.4187042915388718e-05,
+      "loss": 0.5446,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1835463157225508,
+      "grad_norm": 1.8397192781301197,
+      "learning_rate": 1.4181950429197113e-05,
+      "loss": 0.5491,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1840549303833683,
+      "grad_norm": 3.0402865653745814,
+      "learning_rate": 1.4176856628273102e-05,
+      "loss": 0.5359,
+      "step": 2328
+    },
+    {
+      "epoch": 1.1845635450441858,
+      "grad_norm": 2.0318487605334097,
+      "learning_rate": 1.4171761514218078e-05,
+      "loss": 0.5419,
+      "step": 2329
+    },
+    {
+      "epoch": 1.1850721597050036,
+      "grad_norm": 2.8234673182354273,
+      "learning_rate": 1.4166665088633863e-05,
+      "loss": 0.5051,
+      "step": 2330
+    },
+    {
+      "epoch": 1.185580774365821,
+      "grad_norm": 2.2815315386389465,
+      "learning_rate": 1.416156735312268e-05,
+      "loss": 0.5153,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1860893890266386,
+      "grad_norm": 2.5957789011759616,
+      "learning_rate": 1.4156468309287169e-05,
+      "loss": 0.4994,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1865980036874564,
+      "grad_norm": 1.9120182253316111,
+      "learning_rate": 1.415136795873038e-05,
+      "loss": 0.4838,
+      "step": 2333
+    },
+    {
+      "epoch": 1.187106618348274,
+      "grad_norm": 1.975627693191329,
+      "learning_rate": 1.4146266303055774e-05,
+      "loss": 0.4539,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1876152330090914,
+      "grad_norm": 2.076711532491831,
+      "learning_rate": 1.4141163343867228e-05,
+      "loss": 0.4913,
+      "step": 2335
+    },
+    {
+      "epoch": 1.188123847669909,
+      "grad_norm": 2.2359585785688614,
+      "learning_rate": 1.4136059082769018e-05,
+      "loss": 0.5675,
+      "step": 2336
+    },
+    {
+      "epoch": 1.1886324623307267,
+      "grad_norm": 2.203234622473571,
+      "learning_rate": 1.4130953521365835e-05,
+      "loss": 0.5069,
+      "step": 2337
+    },
+    {
+      "epoch": 1.1891410769915443,
+      "grad_norm": 1.8617765160746025,
+      "learning_rate": 1.4125846661262781e-05,
+      "loss": 0.5471,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1896496916523618,
+      "grad_norm": 2.3947234084423807,
+      "learning_rate": 1.4120738504065364e-05,
+      "loss": 0.5927,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1901583063131795,
+      "grad_norm": 1.9765292078162129,
+      "learning_rate": 1.4115629051379501e-05,
+      "loss": 0.5099,
+      "step": 2340
+    },
+    {
+      "epoch": 1.190666920973997,
+      "grad_norm": 2.016278183249959,
+      "learning_rate": 1.4110518304811507e-05,
+      "loss": 0.5073,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1911755356348146,
+      "grad_norm": 1.758613641220525,
+      "learning_rate": 1.410540626596812e-05,
+      "loss": 0.5064,
+      "step": 2342
+    },
+    {
+      "epoch": 1.1916841502956323,
+      "grad_norm": 1.9513325515283275,
+      "learning_rate": 1.4100292936456473e-05,
+      "loss": 0.4439,
+      "step": 2343
+    },
+    {
+      "epoch": 1.1921927649564499,
+      "grad_norm": 2.0727671915983246,
+      "learning_rate": 1.4095178317884111e-05,
+      "loss": 0.4254,
+      "step": 2344
+    },
+    {
+      "epoch": 1.1927013796172674,
+      "grad_norm": 2.016849352966412,
+      "learning_rate": 1.4090062411858973e-05,
+      "loss": 0.5065,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1932099942780852,
+      "grad_norm": 1.9261956997297098,
+      "learning_rate": 1.408494521998942e-05,
+      "loss": 0.4022,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1937186089389027,
+      "grad_norm": 1.9212530106550068,
+      "learning_rate": 1.4079826743884202e-05,
+      "loss": 0.5226,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1942272235997202,
+      "grad_norm": 2.5755886664231107,
+      "learning_rate": 1.4074706985152485e-05,
+      "loss": 0.4574,
+      "step": 2348
+    },
+    {
+      "epoch": 1.194735838260538,
+      "grad_norm": 2.206595858948576,
+      "learning_rate": 1.4069585945403824e-05,
+      "loss": 0.5267,
+      "step": 2349
+    },
+    {
+      "epoch": 1.1952444529213555,
+      "grad_norm": 2.1911087786206633,
+      "learning_rate": 1.4064463626248188e-05,
+      "loss": 0.5119,
+      "step": 2350
+    },
+    {
+      "epoch": 1.195753067582173,
+      "grad_norm": 2.0858042242213877,
+      "learning_rate": 1.4059340029295946e-05,
+      "loss": 0.5119,
+      "step": 2351
+    },
+    {
+      "epoch": 1.1962616822429906,
+      "grad_norm": 2.00704917795765,
+      "learning_rate": 1.4054215156157869e-05,
+      "loss": 0.4792,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1967702969038083,
+      "grad_norm": 2.347301792799966,
+      "learning_rate": 1.4049089008445127e-05,
+      "loss": 0.5304,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1972789115646258,
+      "grad_norm": 1.9461105731863302,
+      "learning_rate": 1.4043961587769287e-05,
+      "loss": 0.5661,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1977875262254434,
+      "grad_norm": 2.2638587117540414,
+      "learning_rate": 1.4038832895742328e-05,
+      "loss": 0.4803,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1982961408862611,
+      "grad_norm": 2.063147472130932,
+      "learning_rate": 1.4033702933976617e-05,
+      "loss": 0.5541,
+      "step": 2356
+    },
+    {
+      "epoch": 1.1988047555470787,
+      "grad_norm": 1.8976492795663074,
+      "learning_rate": 1.4028571704084925e-05,
+      "loss": 0.5249,
+      "step": 2357
+    },
+    {
+      "epoch": 1.1993133702078962,
+      "grad_norm": 1.9515863690470427,
+      "learning_rate": 1.4023439207680426e-05,
+      "loss": 0.5723,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1998219848687137,
+      "grad_norm": 1.9798538487372292,
+      "learning_rate": 1.401830544637668e-05,
+      "loss": 0.4514,
+      "step": 2359
+    },
+    {
+      "epoch": 1.2003305995295315,
+      "grad_norm": 1.9577431973234538,
+      "learning_rate": 1.4013170421787657e-05,
+      "loss": 0.4423,
+      "step": 2360
+    },
+    {
+      "epoch": 1.200839214190349,
+      "grad_norm": 2.1210422685252186,
+      "learning_rate": 1.4008034135527714e-05,
+      "loss": 0.5356,
+      "step": 2361
+    },
+    {
+      "epoch": 1.2013478288511665,
+      "grad_norm": 2.004598100293715,
+      "learning_rate": 1.4002896589211619e-05,
+      "loss": 0.4107,
+      "step": 2362
+    },
+    {
+      "epoch": 1.2018564435119843,
+      "grad_norm": 1.873534710853967,
+      "learning_rate": 1.399775778445452e-05,
+      "loss": 0.4542,
+      "step": 2363
+    },
+    {
+      "epoch": 1.2023650581728018,
+      "grad_norm": 1.785291403698053,
+      "learning_rate": 1.399261772287197e-05,
+      "loss": 0.4627,
+      "step": 2364
+    },
+    {
+      "epoch": 1.2028736728336193,
+      "grad_norm": 2.000741463548247,
+      "learning_rate": 1.3987476406079916e-05,
+      "loss": 0.528,
+      "step": 2365
+    },
+    {
+      "epoch": 1.203382287494437,
+      "grad_norm": 1.9984482467398201,
+      "learning_rate": 1.39823338356947e-05,
+      "loss": 0.5132,
+      "step": 2366
+    },
+    {
+      "epoch": 1.2038909021552546,
+      "grad_norm": 2.138525608582804,
+      "learning_rate": 1.3977190013333049e-05,
+      "loss": 0.5533,
+      "step": 2367
+    },
+    {
+      "epoch": 1.2043995168160722,
+      "grad_norm": 2.015082707754646,
+      "learning_rate": 1.3972044940612097e-05,
+      "loss": 0.5215,
+      "step": 2368
+    },
+    {
+      "epoch": 1.20490813147689,
+      "grad_norm": 2.1760720515804235,
+      "learning_rate": 1.3966898619149364e-05,
+      "loss": 0.4198,
+      "step": 2369
+    },
+    {
+      "epoch": 1.2054167461377074,
+      "grad_norm": 2.180641977118905,
+      "learning_rate": 1.3961751050562766e-05,
+      "loss": 0.5228,
+      "step": 2370
+    },
+    {
+      "epoch": 1.205925360798525,
+      "grad_norm": 2.2358369112880694,
+      "learning_rate": 1.395660223647061e-05,
+      "loss": 0.5569,
+      "step": 2371
+    },
+    {
+      "epoch": 1.2064339754593427,
+      "grad_norm": 1.9887615590505876,
+      "learning_rate": 1.3951452178491588e-05,
+      "loss": 0.4434,
+      "step": 2372
+    },
+    {
+      "epoch": 1.2069425901201603,
+      "grad_norm": 1.997054255264756,
+      "learning_rate": 1.3946300878244792e-05,
+      "loss": 0.4402,
+      "step": 2373
+    },
+    {
+      "epoch": 1.2074512047809778,
+      "grad_norm": 1.9592866065424268,
+      "learning_rate": 1.3941148337349696e-05,
+      "loss": 0.5077,
+      "step": 2374
+    },
+    {
+      "epoch": 1.2079598194417953,
+      "grad_norm": 2.2177280705261873,
+      "learning_rate": 1.3935994557426181e-05,
+      "loss": 0.6147,
+      "step": 2375
+    },
+    {
+      "epoch": 1.208468434102613,
+      "grad_norm": 1.8218641094320658,
+      "learning_rate": 1.3930839540094496e-05,
+      "loss": 0.4695,
+      "step": 2376
+    },
+    {
+      "epoch": 1.2089770487634306,
+      "grad_norm": 1.9895139219415794,
+      "learning_rate": 1.3925683286975289e-05,
+      "loss": 0.49,
+      "step": 2377
+    },
+    {
+      "epoch": 1.2094856634242481,
+      "grad_norm": 2.2080851241440644,
+      "learning_rate": 1.3920525799689598e-05,
+      "loss": 0.5836,
+      "step": 2378
+    },
+    {
+      "epoch": 1.2099942780850659,
+      "grad_norm": 2.401858198517584,
+      "learning_rate": 1.3915367079858846e-05,
+      "loss": 0.5772,
+      "step": 2379
+    },
+    {
+      "epoch": 1.2105028927458834,
+      "grad_norm": 2.044684949264323,
+      "learning_rate": 1.3910207129104852e-05,
+      "loss": 0.4585,
+      "step": 2380
+    },
+    {
+      "epoch": 1.211011507406701,
+      "grad_norm": 2.249196226875767,
+      "learning_rate": 1.3905045949049805e-05,
+      "loss": 0.5735,
+      "step": 2381
+    },
+    {
+      "epoch": 1.2115201220675187,
+      "grad_norm": 2.0786513042131034,
+      "learning_rate": 1.3899883541316294e-05,
+      "loss": 0.5108,
+      "step": 2382
+    },
+    {
+      "epoch": 1.2120287367283362,
+      "grad_norm": 2.139086316507638,
+      "learning_rate": 1.389471990752729e-05,
+      "loss": 0.5055,
+      "step": 2383
+    },
+    {
+      "epoch": 1.2125373513891538,
+      "grad_norm": 2.0956977780177013,
+      "learning_rate": 1.3889555049306152e-05,
+      "loss": 0.5015,
+      "step": 2384
+    },
+    {
+      "epoch": 1.2130459660499713,
+      "grad_norm": 2.0551205022895043,
+      "learning_rate": 1.3884388968276619e-05,
+      "loss": 0.4798,
+      "step": 2385
+    },
+    {
+      "epoch": 1.213554580710789,
+      "grad_norm": 2.184527311895188,
+      "learning_rate": 1.3879221666062815e-05,
+      "loss": 0.4981,
+      "step": 2386
+    },
+    {
+      "epoch": 1.2140631953716066,
+      "grad_norm": 1.805824153027732,
+      "learning_rate": 1.3874053144289252e-05,
+      "loss": 0.4495,
+      "step": 2387
+    },
+    {
+      "epoch": 1.214571810032424,
+      "grad_norm": 2.0171323359457043,
+      "learning_rate": 1.3868883404580824e-05,
+      "loss": 0.4735,
+      "step": 2388
+    },
+    {
+      "epoch": 1.2150804246932418,
+      "grad_norm": 2.1139603121378476,
+      "learning_rate": 1.386371244856281e-05,
+      "loss": 0.4368,
+      "step": 2389
+    },
+    {
+      "epoch": 1.2155890393540594,
+      "grad_norm": 2.283443911397121,
+      "learning_rate": 1.3858540277860861e-05,
+      "loss": 0.6224,
+      "step": 2390
+    },
+    {
+      "epoch": 1.216097654014877,
+      "grad_norm": 1.9119879242743143,
+      "learning_rate": 1.3853366894101022e-05,
+      "loss": 0.4228,
+      "step": 2391
+    },
+    {
+      "epoch": 1.2166062686756947,
+      "grad_norm": 1.9720520907286814,
+      "learning_rate": 1.3848192298909715e-05,
+      "loss": 0.4944,
+      "step": 2392
+    },
+    {
+      "epoch": 1.2171148833365122,
+      "grad_norm": 2.0442287432636195,
+      "learning_rate": 1.384301649391374e-05,
+      "loss": 0.5971,
+      "step": 2393
+    },
+    {
+      "epoch": 1.2176234979973297,
+      "grad_norm": 2.417373016872625,
+      "learning_rate": 1.3837839480740285e-05,
+      "loss": 0.5662,
+      "step": 2394
+    },
+    {
+      "epoch": 1.2181321126581475,
+      "grad_norm": 2.0620980304536753,
+      "learning_rate": 1.3832661261016908e-05,
+      "loss": 0.4925,
+      "step": 2395
+    },
+    {
+      "epoch": 1.218640727318965,
+      "grad_norm": 2.161576368180707,
+      "learning_rate": 1.3827481836371553e-05,
+      "loss": 0.5734,
+      "step": 2396
+    },
+    {
+      "epoch": 1.2191493419797825,
+      "grad_norm": 2.4287714074448385,
+      "learning_rate": 1.3822301208432537e-05,
+      "loss": 0.5994,
+      "step": 2397
+    },
+    {
+      "epoch": 1.2196579566406003,
+      "grad_norm": 2.005134310326246,
+      "learning_rate": 1.381711937882857e-05,
+      "loss": 0.5181,
+      "step": 2398
+    },
+    {
+      "epoch": 1.2201665713014178,
+      "grad_norm": 2.0434412218437132,
+      "learning_rate": 1.3811936349188717e-05,
+      "loss": 0.5412,
+      "step": 2399
+    },
+    {
+      "epoch": 1.2206751859622353,
+      "grad_norm": 2.552214659341647,
+      "learning_rate": 1.3806752121142437e-05,
+      "loss": 0.5148,
+      "step": 2400
+    },
+    {
+      "epoch": 1.2206751859622353,
+      "eval_loss": 0.7196903228759766,
+      "eval_runtime": 170.3013,
+      "eval_samples_per_second": 19.448,
+      "eval_steps_per_second": 9.724,
+      "step": 2400
+    },
+    {
+      "epoch": 1.2211838006230529,
+      "grad_norm": 2.0447911728093824,
+      "learning_rate": 1.3801566696319564e-05,
+      "loss": 0.5319,
+      "step": 2401
+    },
+    {
+      "epoch": 1.2216924152838706,
+      "grad_norm": 2.1165609523576605,
+      "learning_rate": 1.3796380076350302e-05,
+      "loss": 0.5389,
+      "step": 2402
+    },
+    {
+      "epoch": 1.2222010299446882,
+      "grad_norm": 1.9372039076455951,
+      "learning_rate": 1.3791192262865236e-05,
+      "loss": 0.5269,
+      "step": 2403
+    },
+    {
+      "epoch": 1.2227096446055057,
+      "grad_norm": 1.8424113958578616,
+      "learning_rate": 1.3786003257495319e-05,
+      "loss": 0.4823,
+      "step": 2404
+    },
+    {
+      "epoch": 1.2232182592663234,
+      "grad_norm": 2.1616820315381386,
+      "learning_rate": 1.3780813061871892e-05,
+      "loss": 0.5171,
+      "step": 2405
+    },
+    {
+      "epoch": 1.223726873927141,
+      "grad_norm": 1.9389874685865227,
+      "learning_rate": 1.3775621677626656e-05,
+      "loss": 0.5068,
+      "step": 2406
+    },
+    {
+      "epoch": 1.2242354885879585,
+      "grad_norm": 2.3496461074253627,
+      "learning_rate": 1.3770429106391694e-05,
+      "loss": 0.5623,
+      "step": 2407
+    },
+    {
+      "epoch": 1.224744103248776,
+      "grad_norm": 2.1721998013019252,
+      "learning_rate": 1.3765235349799461e-05,
+      "loss": 0.5592,
+      "step": 2408
+    },
+    {
+      "epoch": 1.2252527179095938,
+      "grad_norm": 1.909239753890457,
+      "learning_rate": 1.3760040409482782e-05,
+      "loss": 0.5054,
+      "step": 2409
+    },
+    {
+      "epoch": 1.2257613325704113,
+      "grad_norm": 1.9989338655814768,
+      "learning_rate": 1.3754844287074853e-05,
+      "loss": 0.4716,
+      "step": 2410
+    },
+    {
+      "epoch": 1.2262699472312288,
+      "grad_norm": 2.1611436239421162,
+      "learning_rate": 1.374964698420925e-05,
+      "loss": 0.5284,
+      "step": 2411
+    },
+    {
+      "epoch": 1.2267785618920466,
+      "grad_norm": 2.06114580028689,
+      "learning_rate": 1.3744448502519916e-05,
+      "loss": 0.551,
+      "step": 2412
+    },
+    {
+      "epoch": 1.2272871765528641,
+      "grad_norm": 2.1396840944370306,
+      "learning_rate": 1.3739248843641155e-05,
+      "loss": 0.5079,
+      "step": 2413
+    },
+    {
+      "epoch": 1.2277957912136817,
+      "grad_norm": 1.9716675953071494,
+      "learning_rate": 1.3734048009207652e-05,
+      "loss": 0.5457,
+      "step": 2414
+    },
+    {
+      "epoch": 1.2283044058744994,
+      "grad_norm": 2.0288348585280938,
+      "learning_rate": 1.3728846000854464e-05,
+      "loss": 0.469,
+      "step": 2415
+    },
+    {
+      "epoch": 1.228813020535317,
+      "grad_norm": 2.1325860095371842,
+      "learning_rate": 1.3723642820217009e-05,
+      "loss": 0.4686,
+      "step": 2416
+    },
+    {
+      "epoch": 1.2293216351961345,
+      "grad_norm": 2.3837788607806076,
+      "learning_rate": 1.3718438468931075e-05,
+      "loss": 0.5678,
+      "step": 2417
+    },
+    {
+      "epoch": 1.2298302498569522,
+      "grad_norm": 1.8144273704778755,
+      "learning_rate": 1.3713232948632818e-05,
+      "loss": 0.5466,
+      "step": 2418
+    },
+    {
+      "epoch": 1.2303388645177697,
+      "grad_norm": 2.308887959950154,
+      "learning_rate": 1.3708026260958767e-05,
+      "loss": 0.5322,
+      "step": 2419
+    },
+    {
+      "epoch": 1.2308474791785873,
+      "grad_norm": 2.3310647108472184,
+      "learning_rate": 1.3702818407545811e-05,
+      "loss": 0.4712,
+      "step": 2420
+    },
+    {
+      "epoch": 1.231356093839405,
+      "grad_norm": 2.086235568006366,
+      "learning_rate": 1.369760939003121e-05,
+      "loss": 0.521,
+      "step": 2421
+    },
+    {
+      "epoch": 1.2318647085002226,
+      "grad_norm": 2.7592418737904465,
+      "learning_rate": 1.3692399210052589e-05,
+      "loss": 0.642,
+      "step": 2422
+    },
+    {
+      "epoch": 1.23237332316104,
+      "grad_norm": 2.107369493853479,
+      "learning_rate": 1.3687187869247936e-05,
+      "loss": 0.6187,
+      "step": 2423
+    },
+    {
+      "epoch": 1.2328819378218576,
+      "grad_norm": 2.313558903102798,
+      "learning_rate": 1.3681975369255604e-05,
+      "loss": 0.511,
+      "step": 2424
+    },
+    {
+      "epoch": 1.2333905524826754,
+      "grad_norm": 1.8770918303652657,
+      "learning_rate": 1.3676761711714319e-05,
+      "loss": 0.4606,
+      "step": 2425
+    },
+    {
+      "epoch": 1.233899167143493,
+      "grad_norm": 1.879832715444494,
+      "learning_rate": 1.3671546898263159e-05,
+      "loss": 0.4912,
+      "step": 2426
+    },
+    {
+      "epoch": 1.2344077818043104,
+      "grad_norm": 2.1196360487994435,
+      "learning_rate": 1.3666330930541572e-05,
+      "loss": 0.5408,
+      "step": 2427
+    },
+    {
+      "epoch": 1.2349163964651282,
+      "grad_norm": 2.076989803166124,
+      "learning_rate": 1.3661113810189366e-05,
+      "loss": 0.5259,
+      "step": 2428
+    },
+    {
+      "epoch": 1.2354250111259457,
+      "grad_norm": 1.9836336688501848,
+      "learning_rate": 1.3655895538846713e-05,
+      "loss": 0.5272,
+      "step": 2429
+    },
+    {
+      "epoch": 1.2359336257867632,
+      "grad_norm": 1.9363197915961992,
+      "learning_rate": 1.365067611815415e-05,
+      "loss": 0.4486,
+      "step": 2430
+    },
+    {
+      "epoch": 1.236442240447581,
+      "grad_norm": 2.2551202793565657,
+      "learning_rate": 1.3645455549752567e-05,
+      "loss": 0.5244,
+      "step": 2431
+    },
+    {
+      "epoch": 1.2369508551083985,
+      "grad_norm": 2.3071887909721434,
+      "learning_rate": 1.3640233835283224e-05,
+      "loss": 0.6207,
+      "step": 2432
+    },
+    {
+      "epoch": 1.237459469769216,
+      "grad_norm": 2.025804468177058,
+      "learning_rate": 1.3635010976387734e-05,
+      "loss": 0.4702,
+      "step": 2433
+    },
+    {
+      "epoch": 1.2379680844300336,
+      "grad_norm": 1.9226783376058205,
+      "learning_rate": 1.3629786974708076e-05,
+      "loss": 0.5088,
+      "step": 2434
+    },
+    {
+      "epoch": 1.2384766990908513,
+      "grad_norm": 2.1719068346503443,
+      "learning_rate": 1.3624561831886585e-05,
+      "loss": 0.5347,
+      "step": 2435
+    },
+    {
+      "epoch": 1.2389853137516689,
+      "grad_norm": 2.208957607872751,
+      "learning_rate": 1.3619335549565951e-05,
+      "loss": 0.5333,
+      "step": 2436
+    },
+    {
+      "epoch": 1.2394939284124864,
+      "grad_norm": 2.050171854212573,
+      "learning_rate": 1.361410812938923e-05,
+      "loss": 0.4957,
+      "step": 2437
+    },
+    {
+      "epoch": 1.2400025430733042,
+      "grad_norm": 2.139458777180091,
+      "learning_rate": 1.3608879572999829e-05,
+      "loss": 0.588,
+      "step": 2438
+    },
+    {
+      "epoch": 1.2405111577341217,
+      "grad_norm": 1.876879703542511,
+      "learning_rate": 1.360364988204152e-05,
+      "loss": 0.5506,
+      "step": 2439
+    },
+    {
+      "epoch": 1.2410197723949392,
+      "grad_norm": 1.7765034124769465,
+      "learning_rate": 1.3598419058158421e-05,
+      "loss": 0.504,
+      "step": 2440
+    },
+    {
+      "epoch": 1.241528387055757,
+      "grad_norm": 2.116911922974153,
+      "learning_rate": 1.3593187102995015e-05,
+      "loss": 0.5028,
+      "step": 2441
+    },
+    {
+      "epoch": 1.2420370017165745,
+      "grad_norm": 2.070694682995136,
+      "learning_rate": 1.3587954018196136e-05,
+      "loss": 0.4725,
+      "step": 2442
+    },
+    {
+      "epoch": 1.242545616377392,
+      "grad_norm": 2.1369675625671896,
+      "learning_rate": 1.3582719805406978e-05,
+      "loss": 0.4924,
+      "step": 2443
+    },
+    {
+      "epoch": 1.2430542310382098,
+      "grad_norm": 2.120261013049632,
+      "learning_rate": 1.3577484466273085e-05,
+      "loss": 0.5803,
+      "step": 2444
+    },
+    {
+      "epoch": 1.2435628456990273,
+      "grad_norm": 2.3128834999256913,
+      "learning_rate": 1.3572248002440354e-05,
+      "loss": 0.4762,
+      "step": 2445
+    },
+    {
+      "epoch": 1.2440714603598448,
+      "grad_norm": 1.9322021408797307,
+      "learning_rate": 1.3567010415555038e-05,
+      "loss": 0.4887,
+      "step": 2446
+    },
+    {
+      "epoch": 1.2445800750206626,
+      "grad_norm": 1.7178579616516356,
+      "learning_rate": 1.3561771707263746e-05,
+      "loss": 0.4075,
+      "step": 2447
+    },
+    {
+      "epoch": 1.2450886896814801,
+      "grad_norm": 2.37721159508445,
+      "learning_rate": 1.3556531879213438e-05,
+      "loss": 0.5444,
+      "step": 2448
+    },
+    {
+      "epoch": 1.2455973043422977,
+      "grad_norm": 2.0284054131762312,
+      "learning_rate": 1.355129093305142e-05,
+      "loss": 0.577,
+      "step": 2449
+    },
+    {
+      "epoch": 1.2461059190031152,
+      "grad_norm": 2.1415625976134973,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.5381,
+      "step": 2450
+    },
+    {
+      "epoch": 1.246614533663933,
+      "grad_norm": 2.026746834729533,
+      "learning_rate": 1.3540805692983261e-05,
+      "loss": 0.5402,
+      "step": 2451
+    },
+    {
+      "epoch": 1.2471231483247505,
+      "grad_norm": 1.9831854190753586,
+      "learning_rate": 1.3535561402373497e-05,
+      "loss": 0.533,
+      "step": 2452
+    },
+    {
+      "epoch": 1.247631762985568,
+      "grad_norm": 2.2163779245152586,
+      "learning_rate": 1.3530316000244783e-05,
+      "loss": 0.5228,
+      "step": 2453
+    },
+    {
+      "epoch": 1.2481403776463857,
+      "grad_norm": 2.079300626772517,
+      "learning_rate": 1.3525069488246171e-05,
+      "loss": 0.6249,
+      "step": 2454
+    },
+    {
+      "epoch": 1.2486489923072033,
+      "grad_norm": 1.8922969266717649,
+      "learning_rate": 1.351982186802708e-05,
+      "loss": 0.4734,
+      "step": 2455
+    },
+    {
+      "epoch": 1.2491576069680208,
+      "grad_norm": 1.8768005626651607,
+      "learning_rate": 1.3514573141237273e-05,
+      "loss": 0.4472,
+      "step": 2456
+    },
+    {
+      "epoch": 1.2496662216288383,
+      "grad_norm": 2.020728341223089,
+      "learning_rate": 1.3509323309526856e-05,
+      "loss": 0.4975,
+      "step": 2457
+    },
+    {
+      "epoch": 1.250174836289656,
+      "grad_norm": 1.7624567013014538,
+      "learning_rate": 1.350407237454628e-05,
+      "loss": 0.4012,
+      "step": 2458
+    },
+    {
+      "epoch": 1.2506834509504736,
+      "grad_norm": 2.4361258325696227,
+      "learning_rate": 1.3498820337946352e-05,
+      "loss": 0.5802,
+      "step": 2459
+    },
+    {
+      "epoch": 1.2511920656112911,
+      "grad_norm": 2.1948965092551247,
+      "learning_rate": 1.3493567201378221e-05,
+      "loss": 0.4529,
+      "step": 2460
+    },
+    {
+      "epoch": 1.251700680272109,
+      "grad_norm": 1.8511640546528472,
+      "learning_rate": 1.3488312966493382e-05,
+      "loss": 0.4449,
+      "step": 2461
+    },
+    {
+      "epoch": 1.2522092949329264,
+      "grad_norm": 2.5530397660071737,
+      "learning_rate": 1.3483057634943672e-05,
+      "loss": 0.5591,
+      "step": 2462
+    },
+    {
+      "epoch": 1.252717909593744,
+      "grad_norm": 2.27122604987324,
+      "learning_rate": 1.347780120838128e-05,
+      "loss": 0.5993,
+      "step": 2463
+    },
+    {
+      "epoch": 1.2532265242545617,
+      "grad_norm": 2.121616362369669,
+      "learning_rate": 1.347254368845873e-05,
+      "loss": 0.523,
+      "step": 2464
+    },
+    {
+      "epoch": 1.2537351389153792,
+      "grad_norm": 2.044099746315767,
+      "learning_rate": 1.3467285076828898e-05,
+      "loss": 0.4298,
+      "step": 2465
+    },
+    {
+      "epoch": 1.2542437535761968,
+      "grad_norm": 1.964035322901111,
+      "learning_rate": 1.3462025375145e-05,
+      "loss": 0.5011,
+      "step": 2466
+    },
+    {
+      "epoch": 1.2547523682370145,
+      "grad_norm": 1.9000534903375217,
+      "learning_rate": 1.3456764585060593e-05,
+      "loss": 0.5011,
+      "step": 2467
+    },
+    {
+      "epoch": 1.255260982897832,
+      "grad_norm": 2.126283930934425,
+      "learning_rate": 1.345150270822958e-05,
+      "loss": 0.5365,
+      "step": 2468
+    },
+    {
+      "epoch": 1.2557695975586496,
+      "grad_norm": 2.0203932318964344,
+      "learning_rate": 1.3446239746306202e-05,
+      "loss": 0.5177,
+      "step": 2469
+    },
+    {
+      "epoch": 1.2562782122194673,
+      "grad_norm": 2.8324910389151574,
+      "learning_rate": 1.3440975700945043e-05,
+      "loss": 0.6077,
+      "step": 2470
+    },
+    {
+      "epoch": 1.2567868268802849,
+      "grad_norm": 1.838715095733319,
+      "learning_rate": 1.3435710573801028e-05,
+      "loss": 0.495,
+      "step": 2471
+    },
+    {
+      "epoch": 1.2572954415411024,
+      "grad_norm": 1.8618302363520556,
+      "learning_rate": 1.3430444366529418e-05,
+      "loss": 0.4499,
+      "step": 2472
+    },
+    {
+      "epoch": 1.2578040562019202,
+      "grad_norm": 2.253361627901441,
+      "learning_rate": 1.342517708078582e-05,
+      "loss": 0.4743,
+      "step": 2473
+    },
+    {
+      "epoch": 1.2583126708627377,
+      "grad_norm": 1.8681445362528293,
+      "learning_rate": 1.3419908718226177e-05,
+      "loss": 0.519,
+      "step": 2474
+    },
+    {
+      "epoch": 1.2588212855235552,
+      "grad_norm": 2.366157735808223,
+      "learning_rate": 1.341463928050677e-05,
+      "loss": 0.5052,
+      "step": 2475
+    },
+    {
+      "epoch": 1.2593299001843727,
+      "grad_norm": 2.328269493176257,
+      "learning_rate": 1.3409368769284218e-05,
+      "loss": 0.5663,
+      "step": 2476
+    },
+    {
+      "epoch": 1.2598385148451905,
+      "grad_norm": 2.0891229545160077,
+      "learning_rate": 1.3404097186215478e-05,
+      "loss": 0.5508,
+      "step": 2477
+    },
+    {
+      "epoch": 1.260347129506008,
+      "grad_norm": 2.2401253924014255,
+      "learning_rate": 1.3398824532957843e-05,
+      "loss": 0.5264,
+      "step": 2478
+    },
+    {
+      "epoch": 1.2608557441668256,
+      "grad_norm": 2.0785991156545442,
+      "learning_rate": 1.3393550811168947e-05,
+      "loss": 0.4988,
+      "step": 2479
+    },
+    {
+      "epoch": 1.261364358827643,
+      "grad_norm": 1.8274646446680156,
+      "learning_rate": 1.3388276022506754e-05,
+      "loss": 0.4302,
+      "step": 2480
+    },
+    {
+      "epoch": 1.2618729734884608,
+      "grad_norm": 2.0608648118163906,
+      "learning_rate": 1.3383000168629563e-05,
+      "loss": 0.546,
+      "step": 2481
+    },
+    {
+      "epoch": 1.2623815881492784,
+      "grad_norm": 1.9300019166373583,
+      "learning_rate": 1.3377723251196013e-05,
+      "loss": 0.5228,
+      "step": 2482
+    },
+    {
+      "epoch": 1.262890202810096,
+      "grad_norm": 2.213052445746488,
+      "learning_rate": 1.3372445271865076e-05,
+      "loss": 0.4805,
+      "step": 2483
+    },
+    {
+      "epoch": 1.2633988174709136,
+      "grad_norm": 2.4515212338185615,
+      "learning_rate": 1.3367166232296058e-05,
+      "loss": 0.5026,
+      "step": 2484
+    },
+    {
+      "epoch": 1.2639074321317312,
+      "grad_norm": 2.1671308118121,
+      "learning_rate": 1.3361886134148592e-05,
+      "loss": 0.5349,
+      "step": 2485
+    },
+    {
+      "epoch": 1.2644160467925487,
+      "grad_norm": 2.0707542092182543,
+      "learning_rate": 1.3356604979082652e-05,
+      "loss": 0.5152,
+      "step": 2486
+    },
+    {
+      "epoch": 1.2649246614533665,
+      "grad_norm": 2.2151533176240994,
+      "learning_rate": 1.3351322768758543e-05,
+      "loss": 0.4428,
+      "step": 2487
+    },
+    {
+      "epoch": 1.265433276114184,
+      "grad_norm": 1.9414231973134797,
+      "learning_rate": 1.3346039504836899e-05,
+      "loss": 0.4988,
+      "step": 2488
+    },
+    {
+      "epoch": 1.2659418907750015,
+      "grad_norm": 2.080325345674027,
+      "learning_rate": 1.3340755188978685e-05,
+      "loss": 0.5525,
+      "step": 2489
+    },
+    {
+      "epoch": 1.2664505054358193,
+      "grad_norm": 2.0124140750715043,
+      "learning_rate": 1.3335469822845198e-05,
+      "loss": 0.5221,
+      "step": 2490
+    },
+    {
+      "epoch": 1.2669591200966368,
+      "grad_norm": 2.6856736475760004,
+      "learning_rate": 1.3330183408098063e-05,
+      "loss": 0.577,
+      "step": 2491
+    },
+    {
+      "epoch": 1.2674677347574543,
+      "grad_norm": 2.0949233453970577,
+      "learning_rate": 1.3324895946399241e-05,
+      "loss": 0.5712,
+      "step": 2492
+    },
+    {
+      "epoch": 1.267976349418272,
+      "grad_norm": 1.9512810509488043,
+      "learning_rate": 1.3319607439411018e-05,
+      "loss": 0.5085,
+      "step": 2493
+    },
+    {
+      "epoch": 1.2684849640790896,
+      "grad_norm": 2.4936263854477914,
+      "learning_rate": 1.3314317888796007e-05,
+      "loss": 0.5632,
+      "step": 2494
+    },
+    {
+      "epoch": 1.2689935787399071,
+      "grad_norm": 2.2628314455164067,
+      "learning_rate": 1.3309027296217147e-05,
+      "loss": 0.4713,
+      "step": 2495
+    },
+    {
+      "epoch": 1.269502193400725,
+      "grad_norm": 2.0874929736356065,
+      "learning_rate": 1.3303735663337713e-05,
+      "loss": 0.603,
+      "step": 2496
+    },
+    {
+      "epoch": 1.2700108080615424,
+      "grad_norm": 2.1436605718438337,
+      "learning_rate": 1.32984429918213e-05,
+      "loss": 0.5276,
+      "step": 2497
+    },
+    {
+      "epoch": 1.27051942272236,
+      "grad_norm": 2.011552466770033,
+      "learning_rate": 1.3293149283331832e-05,
+      "loss": 0.5103,
+      "step": 2498
+    },
+    {
+      "epoch": 1.2710280373831775,
+      "grad_norm": 1.8428090167388678,
+      "learning_rate": 1.3287854539533563e-05,
+      "loss": 0.5241,
+      "step": 2499
+    },
+    {
+      "epoch": 1.2715366520439952,
+      "grad_norm": 1.975020700660094,
+      "learning_rate": 1.3282558762091058e-05,
+      "loss": 0.5507,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2715366520439952,
+      "eval_loss": 0.7237278819084167,
+      "eval_runtime": 170.0311,
+      "eval_samples_per_second": 19.479,
+      "eval_steps_per_second": 9.739,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2720452667048128,
+      "grad_norm": 2.363296281751007,
+      "learning_rate": 1.3277261952669229e-05,
+      "loss": 0.5766,
+      "step": 2501
+    },
+    {
+      "epoch": 1.2725538813656303,
+      "grad_norm": 2.1175437711945424,
+      "learning_rate": 1.3271964112933296e-05,
+      "loss": 0.4639,
+      "step": 2502
+    },
+    {
+      "epoch": 1.2730624960264478,
+      "grad_norm": 1.89405178019831,
+      "learning_rate": 1.3266665244548804e-05,
+      "loss": 0.4345,
+      "step": 2503
+    },
+    {
+      "epoch": 1.2735711106872656,
+      "grad_norm": 2.235843629286078,
+      "learning_rate": 1.326136534918163e-05,
+      "loss": 0.4866,
+      "step": 2504
+    },
+    {
+      "epoch": 1.2740797253480831,
+      "grad_norm": 2.1974607118256038,
+      "learning_rate": 1.3256064428497966e-05,
+      "loss": 0.4806,
+      "step": 2505
+    },
+    {
+      "epoch": 1.2745883400089006,
+      "grad_norm": 2.4226559272084396,
+      "learning_rate": 1.325076248416433e-05,
+      "loss": 0.6268,
+      "step": 2506
+    },
+    {
+      "epoch": 1.2750969546697184,
+      "grad_norm": 2.2646512800151952,
+      "learning_rate": 1.3245459517847562e-05,
+      "loss": 0.5244,
+      "step": 2507
+    },
+    {
+      "epoch": 1.275605569330536,
+      "grad_norm": 1.93890305446829,
+      "learning_rate": 1.324015553121482e-05,
+      "loss": 0.5435,
+      "step": 2508
+    },
+    {
+      "epoch": 1.2761141839913535,
+      "grad_norm": 1.8405805598595144,
+      "learning_rate": 1.3234850525933586e-05,
+      "loss": 0.4533,
+      "step": 2509
+    },
+    {
+      "epoch": 1.2766227986521712,
+      "grad_norm": 2.3670468933849618,
+      "learning_rate": 1.3229544503671665e-05,
+      "loss": 0.5197,
+      "step": 2510
+    },
+    {
+      "epoch": 1.2771314133129887,
+      "grad_norm": 2.4322635450241847,
+      "learning_rate": 1.3224237466097171e-05,
+      "loss": 0.5778,
+      "step": 2511
+    },
+    {
+      "epoch": 1.2776400279738063,
+      "grad_norm": 2.3445474920882856,
+      "learning_rate": 1.3218929414878553e-05,
+      "loss": 0.6203,
+      "step": 2512
+    },
+    {
+      "epoch": 1.278148642634624,
+      "grad_norm": 2.0566983554486487,
+      "learning_rate": 1.3213620351684559e-05,
+      "loss": 0.4669,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2786572572954416,
+      "grad_norm": 2.041104284669751,
+      "learning_rate": 1.3208310278184276e-05,
+      "loss": 0.5199,
+      "step": 2514
+    },
+    {
+      "epoch": 1.279165871956259,
+      "grad_norm": 2.3168497159621206,
+      "learning_rate": 1.320299919604709e-05,
+      "loss": 0.6266,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2796744866170768,
+      "grad_norm": 2.009377904625865,
+      "learning_rate": 1.3197687106942721e-05,
+      "loss": 0.5718,
+      "step": 2516
+    },
+    {
+      "epoch": 1.2801831012778944,
+      "grad_norm": 1.831271666831513,
+      "learning_rate": 1.3192374012541195e-05,
+      "loss": 0.5107,
+      "step": 2517
+    },
+    {
+      "epoch": 1.280691715938712,
+      "grad_norm": 1.969781151442062,
+      "learning_rate": 1.318705991451285e-05,
+      "loss": 0.5053,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2812003305995296,
+      "grad_norm": 1.9092723462968777,
+      "learning_rate": 1.3181744814528354e-05,
+      "loss": 0.5939,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2817089452603472,
+      "grad_norm": 1.9363098796256535,
+      "learning_rate": 1.3176428714258679e-05,
+      "loss": 0.5165,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2822175599211647,
+      "grad_norm": 2.009366869872926,
+      "learning_rate": 1.3171111615375116e-05,
+      "loss": 0.5265,
+      "step": 2521
+    },
+    {
+      "epoch": 1.2827261745819825,
+      "grad_norm": 2.090695558734347,
+      "learning_rate": 1.3165793519549264e-05,
+      "loss": 0.557,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2832347892428,
+      "grad_norm": 2.271995165557822,
+      "learning_rate": 1.3160474428453042e-05,
+      "loss": 0.4791,
+      "step": 2523
+    },
+    {
+      "epoch": 1.2837434039036175,
+      "grad_norm": 2.0294416153867725,
+      "learning_rate": 1.3155154343758682e-05,
+      "loss": 0.5292,
+      "step": 2524
+    },
+    {
+      "epoch": 1.284252018564435,
+      "grad_norm": 2.155818398444436,
+      "learning_rate": 1.3149833267138722e-05,
+      "loss": 0.5273,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2847606332252526,
+      "grad_norm": 2.1754505173200576,
+      "learning_rate": 1.3144511200266022e-05,
+      "loss": 0.5143,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2852692478860703,
+      "grad_norm": 2.0108166496047413,
+      "learning_rate": 1.3139188144813737e-05,
+      "loss": 0.4812,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2857778625468879,
+      "grad_norm": 2.22335294990331,
+      "learning_rate": 1.3133864102455353e-05,
+      "loss": 0.5266,
+      "step": 2528
+    },
+    {
+      "epoch": 1.2862864772077054,
+      "grad_norm": 2.0502290690767353,
+      "learning_rate": 1.312853907486465e-05,
+      "loss": 0.5881,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2867950918685231,
+      "grad_norm": 2.024201591298247,
+      "learning_rate": 1.312321306371573e-05,
+      "loss": 0.5102,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2873037065293407,
+      "grad_norm": 1.7228101458326697,
+      "learning_rate": 1.311788607068299e-05,
+      "loss": 0.4349,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2878123211901582,
+      "grad_norm": 1.8915050129432007,
+      "learning_rate": 1.311255809744115e-05,
+      "loss": 0.4786,
+      "step": 2532
+    },
+    {
+      "epoch": 1.288320935850976,
+      "grad_norm": 2.113089110265145,
+      "learning_rate": 1.3107229145665228e-05,
+      "loss": 0.5208,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2888295505117935,
+      "grad_norm": 1.9300026578341352,
+      "learning_rate": 1.3101899217030557e-05,
+      "loss": 0.5336,
+      "step": 2534
+    },
+    {
+      "epoch": 1.289338165172611,
+      "grad_norm": 2.0513562949359123,
+      "learning_rate": 1.3096568313212777e-05,
+      "loss": 0.4445,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2898467798334288,
+      "grad_norm": 1.7705106403814288,
+      "learning_rate": 1.3091236435887821e-05,
+      "loss": 0.4046,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2903553944942463,
+      "grad_norm": 2.3638666105833077,
+      "learning_rate": 1.3085903586731948e-05,
+      "loss": 0.5923,
+      "step": 2537
+    },
+    {
+      "epoch": 1.2908640091550638,
+      "grad_norm": 1.8235875085157278,
+      "learning_rate": 1.3080569767421706e-05,
+      "loss": 0.4683,
+      "step": 2538
+    },
+    {
+      "epoch": 1.2913726238158816,
+      "grad_norm": 1.7977955989218934,
+      "learning_rate": 1.3075234979633964e-05,
+      "loss": 0.4768,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2918812384766991,
+      "grad_norm": 1.8718020824565487,
+      "learning_rate": 1.3069899225045877e-05,
+      "loss": 0.407,
+      "step": 2540
+    },
+    {
+      "epoch": 1.2923898531375166,
+      "grad_norm": 2.3547694241979817,
+      "learning_rate": 1.3064562505334916e-05,
+      "loss": 0.5208,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2928984677983344,
+      "grad_norm": 2.0702878554694744,
+      "learning_rate": 1.3059224822178857e-05,
+      "loss": 0.5202,
+      "step": 2542
+    },
+    {
+      "epoch": 1.293407082459152,
+      "grad_norm": 2.2141465261986477,
+      "learning_rate": 1.3053886177255772e-05,
+      "loss": 0.5628,
+      "step": 2543
+    },
+    {
+      "epoch": 1.2939156971199695,
+      "grad_norm": 2.1826507224423195,
+      "learning_rate": 1.3048546572244036e-05,
+      "loss": 0.5032,
+      "step": 2544
+    },
+    {
+      "epoch": 1.2944243117807872,
+      "grad_norm": 1.7758855518031074,
+      "learning_rate": 1.3043206008822328e-05,
+      "loss": 0.4818,
+      "step": 2545
+    },
+    {
+      "epoch": 1.2949329264416047,
+      "grad_norm": 2.3734669003602544,
+      "learning_rate": 1.3037864488669632e-05,
+      "loss": 0.6027,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2954415411024223,
+      "grad_norm": 2.204954429380465,
+      "learning_rate": 1.3032522013465223e-05,
+      "loss": 0.5162,
+      "step": 2547
+    },
+    {
+      "epoch": 1.2959501557632398,
+      "grad_norm": 2.021158002223575,
+      "learning_rate": 1.302717858488869e-05,
+      "loss": 0.5069,
+      "step": 2548
+    },
+    {
+      "epoch": 1.2964587704240575,
+      "grad_norm": 2.0718325235146695,
+      "learning_rate": 1.3021834204619908e-05,
+      "loss": 0.6114,
+      "step": 2549
+    },
+    {
+      "epoch": 1.296967385084875,
+      "grad_norm": 2.1394779445566017,
+      "learning_rate": 1.3016488874339057e-05,
+      "loss": 0.5114,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2974759997456926,
+      "grad_norm": 2.1967086967804224,
+      "learning_rate": 1.301114259572662e-05,
+      "loss": 0.5349,
+      "step": 2551
+    },
+    {
+      "epoch": 1.2979846144065101,
+      "grad_norm": 1.99658041916892,
+      "learning_rate": 1.300579537046337e-05,
+      "loss": 0.503,
+      "step": 2552
+    },
+    {
+      "epoch": 1.298493229067328,
+      "grad_norm": 2.1944163383690602,
+      "learning_rate": 1.3000447200230385e-05,
+      "loss": 0.5384,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2990018437281454,
+      "grad_norm": 2.055153565435136,
+      "learning_rate": 1.299509808670903e-05,
+      "loss": 0.4681,
+      "step": 2554
+    },
+    {
+      "epoch": 1.299510458388963,
+      "grad_norm": 2.527241676781009,
+      "learning_rate": 1.2989748031580981e-05,
+      "loss": 0.5458,
+      "step": 2555
+    },
+    {
+      "epoch": 1.3000190730497807,
+      "grad_norm": 2.3200445052514955,
+      "learning_rate": 1.29843970365282e-05,
+      "loss": 0.5245,
+      "step": 2556
+    },
+    {
+      "epoch": 1.3005276877105982,
+      "grad_norm": 1.9351973232427395,
+      "learning_rate": 1.2979045103232945e-05,
+      "loss": 0.4658,
+      "step": 2557
+    },
+    {
+      "epoch": 1.3010363023714158,
+      "grad_norm": 2.1560998394413837,
+      "learning_rate": 1.297369223337777e-05,
+      "loss": 0.5771,
+      "step": 2558
+    },
+    {
+      "epoch": 1.3015449170322335,
+      "grad_norm": 1.8694052832729082,
+      "learning_rate": 1.2968338428645529e-05,
+      "loss": 0.4921,
+      "step": 2559
+    },
+    {
+      "epoch": 1.302053531693051,
+      "grad_norm": 1.8952316405304392,
+      "learning_rate": 1.2962983690719357e-05,
+      "loss": 0.5657,
+      "step": 2560
+    },
+    {
+      "epoch": 1.3025621463538686,
+      "grad_norm": 2.060350627502987,
+      "learning_rate": 1.2957628021282695e-05,
+      "loss": 0.5142,
+      "step": 2561
+    },
+    {
+      "epoch": 1.3030707610146863,
+      "grad_norm": 2.1433495773166578,
+      "learning_rate": 1.2952271422019275e-05,
+      "loss": 0.5012,
+      "step": 2562
+    },
+    {
+      "epoch": 1.3035793756755039,
+      "grad_norm": 2.1004449645425765,
+      "learning_rate": 1.2946913894613116e-05,
+      "loss": 0.4831,
+      "step": 2563
+    },
+    {
+      "epoch": 1.3040879903363214,
+      "grad_norm": 2.1221530018691594,
+      "learning_rate": 1.2941555440748527e-05,
+      "loss": 0.5074,
+      "step": 2564
+    },
+    {
+      "epoch": 1.3045966049971391,
+      "grad_norm": 1.8672045463518103,
+      "learning_rate": 1.293619606211012e-05,
+      "loss": 0.489,
+      "step": 2565
+    },
+    {
+      "epoch": 1.3051052196579567,
+      "grad_norm": 1.9857618636913066,
+      "learning_rate": 1.2930835760382785e-05,
+      "loss": 0.4951,
+      "step": 2566
+    },
+    {
+      "epoch": 1.3056138343187742,
+      "grad_norm": 1.9657133995276732,
+      "learning_rate": 1.2925474537251712e-05,
+      "loss": 0.4809,
+      "step": 2567
+    },
+    {
+      "epoch": 1.306122448979592,
+      "grad_norm": 1.698426300826082,
+      "learning_rate": 1.292011239440237e-05,
+      "loss": 0.4664,
+      "step": 2568
+    },
+    {
+      "epoch": 1.3066310636404095,
+      "grad_norm": 2.2614775481272402,
+      "learning_rate": 1.2914749333520532e-05,
+      "loss": 0.4755,
+      "step": 2569
+    },
+    {
+      "epoch": 1.307139678301227,
+      "grad_norm": 2.133137325382027,
+      "learning_rate": 1.2909385356292242e-05,
+      "loss": 0.5056,
+      "step": 2570
+    },
+    {
+      "epoch": 1.3076482929620448,
+      "grad_norm": 2.4297428431448207,
+      "learning_rate": 1.2904020464403843e-05,
+      "loss": 0.5094,
+      "step": 2571
+    },
+    {
+      "epoch": 1.3081569076228623,
+      "grad_norm": 2.401698973432011,
+      "learning_rate": 1.2898654659541969e-05,
+      "loss": 0.5926,
+      "step": 2572
+    },
+    {
+      "epoch": 1.3086655222836798,
+      "grad_norm": 1.964223107665474,
+      "learning_rate": 1.2893287943393533e-05,
+      "loss": 0.5999,
+      "step": 2573
+    },
+    {
+      "epoch": 1.3091741369444974,
+      "grad_norm": 1.9423834262269644,
+      "learning_rate": 1.2887920317645735e-05,
+      "loss": 0.518,
+      "step": 2574
+    },
+    {
+      "epoch": 1.3096827516053149,
+      "grad_norm": 1.9875937361775327,
+      "learning_rate": 1.2882551783986066e-05,
+      "loss": 0.4639,
+      "step": 2575
+    },
+    {
+      "epoch": 1.3101913662661326,
+      "grad_norm": 2.21742853920415,
+      "learning_rate": 1.2877182344102297e-05,
+      "loss": 0.5763,
+      "step": 2576
+    },
+    {
+      "epoch": 1.3106999809269502,
+      "grad_norm": 2.181827492766309,
+      "learning_rate": 1.2871811999682487e-05,
+      "loss": 0.5089,
+      "step": 2577
+    },
+    {
+      "epoch": 1.3112085955877677,
+      "grad_norm": 2.3017043640379082,
+      "learning_rate": 1.286644075241498e-05,
+      "loss": 0.5736,
+      "step": 2578
+    },
+    {
+      "epoch": 1.3117172102485855,
+      "grad_norm": 1.7928153577026764,
+      "learning_rate": 1.28610686039884e-05,
+      "loss": 0.505,
+      "step": 2579
+    },
+    {
+      "epoch": 1.312225824909403,
+      "grad_norm": 2.4574607873933467,
+      "learning_rate": 1.2855695556091662e-05,
+      "loss": 0.5091,
+      "step": 2580
+    },
+    {
+      "epoch": 1.3127344395702205,
+      "grad_norm": 1.8488100090888278,
+      "learning_rate": 1.2850321610413957e-05,
+      "loss": 0.5286,
+      "step": 2581
+    },
+    {
+      "epoch": 1.3132430542310383,
+      "grad_norm": 1.950824695205738,
+      "learning_rate": 1.2844946768644753e-05,
+      "loss": 0.4058,
+      "step": 2582
+    },
+    {
+      "epoch": 1.3137516688918558,
+      "grad_norm": 2.3642623504254776,
+      "learning_rate": 1.2839571032473814e-05,
+      "loss": 0.5829,
+      "step": 2583
+    },
+    {
+      "epoch": 1.3142602835526733,
+      "grad_norm": 2.0259074450229853,
+      "learning_rate": 1.2834194403591177e-05,
+      "loss": 0.5166,
+      "step": 2584
+    },
+    {
+      "epoch": 1.314768898213491,
+      "grad_norm": 1.8663846767962347,
+      "learning_rate": 1.2828816883687163e-05,
+      "loss": 0.402,
+      "step": 2585
+    },
+    {
+      "epoch": 1.3152775128743086,
+      "grad_norm": 2.7372706312823025,
+      "learning_rate": 1.2823438474452363e-05,
+      "loss": 0.5379,
+      "step": 2586
+    },
+    {
+      "epoch": 1.3157861275351261,
+      "grad_norm": 2.343581740379527,
+      "learning_rate": 1.2818059177577657e-05,
+      "loss": 0.5244,
+      "step": 2587
+    },
+    {
+      "epoch": 1.316294742195944,
+      "grad_norm": 2.440719630010152,
+      "learning_rate": 1.281267899475421e-05,
+      "loss": 0.5903,
+      "step": 2588
+    },
+    {
+      "epoch": 1.3168033568567614,
+      "grad_norm": 2.367455999172956,
+      "learning_rate": 1.2807297927673452e-05,
+      "loss": 0.5628,
+      "step": 2589
+    },
+    {
+      "epoch": 1.317311971517579,
+      "grad_norm": 2.115185230702187,
+      "learning_rate": 1.2801915978027095e-05,
+      "loss": 0.4656,
+      "step": 2590
+    },
+    {
+      "epoch": 1.3178205861783967,
+      "grad_norm": 1.724731717463009,
+      "learning_rate": 1.2796533147507132e-05,
+      "loss": 0.5121,
+      "step": 2591
+    },
+    {
+      "epoch": 1.3183292008392142,
+      "grad_norm": 1.8287039598052557,
+      "learning_rate": 1.2791149437805834e-05,
+      "loss": 0.4972,
+      "step": 2592
+    },
+    {
+      "epoch": 1.3188378155000318,
+      "grad_norm": 2.0041357433235993,
+      "learning_rate": 1.2785764850615748e-05,
+      "loss": 0.5114,
+      "step": 2593
+    },
+    {
+      "epoch": 1.3193464301608495,
+      "grad_norm": 2.1226159354134793,
+      "learning_rate": 1.2780379387629683e-05,
+      "loss": 0.5023,
+      "step": 2594
+    },
+    {
+      "epoch": 1.319855044821667,
+      "grad_norm": 2.3897569863218973,
+      "learning_rate": 1.2774993050540745e-05,
+      "loss": 0.5972,
+      "step": 2595
+    },
+    {
+      "epoch": 1.3203636594824846,
+      "grad_norm": 2.9091739507223564,
+      "learning_rate": 1.2769605841042299e-05,
+      "loss": 0.5442,
+      "step": 2596
+    },
+    {
+      "epoch": 1.320872274143302,
+      "grad_norm": 2.015449930320984,
+      "learning_rate": 1.2764217760827996e-05,
+      "loss": 0.4651,
+      "step": 2597
+    },
+    {
+      "epoch": 1.3213808888041199,
+      "grad_norm": 2.1886120830418125,
+      "learning_rate": 1.2758828811591748e-05,
+      "loss": 0.6337,
+      "step": 2598
+    },
+    {
+      "epoch": 1.3218895034649374,
+      "grad_norm": 1.9279404735833485,
+      "learning_rate": 1.275343899502775e-05,
+      "loss": 0.4555,
+      "step": 2599
+    },
+    {
+      "epoch": 1.322398118125755,
+      "grad_norm": 1.9114294509022214,
+      "learning_rate": 1.2748048312830465e-05,
+      "loss": 0.5852,
+      "step": 2600
+    },
+    {
+      "epoch": 1.322398118125755,
+      "eval_loss": 0.7159717082977295,
+      "eval_runtime": 170.0296,
+      "eval_samples_per_second": 19.479,
+      "eval_steps_per_second": 9.739,
+      "step": 2600
+    },
+    {
+      "epoch": 1.3229067327865724,
+      "grad_norm": 2.105146337392524,
+      "learning_rate": 1.2742656766694634e-05,
+      "loss": 0.6015,
+      "step": 2601
+    },
+    {
+      "epoch": 1.3234153474473902,
+      "grad_norm": 2.0848860358691685,
+      "learning_rate": 1.273726435831526e-05,
+      "loss": 0.5081,
+      "step": 2602
+    },
+    {
+      "epoch": 1.3239239621082077,
+      "grad_norm": 1.8524714303226766,
+      "learning_rate": 1.2731871089387626e-05,
+      "loss": 0.4369,
+      "step": 2603
+    },
+    {
+      "epoch": 1.3244325767690253,
+      "grad_norm": 2.3300419026810464,
+      "learning_rate": 1.2726476961607278e-05,
+      "loss": 0.5383,
+      "step": 2604
+    },
+    {
+      "epoch": 1.324941191429843,
+      "grad_norm": 1.8858242402283834,
+      "learning_rate": 1.2721081976670038e-05,
+      "loss": 0.4546,
+      "step": 2605
+    },
+    {
+      "epoch": 1.3254498060906605,
+      "grad_norm": 2.3532126077299,
+      "learning_rate": 1.2715686136271996e-05,
+      "loss": 0.5524,
+      "step": 2606
+    },
+    {
+      "epoch": 1.325958420751478,
+      "grad_norm": 2.083216574893882,
+      "learning_rate": 1.2710289442109512e-05,
+      "loss": 0.4592,
+      "step": 2607
+    },
+    {
+      "epoch": 1.3264670354122958,
+      "grad_norm": 2.845687876642829,
+      "learning_rate": 1.270489189587921e-05,
+      "loss": 0.5834,
+      "step": 2608
+    },
+    {
+      "epoch": 1.3269756500731134,
+      "grad_norm": 1.9694092872804312,
+      "learning_rate": 1.2699493499277984e-05,
+      "loss": 0.4496,
+      "step": 2609
+    },
+    {
+      "epoch": 1.3274842647339309,
+      "grad_norm": 2.1304249797111483,
+      "learning_rate": 1.2694094254002995e-05,
+      "loss": 0.5261,
+      "step": 2610
+    },
+    {
+      "epoch": 1.3279928793947486,
+      "grad_norm": 2.1087031672081773,
+      "learning_rate": 1.2688694161751678e-05,
+      "loss": 0.5087,
+      "step": 2611
+    },
+    {
+      "epoch": 1.3285014940555662,
+      "grad_norm": 2.2330583013977505,
+      "learning_rate": 1.268329322422172e-05,
+      "loss": 0.5237,
+      "step": 2612
+    },
+    {
+      "epoch": 1.3290101087163837,
+      "grad_norm": 2.066943381867935,
+      "learning_rate": 1.2677891443111085e-05,
+      "loss": 0.4812,
+      "step": 2613
+    },
+    {
+      "epoch": 1.3295187233772015,
+      "grad_norm": 2.024616619599835,
+      "learning_rate": 1.2672488820118003e-05,
+      "loss": 0.5148,
+      "step": 2614
+    },
+    {
+      "epoch": 1.330027338038019,
+      "grad_norm": 1.9873954433553613,
+      "learning_rate": 1.2667085356940958e-05,
+      "loss": 0.4675,
+      "step": 2615
+    },
+    {
+      "epoch": 1.3305359526988365,
+      "grad_norm": 2.0150433078021215,
+      "learning_rate": 1.266168105527871e-05,
+      "loss": 0.4805,
+      "step": 2616
+    },
+    {
+      "epoch": 1.3310445673596543,
+      "grad_norm": 2.0672813250066358,
+      "learning_rate": 1.2656275916830274e-05,
+      "loss": 0.523,
+      "step": 2617
+    },
+    {
+      "epoch": 1.3315531820204718,
+      "grad_norm": 2.1033988785644513,
+      "learning_rate": 1.2650869943294933e-05,
+      "loss": 0.4818,
+      "step": 2618
+    },
+    {
+      "epoch": 1.3320617966812893,
+      "grad_norm": 1.9938843923972527,
+      "learning_rate": 1.2645463136372227e-05,
+      "loss": 0.4819,
+      "step": 2619
+    },
+    {
+      "epoch": 1.3325704113421069,
+      "grad_norm": 2.1487814887970065,
+      "learning_rate": 1.2640055497761969e-05,
+      "loss": 0.6374,
+      "step": 2620
+    },
+    {
+      "epoch": 1.3330790260029246,
+      "grad_norm": 1.8753242212357355,
+      "learning_rate": 1.2634647029164221e-05,
+      "loss": 0.4625,
+      "step": 2621
+    },
+    {
+      "epoch": 1.3335876406637421,
+      "grad_norm": 1.9629775298509111,
+      "learning_rate": 1.2629237732279314e-05,
+      "loss": 0.521,
+      "step": 2622
+    },
+    {
+      "epoch": 1.3340962553245597,
+      "grad_norm": 2.1208009184013346,
+      "learning_rate": 1.2623827608807834e-05,
+      "loss": 0.4905,
+      "step": 2623
+    },
+    {
+      "epoch": 1.3346048699853772,
+      "grad_norm": 2.3504045746921713,
+      "learning_rate": 1.2618416660450637e-05,
+      "loss": 0.4943,
+      "step": 2624
+    },
+    {
+      "epoch": 1.335113484646195,
+      "grad_norm": 2.4387501176279764,
+      "learning_rate": 1.2613004888908823e-05,
+      "loss": 0.5789,
+      "step": 2625
+    },
+    {
+      "epoch": 1.3356220993070125,
+      "grad_norm": 2.1657029967691424,
+      "learning_rate": 1.2607592295883761e-05,
+      "loss": 0.4527,
+      "step": 2626
+    },
+    {
+      "epoch": 1.33613071396783,
+      "grad_norm": 2.674856028737827,
+      "learning_rate": 1.2602178883077075e-05,
+      "loss": 0.606,
+      "step": 2627
+    },
+    {
+      "epoch": 1.3366393286286478,
+      "grad_norm": 2.338113339172506,
+      "learning_rate": 1.2596764652190649e-05,
+      "loss": 0.507,
+      "step": 2628
+    },
+    {
+      "epoch": 1.3371479432894653,
+      "grad_norm": 1.916528973263615,
+      "learning_rate": 1.2591349604926626e-05,
+      "loss": 0.4757,
+      "step": 2629
+    },
+    {
+      "epoch": 1.3376565579502828,
+      "grad_norm": 1.9710023499721103,
+      "learning_rate": 1.2585933742987399e-05,
+      "loss": 0.5173,
+      "step": 2630
+    },
+    {
+      "epoch": 1.3381651726111006,
+      "grad_norm": 2.267963439115219,
+      "learning_rate": 1.258051706807562e-05,
+      "loss": 0.5616,
+      "step": 2631
+    },
+    {
+      "epoch": 1.338673787271918,
+      "grad_norm": 1.9447021752560356,
+      "learning_rate": 1.25750995818942e-05,
+      "loss": 0.4338,
+      "step": 2632
+    },
+    {
+      "epoch": 1.3391824019327356,
+      "grad_norm": 1.8897125312127658,
+      "learning_rate": 1.25696812861463e-05,
+      "loss": 0.4877,
+      "step": 2633
+    },
+    {
+      "epoch": 1.3396910165935534,
+      "grad_norm": 2.20373396877921,
+      "learning_rate": 1.2564262182535341e-05,
+      "loss": 0.5232,
+      "step": 2634
+    },
+    {
+      "epoch": 1.340199631254371,
+      "grad_norm": 2.062579297939967,
+      "learning_rate": 1.2558842272764991e-05,
+      "loss": 0.5219,
+      "step": 2635
+    },
+    {
+      "epoch": 1.3407082459151884,
+      "grad_norm": 2.0661514788596094,
+      "learning_rate": 1.2553421558539177e-05,
+      "loss": 0.5871,
+      "step": 2636
+    },
+    {
+      "epoch": 1.3412168605760062,
+      "grad_norm": 2.410421941076084,
+      "learning_rate": 1.2548000041562075e-05,
+      "loss": 0.5057,
+      "step": 2637
+    },
+    {
+      "epoch": 1.3417254752368237,
+      "grad_norm": 1.8215259878570156,
+      "learning_rate": 1.2542577723538118e-05,
+      "loss": 0.4696,
+      "step": 2638
+    },
+    {
+      "epoch": 1.3422340898976413,
+      "grad_norm": 2.4120631163327486,
+      "learning_rate": 1.253715460617199e-05,
+      "loss": 0.555,
+      "step": 2639
+    },
+    {
+      "epoch": 1.342742704558459,
+      "grad_norm": 2.0182359920919533,
+      "learning_rate": 1.2531730691168622e-05,
+      "loss": 0.5185,
+      "step": 2640
+    },
+    {
+      "epoch": 1.3432513192192765,
+      "grad_norm": 1.9639345647862005,
+      "learning_rate": 1.2526305980233199e-05,
+      "loss": 0.4428,
+      "step": 2641
+    },
+    {
+      "epoch": 1.343759933880094,
+      "grad_norm": 2.452734752156216,
+      "learning_rate": 1.2520880475071157e-05,
+      "loss": 0.4871,
+      "step": 2642
+    },
+    {
+      "epoch": 1.3442685485409118,
+      "grad_norm": 2.1525374808243782,
+      "learning_rate": 1.2515454177388182e-05,
+      "loss": 0.5207,
+      "step": 2643
+    },
+    {
+      "epoch": 1.3447771632017294,
+      "grad_norm": 2.1173463881440076,
+      "learning_rate": 1.2510027088890202e-05,
+      "loss": 0.4421,
+      "step": 2644
+    },
+    {
+      "epoch": 1.3452857778625469,
+      "grad_norm": 2.540784140619752,
+      "learning_rate": 1.2504599211283401e-05,
+      "loss": 0.5071,
+      "step": 2645
+    },
+    {
+      "epoch": 1.3457943925233644,
+      "grad_norm": 1.8922185566099399,
+      "learning_rate": 1.2499170546274211e-05,
+      "loss": 0.5457,
+      "step": 2646
+    },
+    {
+      "epoch": 1.3463030071841822,
+      "grad_norm": 2.1979159193956694,
+      "learning_rate": 1.2493741095569307e-05,
+      "loss": 0.4762,
+      "step": 2647
+    },
+    {
+      "epoch": 1.3468116218449997,
+      "grad_norm": 2.0971064068925624,
+      "learning_rate": 1.2488310860875622e-05,
+      "loss": 0.4733,
+      "step": 2648
+    },
+    {
+      "epoch": 1.3473202365058172,
+      "grad_norm": 3.2176184285501606,
+      "learning_rate": 1.2482879843900316e-05,
+      "loss": 0.593,
+      "step": 2649
+    },
+    {
+      "epoch": 1.3478288511666348,
+      "grad_norm": 1.7439903886600194,
+      "learning_rate": 1.2477448046350812e-05,
+      "loss": 0.5119,
+      "step": 2650
+    },
+    {
+      "epoch": 1.3483374658274525,
+      "grad_norm": 2.1130050506892504,
+      "learning_rate": 1.2472015469934771e-05,
+      "loss": 0.4923,
+      "step": 2651
+    },
+    {
+      "epoch": 1.34884608048827,
+      "grad_norm": 2.0100477074422507,
+      "learning_rate": 1.24665821163601e-05,
+      "loss": 0.5723,
+      "step": 2652
+    },
+    {
+      "epoch": 1.3493546951490876,
+      "grad_norm": 2.0415928366669935,
+      "learning_rate": 1.2461147987334956e-05,
+      "loss": 0.5189,
+      "step": 2653
+    },
+    {
+      "epoch": 1.3498633098099053,
+      "grad_norm": 1.8990304279993213,
+      "learning_rate": 1.245571308456773e-05,
+      "loss": 0.4591,
+      "step": 2654
+    },
+    {
+      "epoch": 1.3503719244707229,
+      "grad_norm": 1.9456553942116528,
+      "learning_rate": 1.2450277409767058e-05,
+      "loss": 0.5557,
+      "step": 2655
+    },
+    {
+      "epoch": 1.3508805391315404,
+      "grad_norm": 1.8846891555690308,
+      "learning_rate": 1.2444840964641826e-05,
+      "loss": 0.4745,
+      "step": 2656
+    },
+    {
+      "epoch": 1.3513891537923581,
+      "grad_norm": 2.0470117421374825,
+      "learning_rate": 1.2439403750901158e-05,
+      "loss": 0.531,
+      "step": 2657
+    },
+    {
+      "epoch": 1.3518977684531757,
+      "grad_norm": 2.0065477001795156,
+      "learning_rate": 1.2433965770254417e-05,
+      "loss": 0.3917,
+      "step": 2658
+    },
+    {
+      "epoch": 1.3524063831139932,
+      "grad_norm": 2.7249803734868756,
+      "learning_rate": 1.242852702441121e-05,
+      "loss": 0.5046,
+      "step": 2659
+    },
+    {
+      "epoch": 1.352914997774811,
+      "grad_norm": 2.0174044066215275,
+      "learning_rate": 1.2423087515081385e-05,
+      "loss": 0.542,
+      "step": 2660
+    },
+    {
+      "epoch": 1.3534236124356285,
+      "grad_norm": 2.331372243013102,
+      "learning_rate": 1.2417647243975031e-05,
+      "loss": 0.5307,
+      "step": 2661
+    },
+    {
+      "epoch": 1.353932227096446,
+      "grad_norm": 2.2063131869077757,
+      "learning_rate": 1.2412206212802468e-05,
+      "loss": 0.5296,
+      "step": 2662
+    },
+    {
+      "epoch": 1.3544408417572638,
+      "grad_norm": 2.192495038191119,
+      "learning_rate": 1.2406764423274269e-05,
+      "loss": 0.5073,
+      "step": 2663
+    },
+    {
+      "epoch": 1.3549494564180813,
+      "grad_norm": 1.928351430101908,
+      "learning_rate": 1.2401321877101231e-05,
+      "loss": 0.5901,
+      "step": 2664
+    },
+    {
+      "epoch": 1.3554580710788988,
+      "grad_norm": 2.27899131263058,
+      "learning_rate": 1.23958785759944e-05,
+      "loss": 0.5715,
+      "step": 2665
+    },
+    {
+      "epoch": 1.3559666857397166,
+      "grad_norm": 2.4058596802253116,
+      "learning_rate": 1.2390434521665058e-05,
+      "loss": 0.5373,
+      "step": 2666
+    },
+    {
+      "epoch": 1.356475300400534,
+      "grad_norm": 1.9415308952614272,
+      "learning_rate": 1.2384989715824714e-05,
+      "loss": 0.5003,
+      "step": 2667
+    },
+    {
+      "epoch": 1.3569839150613516,
+      "grad_norm": 2.0376021611163657,
+      "learning_rate": 1.2379544160185122e-05,
+      "loss": 0.4761,
+      "step": 2668
+    },
+    {
+      "epoch": 1.3574925297221692,
+      "grad_norm": 2.174179720485456,
+      "learning_rate": 1.2374097856458272e-05,
+      "loss": 0.487,
+      "step": 2669
+    },
+    {
+      "epoch": 1.358001144382987,
+      "grad_norm": 1.9099643353432874,
+      "learning_rate": 1.2368650806356386e-05,
+      "loss": 0.5648,
+      "step": 2670
+    },
+    {
+      "epoch": 1.3585097590438044,
+      "grad_norm": 1.949089581141026,
+      "learning_rate": 1.2363203011591926e-05,
+      "loss": 0.5202,
+      "step": 2671
+    },
+    {
+      "epoch": 1.359018373704622,
+      "grad_norm": 2.1758060145723737,
+      "learning_rate": 1.2357754473877576e-05,
+      "loss": 0.5422,
+      "step": 2672
+    },
+    {
+      "epoch": 1.3595269883654395,
+      "grad_norm": 2.0263922485272583,
+      "learning_rate": 1.2352305194926265e-05,
+      "loss": 0.5791,
+      "step": 2673
+    },
+    {
+      "epoch": 1.3600356030262573,
+      "grad_norm": 2.4936276283929146,
+      "learning_rate": 1.2346855176451151e-05,
+      "loss": 0.5676,
+      "step": 2674
+    },
+    {
+      "epoch": 1.3605442176870748,
+      "grad_norm": 2.38244105711453,
+      "learning_rate": 1.2341404420165628e-05,
+      "loss": 0.4494,
+      "step": 2675
+    },
+    {
+      "epoch": 1.3610528323478923,
+      "grad_norm": 2.3035590043688674,
+      "learning_rate": 1.2335952927783318e-05,
+      "loss": 0.4922,
+      "step": 2676
+    },
+    {
+      "epoch": 1.36156144700871,
+      "grad_norm": 2.28793400884042,
+      "learning_rate": 1.2330500701018069e-05,
+      "loss": 0.6992,
+      "step": 2677
+    },
+    {
+      "epoch": 1.3620700616695276,
+      "grad_norm": 1.9820835369763026,
+      "learning_rate": 1.2325047741583972e-05,
+      "loss": 0.5369,
+      "step": 2678
+    },
+    {
+      "epoch": 1.3625786763303451,
+      "grad_norm": 2.492564970985214,
+      "learning_rate": 1.2319594051195347e-05,
+      "loss": 0.5556,
+      "step": 2679
+    },
+    {
+      "epoch": 1.3630872909911629,
+      "grad_norm": 1.6159484834579925,
+      "learning_rate": 1.2314139631566728e-05,
+      "loss": 0.441,
+      "step": 2680
+    },
+    {
+      "epoch": 1.3635959056519804,
+      "grad_norm": 2.293216516473,
+      "learning_rate": 1.2308684484412899e-05,
+      "loss": 0.5167,
+      "step": 2681
+    },
+    {
+      "epoch": 1.364104520312798,
+      "grad_norm": 1.9657765292177762,
+      "learning_rate": 1.2303228611448857e-05,
+      "loss": 0.4378,
+      "step": 2682
+    },
+    {
+      "epoch": 1.3646131349736157,
+      "grad_norm": 1.9053084127323807,
+      "learning_rate": 1.2297772014389836e-05,
+      "loss": 0.5241,
+      "step": 2683
+    },
+    {
+      "epoch": 1.3651217496344332,
+      "grad_norm": 2.0865742715158158,
+      "learning_rate": 1.2292314694951298e-05,
+      "loss": 0.5301,
+      "step": 2684
+    },
+    {
+      "epoch": 1.3656303642952508,
+      "grad_norm": 2.041554298682201,
+      "learning_rate": 1.2286856654848924e-05,
+      "loss": 0.5105,
+      "step": 2685
+    },
+    {
+      "epoch": 1.3661389789560685,
+      "grad_norm": 1.893958187934798,
+      "learning_rate": 1.228139789579863e-05,
+      "loss": 0.5379,
+      "step": 2686
+    },
+    {
+      "epoch": 1.366647593616886,
+      "grad_norm": 2.418681346303406,
+      "learning_rate": 1.2275938419516553e-05,
+      "loss": 0.5016,
+      "step": 2687
+    },
+    {
+      "epoch": 1.3671562082777036,
+      "grad_norm": 1.9198576721446632,
+      "learning_rate": 1.227047822771906e-05,
+      "loss": 0.5321,
+      "step": 2688
+    },
+    {
+      "epoch": 1.3676648229385213,
+      "grad_norm": 2.1779804723615594,
+      "learning_rate": 1.2265017322122739e-05,
+      "loss": 0.4725,
+      "step": 2689
+    },
+    {
+      "epoch": 1.3681734375993388,
+      "grad_norm": 2.0167273530682746,
+      "learning_rate": 1.2259555704444404e-05,
+      "loss": 0.5625,
+      "step": 2690
+    },
+    {
+      "epoch": 1.3686820522601564,
+      "grad_norm": 2.2026816929881177,
+      "learning_rate": 1.2254093376401087e-05,
+      "loss": 0.4759,
+      "step": 2691
+    },
+    {
+      "epoch": 1.3691906669209741,
+      "grad_norm": 2.184486274995687,
+      "learning_rate": 1.2248630339710057e-05,
+      "loss": 0.4544,
+      "step": 2692
+    },
+    {
+      "epoch": 1.3696992815817917,
+      "grad_norm": 2.3342389211351917,
+      "learning_rate": 1.2243166596088796e-05,
+      "loss": 0.5611,
+      "step": 2693
+    },
+    {
+      "epoch": 1.3702078962426092,
+      "grad_norm": 2.357391331940189,
+      "learning_rate": 1.2237702147255006e-05,
+      "loss": 0.5427,
+      "step": 2694
+    },
+    {
+      "epoch": 1.3707165109034267,
+      "grad_norm": 2.3513305144084446,
+      "learning_rate": 1.2232236994926616e-05,
+      "loss": 0.4676,
+      "step": 2695
+    },
+    {
+      "epoch": 1.3712251255642443,
+      "grad_norm": 2.2610102517234276,
+      "learning_rate": 1.2226771140821776e-05,
+      "loss": 0.6016,
+      "step": 2696
+    },
+    {
+      "epoch": 1.371733740225062,
+      "grad_norm": 2.4087584660395662,
+      "learning_rate": 1.2221304586658855e-05,
+      "loss": 0.5097,
+      "step": 2697
+    },
+    {
+      "epoch": 1.3722423548858795,
+      "grad_norm": 1.9507361489713164,
+      "learning_rate": 1.2215837334156446e-05,
+      "loss": 0.5846,
+      "step": 2698
+    },
+    {
+      "epoch": 1.372750969546697,
+      "grad_norm": 2.0035311996738177,
+      "learning_rate": 1.2210369385033352e-05,
+      "loss": 0.4314,
+      "step": 2699
+    },
+    {
+      "epoch": 1.3732595842075148,
+      "grad_norm": 2.146885314650099,
+      "learning_rate": 1.22049007410086e-05,
+      "loss": 0.5252,
+      "step": 2700
+    },
+    {
+      "epoch": 1.3732595842075148,
+      "eval_loss": 0.7148016095161438,
+      "eval_runtime": 170.3971,
+      "eval_samples_per_second": 19.437,
+      "eval_steps_per_second": 9.718,
+      "step": 2700
+    },
+    {
+      "epoch": 1.3737681988683323,
+      "grad_norm": 1.8654600955536484,
+      "learning_rate": 1.2199431403801443e-05,
+      "loss": 0.3981,
+      "step": 2701
+    },
+    {
+      "epoch": 1.3742768135291499,
+      "grad_norm": 2.099739957785207,
+      "learning_rate": 1.2193961375131345e-05,
+      "loss": 0.5676,
+      "step": 2702
+    },
+    {
+      "epoch": 1.3747854281899676,
+      "grad_norm": 1.9839844807510063,
+      "learning_rate": 1.218849065671798e-05,
+      "loss": 0.4478,
+      "step": 2703
+    },
+    {
+      "epoch": 1.3752940428507852,
+      "grad_norm": 2.0763291939015245,
+      "learning_rate": 1.2183019250281253e-05,
+      "loss": 0.5166,
+      "step": 2704
+    },
+    {
+      "epoch": 1.3758026575116027,
+      "grad_norm": 2.171270601990708,
+      "learning_rate": 1.217754715754128e-05,
+      "loss": 0.6072,
+      "step": 2705
+    },
+    {
+      "epoch": 1.3763112721724204,
+      "grad_norm": 2.03465830153166,
+      "learning_rate": 1.2172074380218386e-05,
+      "loss": 0.4417,
+      "step": 2706
+    },
+    {
+      "epoch": 1.376819886833238,
+      "grad_norm": 1.9968423354174107,
+      "learning_rate": 1.2166600920033123e-05,
+      "loss": 0.4719,
+      "step": 2707
+    },
+    {
+      "epoch": 1.3773285014940555,
+      "grad_norm": 2.080774668807527,
+      "learning_rate": 1.2161126778706244e-05,
+      "loss": 0.5403,
+      "step": 2708
+    },
+    {
+      "epoch": 1.3778371161548733,
+      "grad_norm": 2.079493824734316,
+      "learning_rate": 1.2155651957958732e-05,
+      "loss": 0.4841,
+      "step": 2709
+    },
+    {
+      "epoch": 1.3783457308156908,
+      "grad_norm": 2.0060533468772905,
+      "learning_rate": 1.2150176459511771e-05,
+      "loss": 0.6105,
+      "step": 2710
+    },
+    {
+      "epoch": 1.3788543454765083,
+      "grad_norm": 2.129956467591065,
+      "learning_rate": 1.2144700285086766e-05,
+      "loss": 0.4795,
+      "step": 2711
+    },
+    {
+      "epoch": 1.379362960137326,
+      "grad_norm": 1.9508459603695234,
+      "learning_rate": 1.2139223436405325e-05,
+      "loss": 0.4783,
+      "step": 2712
+    },
+    {
+      "epoch": 1.3798715747981436,
+      "grad_norm": 2.225588508527475,
+      "learning_rate": 1.2133745915189279e-05,
+      "loss": 0.5083,
+      "step": 2713
+    },
+    {
+      "epoch": 1.3803801894589611,
+      "grad_norm": 2.6110567623822036,
+      "learning_rate": 1.2128267723160662e-05,
+      "loss": 0.4864,
+      "step": 2714
+    },
+    {
+      "epoch": 1.3808888041197789,
+      "grad_norm": 2.3821251049675274,
+      "learning_rate": 1.2122788862041723e-05,
+      "loss": 0.6051,
+      "step": 2715
+    },
+    {
+      "epoch": 1.3813974187805964,
+      "grad_norm": 1.7708446352728915,
+      "learning_rate": 1.2117309333554925e-05,
+      "loss": 0.5039,
+      "step": 2716
+    },
+    {
+      "epoch": 1.381906033441414,
+      "grad_norm": 2.1997205253297256,
+      "learning_rate": 1.211182913942293e-05,
+      "loss": 0.4489,
+      "step": 2717
+    },
+    {
+      "epoch": 1.3824146481022315,
+      "grad_norm": 2.0466940996441823,
+      "learning_rate": 1.2106348281368617e-05,
+      "loss": 0.5164,
+      "step": 2718
+    },
+    {
+      "epoch": 1.3829232627630492,
+      "grad_norm": 2.1679040847619477,
+      "learning_rate": 1.2100866761115073e-05,
+      "loss": 0.5417,
+      "step": 2719
+    },
+    {
+      "epoch": 1.3834318774238668,
+      "grad_norm": 1.945275731486102,
+      "learning_rate": 1.2095384580385594e-05,
+      "loss": 0.5279,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3839404920846843,
+      "grad_norm": 2.2485291653862918,
+      "learning_rate": 1.2089901740903685e-05,
+      "loss": 0.4977,
+      "step": 2721
+    },
+    {
+      "epoch": 1.3844491067455018,
+      "grad_norm": 1.9884868283218025,
+      "learning_rate": 1.2084418244393047e-05,
+      "loss": 0.5075,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3849577214063196,
+      "grad_norm": 1.7113337645778453,
+      "learning_rate": 1.2078934092577601e-05,
+      "loss": 0.485,
+      "step": 2723
+    },
+    {
+      "epoch": 1.385466336067137,
+      "grad_norm": 2.052068745775866,
+      "learning_rate": 1.2073449287181468e-05,
+      "loss": 0.5128,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3859749507279546,
+      "grad_norm": 2.0507461399524285,
+      "learning_rate": 1.2067963829928978e-05,
+      "loss": 0.4194,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3864835653887724,
+      "grad_norm": 2.607534890458046,
+      "learning_rate": 1.2062477722544656e-05,
+      "loss": 0.5328,
+      "step": 2726
+    },
+    {
+      "epoch": 1.38699218004959,
+      "grad_norm": 2.00607842397997,
+      "learning_rate": 1.2056990966753246e-05,
+      "loss": 0.4961,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3875007947104074,
+      "grad_norm": 2.121828405124278,
+      "learning_rate": 1.2051503564279685e-05,
+      "loss": 0.487,
+      "step": 2728
+    },
+    {
+      "epoch": 1.3880094093712252,
+      "grad_norm": 2.2948368917110655,
+      "learning_rate": 1.204601551684912e-05,
+      "loss": 0.5028,
+      "step": 2729
+    },
+    {
+      "epoch": 1.3885180240320427,
+      "grad_norm": 2.2727739866831143,
+      "learning_rate": 1.2040526826186896e-05,
+      "loss": 0.552,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3890266386928602,
+      "grad_norm": 2.0733279741878285,
+      "learning_rate": 1.2035037494018558e-05,
+      "loss": 0.5035,
+      "step": 2731
+    },
+    {
+      "epoch": 1.389535253353678,
+      "grad_norm": 2.1205856248531876,
+      "learning_rate": 1.2029547522069862e-05,
+      "loss": 0.5572,
+      "step": 2732
+    },
+    {
+      "epoch": 1.3900438680144955,
+      "grad_norm": 1.734340633017473,
+      "learning_rate": 1.2024056912066755e-05,
+      "loss": 0.4777,
+      "step": 2733
+    },
+    {
+      "epoch": 1.390552482675313,
+      "grad_norm": 1.8235952222483294,
+      "learning_rate": 1.2018565665735394e-05,
+      "loss": 0.4851,
+      "step": 2734
+    },
+    {
+      "epoch": 1.3910610973361308,
+      "grad_norm": 2.2758101309905747,
+      "learning_rate": 1.2013073784802127e-05,
+      "loss": 0.639,
+      "step": 2735
+    },
+    {
+      "epoch": 1.3915697119969483,
+      "grad_norm": 2.0476621322809323,
+      "learning_rate": 1.2007581270993507e-05,
+      "loss": 0.5351,
+      "step": 2736
+    },
+    {
+      "epoch": 1.3920783266577659,
+      "grad_norm": 2.2451490455547423,
+      "learning_rate": 1.2002088126036284e-05,
+      "loss": 0.5376,
+      "step": 2737
+    },
+    {
+      "epoch": 1.3925869413185836,
+      "grad_norm": 2.0233680978309576,
+      "learning_rate": 1.1996594351657412e-05,
+      "loss": 0.4557,
+      "step": 2738
+    },
+    {
+      "epoch": 1.3930955559794012,
+      "grad_norm": 2.168384409055408,
+      "learning_rate": 1.1991099949584033e-05,
+      "loss": 0.5409,
+      "step": 2739
+    },
+    {
+      "epoch": 1.3936041706402187,
+      "grad_norm": 2.596069401946273,
+      "learning_rate": 1.198560492154349e-05,
+      "loss": 0.5091,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3941127853010364,
+      "grad_norm": 2.120335901928186,
+      "learning_rate": 1.198010926926333e-05,
+      "loss": 0.5237,
+      "step": 2741
+    },
+    {
+      "epoch": 1.394621399961854,
+      "grad_norm": 2.32084921849109,
+      "learning_rate": 1.1974612994471282e-05,
+      "loss": 0.4919,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3951300146226715,
+      "grad_norm": 2.294077928265426,
+      "learning_rate": 1.196911609889529e-05,
+      "loss": 0.5125,
+      "step": 2743
+    },
+    {
+      "epoch": 1.395638629283489,
+      "grad_norm": 1.9767541603979577,
+      "learning_rate": 1.1963618584263475e-05,
+      "loss": 0.4762,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3961472439443066,
+      "grad_norm": 2.053821801981178,
+      "learning_rate": 1.195812045230416e-05,
+      "loss": 0.5701,
+      "step": 2745
+    },
+    {
+      "epoch": 1.3966558586051243,
+      "grad_norm": 1.9302510662532926,
+      "learning_rate": 1.1952621704745863e-05,
+      "loss": 0.5233,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3971644732659418,
+      "grad_norm": 2.098346731533496,
+      "learning_rate": 1.1947122343317298e-05,
+      "loss": 0.5223,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3976730879267594,
+      "grad_norm": 2.103874438045677,
+      "learning_rate": 1.1941622369747365e-05,
+      "loss": 0.5292,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3981817025875771,
+      "grad_norm": 1.9237986915776668,
+      "learning_rate": 1.1936121785765158e-05,
+      "loss": 0.4643,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3986903172483947,
+      "grad_norm": 1.8995729669231027,
+      "learning_rate": 1.1930620593099968e-05,
+      "loss": 0.5225,
+      "step": 2750
+    },
+    {
+      "epoch": 1.3991989319092122,
+      "grad_norm": 1.6704197191702297,
+      "learning_rate": 1.1925118793481274e-05,
+      "loss": 0.461,
+      "step": 2751
+    },
+    {
+      "epoch": 1.39970754657003,
+      "grad_norm": 2.2021683584187657,
+      "learning_rate": 1.191961638863875e-05,
+      "loss": 0.555,
+      "step": 2752
+    },
+    {
+      "epoch": 1.4002161612308475,
+      "grad_norm": 2.307581232527764,
+      "learning_rate": 1.1914113380302246e-05,
+      "loss": 0.5256,
+      "step": 2753
+    },
+    {
+      "epoch": 1.400724775891665,
+      "grad_norm": 1.8022182465548762,
+      "learning_rate": 1.1908609770201817e-05,
+      "loss": 0.4891,
+      "step": 2754
+    },
+    {
+      "epoch": 1.4012333905524827,
+      "grad_norm": 1.8906275221122124,
+      "learning_rate": 1.1903105560067705e-05,
+      "loss": 0.4286,
+      "step": 2755
+    },
+    {
+      "epoch": 1.4017420052133003,
+      "grad_norm": 2.2581008674183054,
+      "learning_rate": 1.1897600751630337e-05,
+      "loss": 0.3859,
+      "step": 2756
+    },
+    {
+      "epoch": 1.4022506198741178,
+      "grad_norm": 1.8557253369050857,
+      "learning_rate": 1.1892095346620328e-05,
+      "loss": 0.4923,
+      "step": 2757
+    },
+    {
+      "epoch": 1.4027592345349356,
+      "grad_norm": 2.208734607826404,
+      "learning_rate": 1.1886589346768477e-05,
+      "loss": 0.4672,
+      "step": 2758
+    },
+    {
+      "epoch": 1.403267849195753,
+      "grad_norm": 2.016884580246024,
+      "learning_rate": 1.1881082753805783e-05,
+      "loss": 0.5719,
+      "step": 2759
+    },
+    {
+      "epoch": 1.4037764638565706,
+      "grad_norm": 2.1104627277234402,
+      "learning_rate": 1.1875575569463415e-05,
+      "loss": 0.5368,
+      "step": 2760
+    },
+    {
+      "epoch": 1.4042850785173884,
+      "grad_norm": 2.243493527176295,
+      "learning_rate": 1.1870067795472744e-05,
+      "loss": 0.5289,
+      "step": 2761
+    },
+    {
+      "epoch": 1.404793693178206,
+      "grad_norm": 1.8765074074667838,
+      "learning_rate": 1.1864559433565311e-05,
+      "loss": 0.4685,
+      "step": 2762
+    },
+    {
+      "epoch": 1.4053023078390234,
+      "grad_norm": 1.9292600212276847,
+      "learning_rate": 1.1859050485472852e-05,
+      "loss": 0.5615,
+      "step": 2763
+    },
+    {
+      "epoch": 1.4058109224998412,
+      "grad_norm": 2.207040540287416,
+      "learning_rate": 1.1853540952927286e-05,
+      "loss": 0.4865,
+      "step": 2764
+    },
+    {
+      "epoch": 1.4063195371606587,
+      "grad_norm": 1.9033167424567274,
+      "learning_rate": 1.1848030837660708e-05,
+      "loss": 0.5074,
+      "step": 2765
+    },
+    {
+      "epoch": 1.4068281518214762,
+      "grad_norm": 2.0386069070171158,
+      "learning_rate": 1.1842520141405413e-05,
+      "loss": 0.5223,
+      "step": 2766
+    },
+    {
+      "epoch": 1.4073367664822938,
+      "grad_norm": 2.0965850766089775,
+      "learning_rate": 1.183700886589386e-05,
+      "loss": 0.5473,
+      "step": 2767
+    },
+    {
+      "epoch": 1.4078453811431115,
+      "grad_norm": 1.9250019965223388,
+      "learning_rate": 1.1831497012858697e-05,
+      "loss": 0.4673,
+      "step": 2768
+    },
+    {
+      "epoch": 1.408353995803929,
+      "grad_norm": 2.1234208017216187,
+      "learning_rate": 1.1825984584032759e-05,
+      "loss": 0.5614,
+      "step": 2769
+    },
+    {
+      "epoch": 1.4088626104647466,
+      "grad_norm": 1.879928025531778,
+      "learning_rate": 1.182047158114906e-05,
+      "loss": 0.5262,
+      "step": 2770
+    },
+    {
+      "epoch": 1.4093712251255641,
+      "grad_norm": 2.0181670017307787,
+      "learning_rate": 1.1814958005940782e-05,
+      "loss": 0.4844,
+      "step": 2771
+    },
+    {
+      "epoch": 1.4098798397863819,
+      "grad_norm": 2.0115612375760055,
+      "learning_rate": 1.1809443860141307e-05,
+      "loss": 0.5603,
+      "step": 2772
+    },
+    {
+      "epoch": 1.4103884544471994,
+      "grad_norm": 2.3574836678611746,
+      "learning_rate": 1.1803929145484182e-05,
+      "loss": 0.4882,
+      "step": 2773
+    },
+    {
+      "epoch": 1.410897069108017,
+      "grad_norm": 2.02319369817275,
+      "learning_rate": 1.1798413863703136e-05,
+      "loss": 0.4587,
+      "step": 2774
+    },
+    {
+      "epoch": 1.4114056837688347,
+      "grad_norm": 1.822567243238196,
+      "learning_rate": 1.1792898016532082e-05,
+      "loss": 0.4786,
+      "step": 2775
+    },
+    {
+      "epoch": 1.4119142984296522,
+      "grad_norm": 2.126221586168758,
+      "learning_rate": 1.1787381605705103e-05,
+      "loss": 0.4933,
+      "step": 2776
+    },
+    {
+      "epoch": 1.4124229130904697,
+      "grad_norm": 2.157256402931344,
+      "learning_rate": 1.178186463295646e-05,
+      "loss": 0.504,
+      "step": 2777
+    },
+    {
+      "epoch": 1.4129315277512875,
+      "grad_norm": 2.1492370227520885,
+      "learning_rate": 1.1776347100020603e-05,
+      "loss": 0.4745,
+      "step": 2778
+    },
+    {
+      "epoch": 1.413440142412105,
+      "grad_norm": 2.389474928729757,
+      "learning_rate": 1.1770829008632138e-05,
+      "loss": 0.5455,
+      "step": 2779
+    },
+    {
+      "epoch": 1.4139487570729226,
+      "grad_norm": 2.5566707460685056,
+      "learning_rate": 1.1765310360525864e-05,
+      "loss": 0.4418,
+      "step": 2780
+    },
+    {
+      "epoch": 1.4144573717337403,
+      "grad_norm": 1.958703094096074,
+      "learning_rate": 1.1759791157436743e-05,
+      "loss": 0.4544,
+      "step": 2781
+    },
+    {
+      "epoch": 1.4149659863945578,
+      "grad_norm": 2.347234653186673,
+      "learning_rate": 1.1754271401099915e-05,
+      "loss": 0.539,
+      "step": 2782
+    },
+    {
+      "epoch": 1.4154746010553754,
+      "grad_norm": 1.7774776198725961,
+      "learning_rate": 1.17487510932507e-05,
+      "loss": 0.4859,
+      "step": 2783
+    },
+    {
+      "epoch": 1.4159832157161931,
+      "grad_norm": 1.9198861104247702,
+      "learning_rate": 1.174323023562459e-05,
+      "loss": 0.4923,
+      "step": 2784
+    },
+    {
+      "epoch": 1.4164918303770107,
+      "grad_norm": 3.425013944500012,
+      "learning_rate": 1.1737708829957235e-05,
+      "loss": 0.4691,
+      "step": 2785
+    },
+    {
+      "epoch": 1.4170004450378282,
+      "grad_norm": 2.1674598440214656,
+      "learning_rate": 1.1732186877984479e-05,
+      "loss": 0.5042,
+      "step": 2786
+    },
+    {
+      "epoch": 1.417509059698646,
+      "grad_norm": 1.842943758793073,
+      "learning_rate": 1.1726664381442323e-05,
+      "loss": 0.5422,
+      "step": 2787
+    },
+    {
+      "epoch": 1.4180176743594635,
+      "grad_norm": 2.1289398470686116,
+      "learning_rate": 1.1721141342066943e-05,
+      "loss": 0.49,
+      "step": 2788
+    },
+    {
+      "epoch": 1.418526289020281,
+      "grad_norm": 2.1867612408717276,
+      "learning_rate": 1.1715617761594694e-05,
+      "loss": 0.4362,
+      "step": 2789
+    },
+    {
+      "epoch": 1.4190349036810985,
+      "grad_norm": 1.8901507831149758,
+      "learning_rate": 1.171009364176208e-05,
+      "loss": 0.4969,
+      "step": 2790
+    },
+    {
+      "epoch": 1.4195435183419163,
+      "grad_norm": 2.0899969597388246,
+      "learning_rate": 1.1704568984305802e-05,
+      "loss": 0.4326,
+      "step": 2791
+    },
+    {
+      "epoch": 1.4200521330027338,
+      "grad_norm": 1.955987160669793,
+      "learning_rate": 1.1699043790962707e-05,
+      "loss": 0.5114,
+      "step": 2792
+    },
+    {
+      "epoch": 1.4205607476635513,
+      "grad_norm": 2.197681058817544,
+      "learning_rate": 1.1693518063469824e-05,
+      "loss": 0.4861,
+      "step": 2793
+    },
+    {
+      "epoch": 1.4210693623243689,
+      "grad_norm": 1.9357668600549025,
+      "learning_rate": 1.168799180356434e-05,
+      "loss": 0.4664,
+      "step": 2794
+    },
+    {
+      "epoch": 1.4215779769851866,
+      "grad_norm": 1.9651771702380436,
+      "learning_rate": 1.1682465012983622e-05,
+      "loss": 0.4539,
+      "step": 2795
+    },
+    {
+      "epoch": 1.4220865916460042,
+      "grad_norm": 2.1934193766837473,
+      "learning_rate": 1.1676937693465194e-05,
+      "loss": 0.4552,
+      "step": 2796
+    },
+    {
+      "epoch": 1.4225952063068217,
+      "grad_norm": 2.060762424855907,
+      "learning_rate": 1.1671409846746751e-05,
+      "loss": 0.4945,
+      "step": 2797
+    },
+    {
+      "epoch": 1.4231038209676394,
+      "grad_norm": 2.3043868563838217,
+      "learning_rate": 1.1665881474566146e-05,
+      "loss": 0.6029,
+      "step": 2798
+    },
+    {
+      "epoch": 1.423612435628457,
+      "grad_norm": 2.902999372607724,
+      "learning_rate": 1.1660352578661406e-05,
+      "loss": 0.4462,
+      "step": 2799
+    },
+    {
+      "epoch": 1.4241210502892745,
+      "grad_norm": 1.9281367816286876,
+      "learning_rate": 1.1654823160770723e-05,
+      "loss": 0.4588,
+      "step": 2800
+    },
+    {
+      "epoch": 1.4241210502892745,
+      "eval_loss": 0.7160014510154724,
+      "eval_runtime": 169.8193,
+      "eval_samples_per_second": 19.503,
+      "eval_steps_per_second": 9.752,
+      "step": 2800
+    },
+    {
+      "epoch": 1.4246296649500922,
+      "grad_norm": 2.2122940915947584,
+      "learning_rate": 1.1649293222632445e-05,
+      "loss": 0.5544,
+      "step": 2801
+    },
+    {
+      "epoch": 1.4251382796109098,
+      "grad_norm": 1.9719472687501562,
+      "learning_rate": 1.1643762765985092e-05,
+      "loss": 0.535,
+      "step": 2802
+    },
+    {
+      "epoch": 1.4256468942717273,
+      "grad_norm": 2.044644254040256,
+      "learning_rate": 1.1638231792567338e-05,
+      "loss": 0.5864,
+      "step": 2803
+    },
+    {
+      "epoch": 1.426155508932545,
+      "grad_norm": 2.0814153107296356,
+      "learning_rate": 1.1632700304118031e-05,
+      "loss": 0.553,
+      "step": 2804
+    },
+    {
+      "epoch": 1.4266641235933626,
+      "grad_norm": 1.9469946516812873,
+      "learning_rate": 1.1627168302376171e-05,
+      "loss": 0.4931,
+      "step": 2805
+    },
+    {
+      "epoch": 1.4271727382541801,
+      "grad_norm": 2.7708943486964968,
+      "learning_rate": 1.1621635789080923e-05,
+      "loss": 0.443,
+      "step": 2806
+    },
+    {
+      "epoch": 1.4276813529149979,
+      "grad_norm": 1.8015849208842731,
+      "learning_rate": 1.1616102765971617e-05,
+      "loss": 0.5143,
+      "step": 2807
+    },
+    {
+      "epoch": 1.4281899675758154,
+      "grad_norm": 2.085110046629476,
+      "learning_rate": 1.1610569234787733e-05,
+      "loss": 0.4979,
+      "step": 2808
+    },
+    {
+      "epoch": 1.428698582236633,
+      "grad_norm": 2.419937146056285,
+      "learning_rate": 1.160503519726892e-05,
+      "loss": 0.445,
+      "step": 2809
+    },
+    {
+      "epoch": 1.4292071968974507,
+      "grad_norm": 2.2106338130655536,
+      "learning_rate": 1.1599500655154983e-05,
+      "loss": 0.5876,
+      "step": 2810
+    },
+    {
+      "epoch": 1.4297158115582682,
+      "grad_norm": 2.074374378062882,
+      "learning_rate": 1.1593965610185886e-05,
+      "loss": 0.514,
+      "step": 2811
+    },
+    {
+      "epoch": 1.4302244262190857,
+      "grad_norm": 1.9474638414270422,
+      "learning_rate": 1.158843006410175e-05,
+      "loss": 0.4415,
+      "step": 2812
+    },
+    {
+      "epoch": 1.4307330408799035,
+      "grad_norm": 2.2656439418658607,
+      "learning_rate": 1.1582894018642854e-05,
+      "loss": 0.6027,
+      "step": 2813
+    },
+    {
+      "epoch": 1.431241655540721,
+      "grad_norm": 2.2843052510312445,
+      "learning_rate": 1.1577357475549636e-05,
+      "loss": 0.5074,
+      "step": 2814
+    },
+    {
+      "epoch": 1.4317502702015386,
+      "grad_norm": 2.0543846226238505,
+      "learning_rate": 1.1571820436562686e-05,
+      "loss": 0.4862,
+      "step": 2815
+    },
+    {
+      "epoch": 1.432258884862356,
+      "grad_norm": 2.107876846089166,
+      "learning_rate": 1.1566282903422757e-05,
+      "loss": 0.5252,
+      "step": 2816
+    },
+    {
+      "epoch": 1.4327674995231738,
+      "grad_norm": 2.0366764065665164,
+      "learning_rate": 1.1560744877870748e-05,
+      "loss": 0.4755,
+      "step": 2817
+    },
+    {
+      "epoch": 1.4332761141839914,
+      "grad_norm": 2.03956521805176,
+      "learning_rate": 1.1555206361647722e-05,
+      "loss": 0.4989,
+      "step": 2818
+    },
+    {
+      "epoch": 1.433784728844809,
+      "grad_norm": 2.236285160101739,
+      "learning_rate": 1.1549667356494889e-05,
+      "loss": 0.5068,
+      "step": 2819
+    },
+    {
+      "epoch": 1.4342933435056264,
+      "grad_norm": 2.3204632334444204,
+      "learning_rate": 1.1544127864153618e-05,
+      "loss": 0.5384,
+      "step": 2820
+    },
+    {
+      "epoch": 1.4348019581664442,
+      "grad_norm": 2.336176831287015,
+      "learning_rate": 1.1538587886365426e-05,
+      "loss": 0.5823,
+      "step": 2821
+    },
+    {
+      "epoch": 1.4353105728272617,
+      "grad_norm": 2.0564945527720355,
+      "learning_rate": 1.153304742487199e-05,
+      "loss": 0.546,
+      "step": 2822
+    },
+    {
+      "epoch": 1.4358191874880792,
+      "grad_norm": 2.1574474824279766,
+      "learning_rate": 1.1527506481415132e-05,
+      "loss": 0.4953,
+      "step": 2823
+    },
+    {
+      "epoch": 1.436327802148897,
+      "grad_norm": 2.175164236247663,
+      "learning_rate": 1.1521965057736825e-05,
+      "loss": 0.53,
+      "step": 2824
+    },
+    {
+      "epoch": 1.4368364168097145,
+      "grad_norm": 1.9922897013460055,
+      "learning_rate": 1.1516423155579207e-05,
+      "loss": 0.5079,
+      "step": 2825
+    },
+    {
+      "epoch": 1.437345031470532,
+      "grad_norm": 2.3121627355477203,
+      "learning_rate": 1.151088077668454e-05,
+      "loss": 0.4937,
+      "step": 2826
+    },
+    {
+      "epoch": 1.4378536461313498,
+      "grad_norm": 2.2409388025822894,
+      "learning_rate": 1.1505337922795263e-05,
+      "loss": 0.633,
+      "step": 2827
+    },
+    {
+      "epoch": 1.4383622607921673,
+      "grad_norm": 2.249555967809454,
+      "learning_rate": 1.149979459565395e-05,
+      "loss": 0.5554,
+      "step": 2828
+    },
+    {
+      "epoch": 1.4388708754529849,
+      "grad_norm": 1.989137476416713,
+      "learning_rate": 1.1494250797003323e-05,
+      "loss": 0.4981,
+      "step": 2829
+    },
+    {
+      "epoch": 1.4393794901138026,
+      "grad_norm": 1.9491124554168453,
+      "learning_rate": 1.1488706528586263e-05,
+      "loss": 0.4815,
+      "step": 2830
+    },
+    {
+      "epoch": 1.4398881047746201,
+      "grad_norm": 2.3323951833383134,
+      "learning_rate": 1.1483161792145784e-05,
+      "loss": 0.5422,
+      "step": 2831
+    },
+    {
+      "epoch": 1.4403967194354377,
+      "grad_norm": 2.3446155221520453,
+      "learning_rate": 1.1477616589425056e-05,
+      "loss": 0.5564,
+      "step": 2832
+    },
+    {
+      "epoch": 1.4409053340962554,
+      "grad_norm": 2.1144827707616587,
+      "learning_rate": 1.14720709221674e-05,
+      "loss": 0.4665,
+      "step": 2833
+    },
+    {
+      "epoch": 1.441413948757073,
+      "grad_norm": 2.107357840851186,
+      "learning_rate": 1.1466524792116271e-05,
+      "loss": 0.4918,
+      "step": 2834
+    },
+    {
+      "epoch": 1.4419225634178905,
+      "grad_norm": 2.0902491666483964,
+      "learning_rate": 1.1460978201015278e-05,
+      "loss": 0.5873,
+      "step": 2835
+    },
+    {
+      "epoch": 1.4424311780787082,
+      "grad_norm": 2.3172903191152345,
+      "learning_rate": 1.1455431150608174e-05,
+      "loss": 0.5305,
+      "step": 2836
+    },
+    {
+      "epoch": 1.4429397927395258,
+      "grad_norm": 2.321873714400458,
+      "learning_rate": 1.1449883642638853e-05,
+      "loss": 0.5513,
+      "step": 2837
+    },
+    {
+      "epoch": 1.4434484074003433,
+      "grad_norm": 1.979744381828829,
+      "learning_rate": 1.1444335678851358e-05,
+      "loss": 0.4658,
+      "step": 2838
+    },
+    {
+      "epoch": 1.4439570220611608,
+      "grad_norm": 2.277893314170931,
+      "learning_rate": 1.1438787260989866e-05,
+      "loss": 0.4601,
+      "step": 2839
+    },
+    {
+      "epoch": 1.4444656367219786,
+      "grad_norm": 2.336674398589935,
+      "learning_rate": 1.1433238390798706e-05,
+      "loss": 0.5666,
+      "step": 2840
+    },
+    {
+      "epoch": 1.4449742513827961,
+      "grad_norm": 1.842283150322755,
+      "learning_rate": 1.142768907002235e-05,
+      "loss": 0.4546,
+      "step": 2841
+    },
+    {
+      "epoch": 1.4454828660436136,
+      "grad_norm": 2.49138300715287,
+      "learning_rate": 1.1422139300405402e-05,
+      "loss": 0.5384,
+      "step": 2842
+    },
+    {
+      "epoch": 1.4459914807044312,
+      "grad_norm": 2.597373453679628,
+      "learning_rate": 1.141658908369262e-05,
+      "loss": 0.5071,
+      "step": 2843
+    },
+    {
+      "epoch": 1.446500095365249,
+      "grad_norm": 2.1340831283352153,
+      "learning_rate": 1.1411038421628888e-05,
+      "loss": 0.4919,
+      "step": 2844
+    },
+    {
+      "epoch": 1.4470087100260665,
+      "grad_norm": 2.0779353463961505,
+      "learning_rate": 1.1405487315959241e-05,
+      "loss": 0.531,
+      "step": 2845
+    },
+    {
+      "epoch": 1.447517324686884,
+      "grad_norm": 2.083709784510418,
+      "learning_rate": 1.1399935768428847e-05,
+      "loss": 0.4784,
+      "step": 2846
+    },
+    {
+      "epoch": 1.4480259393477017,
+      "grad_norm": 2.08936590988642,
+      "learning_rate": 1.1394383780783025e-05,
+      "loss": 0.4408,
+      "step": 2847
+    },
+    {
+      "epoch": 1.4485345540085193,
+      "grad_norm": 2.5391113276555077,
+      "learning_rate": 1.1388831354767214e-05,
+      "loss": 0.6476,
+      "step": 2848
+    },
+    {
+      "epoch": 1.4490431686693368,
+      "grad_norm": 2.255380766328171,
+      "learning_rate": 1.1383278492127001e-05,
+      "loss": 0.5834,
+      "step": 2849
+    },
+    {
+      "epoch": 1.4495517833301546,
+      "grad_norm": 2.2962113024877757,
+      "learning_rate": 1.1377725194608113e-05,
+      "loss": 0.5991,
+      "step": 2850
+    },
+    {
+      "epoch": 1.450060397990972,
+      "grad_norm": 2.255157493623211,
+      "learning_rate": 1.137217146395641e-05,
+      "loss": 0.5363,
+      "step": 2851
+    },
+    {
+      "epoch": 1.4505690126517896,
+      "grad_norm": 2.214677108066357,
+      "learning_rate": 1.1366617301917886e-05,
+      "loss": 0.5643,
+      "step": 2852
+    },
+    {
+      "epoch": 1.4510776273126074,
+      "grad_norm": 2.009380159030301,
+      "learning_rate": 1.1361062710238673e-05,
+      "loss": 0.4625,
+      "step": 2853
+    },
+    {
+      "epoch": 1.451586241973425,
+      "grad_norm": 2.038727129837763,
+      "learning_rate": 1.1355507690665038e-05,
+      "loss": 0.5487,
+      "step": 2854
+    },
+    {
+      "epoch": 1.4520948566342424,
+      "grad_norm": 2.010508349078986,
+      "learning_rate": 1.1349952244943384e-05,
+      "loss": 0.5276,
+      "step": 2855
+    },
+    {
+      "epoch": 1.4526034712950602,
+      "grad_norm": 2.7148862547257173,
+      "learning_rate": 1.1344396374820243e-05,
+      "loss": 0.5163,
+      "step": 2856
+    },
+    {
+      "epoch": 1.4531120859558777,
+      "grad_norm": 2.7287852673560367,
+      "learning_rate": 1.133884008204229e-05,
+      "loss": 0.541,
+      "step": 2857
+    },
+    {
+      "epoch": 1.4536207006166952,
+      "grad_norm": 2.0211889078678027,
+      "learning_rate": 1.133328336835632e-05,
+      "loss": 0.5601,
+      "step": 2858
+    },
+    {
+      "epoch": 1.454129315277513,
+      "grad_norm": 1.8917025813363848,
+      "learning_rate": 1.132772623550927e-05,
+      "loss": 0.5757,
+      "step": 2859
+    },
+    {
+      "epoch": 1.4546379299383305,
+      "grad_norm": 2.216502165845161,
+      "learning_rate": 1.1322168685248206e-05,
+      "loss": 0.5129,
+      "step": 2860
+    },
+    {
+      "epoch": 1.455146544599148,
+      "grad_norm": 2.1658572249974113,
+      "learning_rate": 1.1316610719320328e-05,
+      "loss": 0.51,
+      "step": 2861
+    },
+    {
+      "epoch": 1.4556551592599658,
+      "grad_norm": 2.490949173606708,
+      "learning_rate": 1.131105233947296e-05,
+      "loss": 0.5357,
+      "step": 2862
+    },
+    {
+      "epoch": 1.4561637739207833,
+      "grad_norm": 1.9213398133966544,
+      "learning_rate": 1.1305493547453556e-05,
+      "loss": 0.5223,
+      "step": 2863
+    },
+    {
+      "epoch": 1.4566723885816009,
+      "grad_norm": 1.7218909953459642,
+      "learning_rate": 1.1299934345009706e-05,
+      "loss": 0.4196,
+      "step": 2864
+    },
+    {
+      "epoch": 1.4571810032424184,
+      "grad_norm": 2.7142686413106105,
+      "learning_rate": 1.1294374733889136e-05,
+      "loss": 0.4919,
+      "step": 2865
+    },
+    {
+      "epoch": 1.457689617903236,
+      "grad_norm": 1.9659580840011102,
+      "learning_rate": 1.128881471583968e-05,
+      "loss": 0.4951,
+      "step": 2866
+    },
+    {
+      "epoch": 1.4581982325640537,
+      "grad_norm": 2.1428588571995735,
+      "learning_rate": 1.1283254292609312e-05,
+      "loss": 0.5661,
+      "step": 2867
+    },
+    {
+      "epoch": 1.4587068472248712,
+      "grad_norm": 2.0712431345464988,
+      "learning_rate": 1.1277693465946136e-05,
+      "loss": 0.4893,
+      "step": 2868
+    },
+    {
+      "epoch": 1.4592154618856887,
+      "grad_norm": 2.1998393130140834,
+      "learning_rate": 1.1272132237598377e-05,
+      "loss": 0.5417,
+      "step": 2869
+    },
+    {
+      "epoch": 1.4597240765465065,
+      "grad_norm": 2.186031284460917,
+      "learning_rate": 1.1266570609314393e-05,
+      "loss": 0.5915,
+      "step": 2870
+    },
+    {
+      "epoch": 1.460232691207324,
+      "grad_norm": 2.0007968745576714,
+      "learning_rate": 1.1261008582842654e-05,
+      "loss": 0.5416,
+      "step": 2871
+    },
+    {
+      "epoch": 1.4607413058681415,
+      "grad_norm": 2.393608375921486,
+      "learning_rate": 1.125544615993177e-05,
+      "loss": 0.5449,
+      "step": 2872
+    },
+    {
+      "epoch": 1.4612499205289593,
+      "grad_norm": 2.2277765544635377,
+      "learning_rate": 1.124988334233047e-05,
+      "loss": 0.5569,
+      "step": 2873
+    },
+    {
+      "epoch": 1.4617585351897768,
+      "grad_norm": 1.919634187241553,
+      "learning_rate": 1.1244320131787608e-05,
+      "loss": 0.4929,
+      "step": 2874
+    },
+    {
+      "epoch": 1.4622671498505944,
+      "grad_norm": 1.680446435557623,
+      "learning_rate": 1.1238756530052157e-05,
+      "loss": 0.3858,
+      "step": 2875
+    },
+    {
+      "epoch": 1.4627757645114121,
+      "grad_norm": 1.9716763016934773,
+      "learning_rate": 1.1233192538873219e-05,
+      "loss": 0.5126,
+      "step": 2876
+    },
+    {
+      "epoch": 1.4632843791722296,
+      "grad_norm": 2.2570839732803396,
+      "learning_rate": 1.1227628160000013e-05,
+      "loss": 0.4731,
+      "step": 2877
+    },
+    {
+      "epoch": 1.4637929938330472,
+      "grad_norm": 2.204834403396075,
+      "learning_rate": 1.1222063395181888e-05,
+      "loss": 0.5568,
+      "step": 2878
+    },
+    {
+      "epoch": 1.464301608493865,
+      "grad_norm": 2.377159441956237,
+      "learning_rate": 1.1216498246168303e-05,
+      "loss": 0.5253,
+      "step": 2879
+    },
+    {
+      "epoch": 1.4648102231546825,
+      "grad_norm": 2.053326870590212,
+      "learning_rate": 1.1210932714708851e-05,
+      "loss": 0.4809,
+      "step": 2880
+    },
+    {
+      "epoch": 1.4653188378155,
+      "grad_norm": 1.939720173839119,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.474,
+      "step": 2881
+    },
+    {
+      "epoch": 1.4658274524763177,
+      "grad_norm": 1.6274956832336274,
+      "learning_rate": 1.1199800511451273e-05,
+      "loss": 0.4177,
+      "step": 2882
+    },
+    {
+      "epoch": 1.4663360671371353,
+      "grad_norm": 2.167844476660582,
+      "learning_rate": 1.1194233843152924e-05,
+      "loss": 0.5631,
+      "step": 2883
+    },
+    {
+      "epoch": 1.4668446817979528,
+      "grad_norm": 1.854314633869629,
+      "learning_rate": 1.1188666799408242e-05,
+      "loss": 0.5313,
+      "step": 2884
+    },
+    {
+      "epoch": 1.4673532964587706,
+      "grad_norm": 2.1650947835494248,
+      "learning_rate": 1.1183099381967416e-05,
+      "loss": 0.5916,
+      "step": 2885
+    },
+    {
+      "epoch": 1.467861911119588,
+      "grad_norm": 1.709297683334142,
+      "learning_rate": 1.1177531592580737e-05,
+      "loss": 0.4957,
+      "step": 2886
+    },
+    {
+      "epoch": 1.4683705257804056,
+      "grad_norm": 1.7711323953054552,
+      "learning_rate": 1.1171963432998624e-05,
+      "loss": 0.3951,
+      "step": 2887
+    },
+    {
+      "epoch": 1.4688791404412231,
+      "grad_norm": 1.9481539522234173,
+      "learning_rate": 1.1166394904971614e-05,
+      "loss": 0.5591,
+      "step": 2888
+    },
+    {
+      "epoch": 1.469387755102041,
+      "grad_norm": 1.9437113227553302,
+      "learning_rate": 1.116082601025035e-05,
+      "loss": 0.5218,
+      "step": 2889
+    },
+    {
+      "epoch": 1.4698963697628584,
+      "grad_norm": 1.7400906521651927,
+      "learning_rate": 1.1155256750585594e-05,
+      "loss": 0.412,
+      "step": 2890
+    },
+    {
+      "epoch": 1.470404984423676,
+      "grad_norm": 2.1992943108758385,
+      "learning_rate": 1.114968712772823e-05,
+      "loss": 0.563,
+      "step": 2891
+    },
+    {
+      "epoch": 1.4709135990844935,
+      "grad_norm": 2.190344242056978,
+      "learning_rate": 1.1144117143429246e-05,
+      "loss": 0.5867,
+      "step": 2892
+    },
+    {
+      "epoch": 1.4714222137453112,
+      "grad_norm": 1.7836176880254786,
+      "learning_rate": 1.1138546799439753e-05,
+      "loss": 0.4401,
+      "step": 2893
+    },
+    {
+      "epoch": 1.4719308284061288,
+      "grad_norm": 2.378448793121059,
+      "learning_rate": 1.1132976097510966e-05,
+      "loss": 0.5189,
+      "step": 2894
+    },
+    {
+      "epoch": 1.4724394430669463,
+      "grad_norm": 2.2913531927084816,
+      "learning_rate": 1.1127405039394216e-05,
+      "loss": 0.5444,
+      "step": 2895
+    },
+    {
+      "epoch": 1.472948057727764,
+      "grad_norm": 2.466498785062313,
+      "learning_rate": 1.112183362684095e-05,
+      "loss": 0.6324,
+      "step": 2896
+    },
+    {
+      "epoch": 1.4734566723885816,
+      "grad_norm": 3.201422113083899,
+      "learning_rate": 1.111626186160272e-05,
+      "loss": 0.4752,
+      "step": 2897
+    },
+    {
+      "epoch": 1.473965287049399,
+      "grad_norm": 2.1548593431845875,
+      "learning_rate": 1.1110689745431196e-05,
+      "loss": 0.5335,
+      "step": 2898
+    },
+    {
+      "epoch": 1.4744739017102169,
+      "grad_norm": 2.1075206376374176,
+      "learning_rate": 1.1105117280078149e-05,
+      "loss": 0.5924,
+      "step": 2899
+    },
+    {
+      "epoch": 1.4749825163710344,
+      "grad_norm": 2.267031318653507,
+      "learning_rate": 1.109954446729547e-05,
+      "loss": 0.5128,
+      "step": 2900
+    },
+    {
+      "epoch": 1.4749825163710344,
+      "eval_loss": 0.7069129943847656,
+      "eval_runtime": 170.7678,
+      "eval_samples_per_second": 19.395,
+      "eval_steps_per_second": 9.697,
+      "step": 2900
+    },
+    {
+      "epoch": 1.475491131031852,
+      "grad_norm": 1.8090977141695392,
+      "learning_rate": 1.109397130883515e-05,
+      "loss": 0.5213,
+      "step": 2901
+    },
+    {
+      "epoch": 1.4759997456926697,
+      "grad_norm": 2.312378957518916,
+      "learning_rate": 1.10883978064493e-05,
+      "loss": 0.535,
+      "step": 2902
+    },
+    {
+      "epoch": 1.4765083603534872,
+      "grad_norm": 2.457098203392718,
+      "learning_rate": 1.1082823961890123e-05,
+      "loss": 0.592,
+      "step": 2903
+    },
+    {
+      "epoch": 1.4770169750143047,
+      "grad_norm": 2.0008358400904553,
+      "learning_rate": 1.1077249776909946e-05,
+      "loss": 0.4874,
+      "step": 2904
+    },
+    {
+      "epoch": 1.4775255896751225,
+      "grad_norm": 1.9073610663657243,
+      "learning_rate": 1.1071675253261192e-05,
+      "loss": 0.5336,
+      "step": 2905
+    },
+    {
+      "epoch": 1.47803420433594,
+      "grad_norm": 2.127878371412565,
+      "learning_rate": 1.1066100392696394e-05,
+      "loss": 0.5659,
+      "step": 2906
+    },
+    {
+      "epoch": 1.4785428189967575,
+      "grad_norm": 2.6188105318998827,
+      "learning_rate": 1.1060525196968193e-05,
+      "loss": 0.6201,
+      "step": 2907
+    },
+    {
+      "epoch": 1.4790514336575753,
+      "grad_norm": 2.3615302542857117,
+      "learning_rate": 1.1054949667829331e-05,
+      "loss": 0.5628,
+      "step": 2908
+    },
+    {
+      "epoch": 1.4795600483183928,
+      "grad_norm": 2.124568222160737,
+      "learning_rate": 1.1049373807032662e-05,
+      "loss": 0.5151,
+      "step": 2909
+    },
+    {
+      "epoch": 1.4800686629792104,
+      "grad_norm": 2.0255153989109997,
+      "learning_rate": 1.1043797616331133e-05,
+      "loss": 0.5687,
+      "step": 2910
+    },
+    {
+      "epoch": 1.480577277640028,
+      "grad_norm": 2.0064285198895138,
+      "learning_rate": 1.1038221097477808e-05,
+      "loss": 0.5539,
+      "step": 2911
+    },
+    {
+      "epoch": 1.4810858923008456,
+      "grad_norm": 1.9762398075658876,
+      "learning_rate": 1.1032644252225844e-05,
+      "loss": 0.557,
+      "step": 2912
+    },
+    {
+      "epoch": 1.4815945069616632,
+      "grad_norm": 2.0403049235496127,
+      "learning_rate": 1.1027067082328502e-05,
+      "loss": 0.5036,
+      "step": 2913
+    },
+    {
+      "epoch": 1.4821031216224807,
+      "grad_norm": 1.8350792359959345,
+      "learning_rate": 1.1021489589539149e-05,
+      "loss": 0.4483,
+      "step": 2914
+    },
+    {
+      "epoch": 1.4826117362832982,
+      "grad_norm": 2.1842996346893444,
+      "learning_rate": 1.1015911775611252e-05,
+      "loss": 0.4911,
+      "step": 2915
+    },
+    {
+      "epoch": 1.483120350944116,
+      "grad_norm": 2.041542153227143,
+      "learning_rate": 1.1010333642298381e-05,
+      "loss": 0.4797,
+      "step": 2916
+    },
+    {
+      "epoch": 1.4836289656049335,
+      "grad_norm": 1.8245431746321603,
+      "learning_rate": 1.10047551913542e-05,
+      "loss": 0.476,
+      "step": 2917
+    },
+    {
+      "epoch": 1.484137580265751,
+      "grad_norm": 2.0398680750880502,
+      "learning_rate": 1.0999176424532478e-05,
+      "loss": 0.5583,
+      "step": 2918
+    },
+    {
+      "epoch": 1.4846461949265688,
+      "grad_norm": 1.981690941970189,
+      "learning_rate": 1.0993597343587083e-05,
+      "loss": 0.4764,
+      "step": 2919
+    },
+    {
+      "epoch": 1.4851548095873863,
+      "grad_norm": 1.8937920140808588,
+      "learning_rate": 1.0988017950271983e-05,
+      "loss": 0.4813,
+      "step": 2920
+    },
+    {
+      "epoch": 1.4856634242482039,
+      "grad_norm": 1.9991504533795856,
+      "learning_rate": 1.0982438246341238e-05,
+      "loss": 0.5043,
+      "step": 2921
+    },
+    {
+      "epoch": 1.4861720389090216,
+      "grad_norm": 2.0049524977995405,
+      "learning_rate": 1.0976858233549013e-05,
+      "loss": 0.5641,
+      "step": 2922
+    },
+    {
+      "epoch": 1.4866806535698391,
+      "grad_norm": 2.0069623401493693,
+      "learning_rate": 1.0971277913649563e-05,
+      "loss": 0.4988,
+      "step": 2923
+    },
+    {
+      "epoch": 1.4871892682306567,
+      "grad_norm": 1.9600043641255898,
+      "learning_rate": 1.0965697288397245e-05,
+      "loss": 0.5405,
+      "step": 2924
+    },
+    {
+      "epoch": 1.4876978828914744,
+      "grad_norm": 2.1315321660005098,
+      "learning_rate": 1.0960116359546515e-05,
+      "loss": 0.5872,
+      "step": 2925
+    },
+    {
+      "epoch": 1.488206497552292,
+      "grad_norm": 2.1127683119264096,
+      "learning_rate": 1.0954535128851913e-05,
+      "loss": 0.4285,
+      "step": 2926
+    },
+    {
+      "epoch": 1.4887151122131095,
+      "grad_norm": 1.9895398066953918,
+      "learning_rate": 1.0948953598068086e-05,
+      "loss": 0.483,
+      "step": 2927
+    },
+    {
+      "epoch": 1.4892237268739272,
+      "grad_norm": 2.2690880978034462,
+      "learning_rate": 1.0943371768949767e-05,
+      "loss": 0.508,
+      "step": 2928
+    },
+    {
+      "epoch": 1.4897323415347448,
+      "grad_norm": 2.9697203004188335,
+      "learning_rate": 1.0937789643251786e-05,
+      "loss": 0.4913,
+      "step": 2929
+    },
+    {
+      "epoch": 1.4902409561955623,
+      "grad_norm": 1.8283275263835224,
+      "learning_rate": 1.0932207222729068e-05,
+      "loss": 0.4212,
+      "step": 2930
+    },
+    {
+      "epoch": 1.49074957085638,
+      "grad_norm": 2.0899223526763056,
+      "learning_rate": 1.0926624509136628e-05,
+      "loss": 0.536,
+      "step": 2931
+    },
+    {
+      "epoch": 1.4912581855171976,
+      "grad_norm": 2.133506131127834,
+      "learning_rate": 1.0921041504229572e-05,
+      "loss": 0.5844,
+      "step": 2932
+    },
+    {
+      "epoch": 1.491766800178015,
+      "grad_norm": 2.424289508835557,
+      "learning_rate": 1.0915458209763104e-05,
+      "loss": 0.5731,
+      "step": 2933
+    },
+    {
+      "epoch": 1.4922754148388329,
+      "grad_norm": 1.8479259205457093,
+      "learning_rate": 1.090987462749251e-05,
+      "loss": 0.505,
+      "step": 2934
+    },
+    {
+      "epoch": 1.4927840294996504,
+      "grad_norm": 2.034630178453502,
+      "learning_rate": 1.0904290759173173e-05,
+      "loss": 0.4531,
+      "step": 2935
+    },
+    {
+      "epoch": 1.493292644160468,
+      "grad_norm": 2.1430552254813433,
+      "learning_rate": 1.0898706606560562e-05,
+      "loss": 0.5192,
+      "step": 2936
+    },
+    {
+      "epoch": 1.4938012588212854,
+      "grad_norm": 2.4611567323913492,
+      "learning_rate": 1.0893122171410242e-05,
+      "loss": 0.5449,
+      "step": 2937
+    },
+    {
+      "epoch": 1.4943098734821032,
+      "grad_norm": 2.0674597320878902,
+      "learning_rate": 1.0887537455477864e-05,
+      "loss": 0.4908,
+      "step": 2938
+    },
+    {
+      "epoch": 1.4948184881429207,
+      "grad_norm": 1.9366819438999276,
+      "learning_rate": 1.0881952460519154e-05,
+      "loss": 0.4627,
+      "step": 2939
+    },
+    {
+      "epoch": 1.4953271028037383,
+      "grad_norm": 2.0556312395316025,
+      "learning_rate": 1.0876367188289947e-05,
+      "loss": 0.4729,
+      "step": 2940
+    },
+    {
+      "epoch": 1.4958357174645558,
+      "grad_norm": 2.0775831849925925,
+      "learning_rate": 1.0870781640546151e-05,
+      "loss": 0.5139,
+      "step": 2941
+    },
+    {
+      "epoch": 1.4963443321253735,
+      "grad_norm": 2.1405924773704137,
+      "learning_rate": 1.0865195819043769e-05,
+      "loss": 0.5722,
+      "step": 2942
+    },
+    {
+      "epoch": 1.496852946786191,
+      "grad_norm": 1.8977240332428886,
+      "learning_rate": 1.0859609725538885e-05,
+      "loss": 0.4855,
+      "step": 2943
+    },
+    {
+      "epoch": 1.4973615614470086,
+      "grad_norm": 1.9990509881085727,
+      "learning_rate": 1.0854023361787665e-05,
+      "loss": 0.4927,
+      "step": 2944
+    },
+    {
+      "epoch": 1.4978701761078264,
+      "grad_norm": 2.11750909256137,
+      "learning_rate": 1.0848436729546365e-05,
+      "loss": 0.5355,
+      "step": 2945
+    },
+    {
+      "epoch": 1.4983787907686439,
+      "grad_norm": 2.191148385772657,
+      "learning_rate": 1.084284983057133e-05,
+      "loss": 0.5171,
+      "step": 2946
+    },
+    {
+      "epoch": 1.4988874054294614,
+      "grad_norm": 2.0815886121806364,
+      "learning_rate": 1.0837262666618983e-05,
+      "loss": 0.4407,
+      "step": 2947
+    },
+    {
+      "epoch": 1.4993960200902792,
+      "grad_norm": 1.941473055943122,
+      "learning_rate": 1.0831675239445831e-05,
+      "loss": 0.5224,
+      "step": 2948
+    },
+    {
+      "epoch": 1.4999046347510967,
+      "grad_norm": 1.964360445012961,
+      "learning_rate": 1.082608755080846e-05,
+      "loss": 0.482,
+      "step": 2949
+    },
+    {
+      "epoch": 1.5004132494119142,
+      "grad_norm": 2.3656739284036092,
+      "learning_rate": 1.082049960246354e-05,
+      "loss": 0.5464,
+      "step": 2950
+    },
+    {
+      "epoch": 1.500921864072732,
+      "grad_norm": 2.3956943195205453,
+      "learning_rate": 1.0814911396167835e-05,
+      "loss": 0.5608,
+      "step": 2951
+    },
+    {
+      "epoch": 1.5014304787335495,
+      "grad_norm": 2.0035653521487164,
+      "learning_rate": 1.080932293367817e-05,
+      "loss": 0.4993,
+      "step": 2952
+    },
+    {
+      "epoch": 1.501939093394367,
+      "grad_norm": 1.9060276089376866,
+      "learning_rate": 1.0803734216751466e-05,
+      "loss": 0.5099,
+      "step": 2953
+    },
+    {
+      "epoch": 1.5024477080551848,
+      "grad_norm": 1.7426960976812025,
+      "learning_rate": 1.0798145247144716e-05,
+      "loss": 0.4648,
+      "step": 2954
+    },
+    {
+      "epoch": 1.5029563227160023,
+      "grad_norm": 2.0611054300719482,
+      "learning_rate": 1.0792556026614995e-05,
+      "loss": 0.5787,
+      "step": 2955
+    },
+    {
+      "epoch": 1.5034649373768199,
+      "grad_norm": 2.255928071532473,
+      "learning_rate": 1.0786966556919459e-05,
+      "loss": 0.523,
+      "step": 2956
+    },
+    {
+      "epoch": 1.5039735520376376,
+      "grad_norm": 1.881958700628896,
+      "learning_rate": 1.0781376839815335e-05,
+      "loss": 0.4954,
+      "step": 2957
+    },
+    {
+      "epoch": 1.5044821666984551,
+      "grad_norm": 2.0561547208005435,
+      "learning_rate": 1.0775786877059934e-05,
+      "loss": 0.4971,
+      "step": 2958
+    },
+    {
+      "epoch": 1.5049907813592727,
+      "grad_norm": 1.902851702730625,
+      "learning_rate": 1.0770196670410646e-05,
+      "loss": 0.4699,
+      "step": 2959
+    },
+    {
+      "epoch": 1.5054993960200904,
+      "grad_norm": 2.114522234655685,
+      "learning_rate": 1.0764606221624934e-05,
+      "loss": 0.4905,
+      "step": 2960
+    },
+    {
+      "epoch": 1.5060080106809077,
+      "grad_norm": 2.30424686680908,
+      "learning_rate": 1.075901553246034e-05,
+      "loss": 0.4719,
+      "step": 2961
+    },
+    {
+      "epoch": 1.5065166253417255,
+      "grad_norm": 1.9664642143572448,
+      "learning_rate": 1.0753424604674471e-05,
+      "loss": 0.539,
+      "step": 2962
+    },
+    {
+      "epoch": 1.5070252400025432,
+      "grad_norm": 2.0733154398932063,
+      "learning_rate": 1.0747833440025025e-05,
+      "loss": 0.5029,
+      "step": 2963
+    },
+    {
+      "epoch": 1.5075338546633605,
+      "grad_norm": 1.9745820870858057,
+      "learning_rate": 1.074224204026976e-05,
+      "loss": 0.4875,
+      "step": 2964
+    },
+    {
+      "epoch": 1.5080424693241783,
+      "grad_norm": 2.047109575880495,
+      "learning_rate": 1.0736650407166525e-05,
+      "loss": 0.5151,
+      "step": 2965
+    },
+    {
+      "epoch": 1.5085510839849958,
+      "grad_norm": 2.2482150733259876,
+      "learning_rate": 1.0731058542473225e-05,
+      "loss": 0.554,
+      "step": 2966
+    },
+    {
+      "epoch": 1.5090596986458134,
+      "grad_norm": 2.030098221756284,
+      "learning_rate": 1.0725466447947841e-05,
+      "loss": 0.5048,
+      "step": 2967
+    },
+    {
+      "epoch": 1.509568313306631,
+      "grad_norm": 2.3265889880412702,
+      "learning_rate": 1.071987412534844e-05,
+      "loss": 0.4931,
+      "step": 2968
+    },
+    {
+      "epoch": 1.5100769279674486,
+      "grad_norm": 1.9060393045021413,
+      "learning_rate": 1.0714281576433146e-05,
+      "loss": 0.5749,
+      "step": 2969
+    },
+    {
+      "epoch": 1.5105855426282662,
+      "grad_norm": 2.1011914914873717,
+      "learning_rate": 1.070868880296016e-05,
+      "loss": 0.514,
+      "step": 2970
+    },
+    {
+      "epoch": 1.511094157289084,
+      "grad_norm": 1.9441649947613677,
+      "learning_rate": 1.0703095806687749e-05,
+      "loss": 0.5632,
+      "step": 2971
+    },
+    {
+      "epoch": 1.5116027719499014,
+      "grad_norm": 1.7275583838659483,
+      "learning_rate": 1.0697502589374259e-05,
+      "loss": 0.5965,
+      "step": 2972
+    },
+    {
+      "epoch": 1.512111386610719,
+      "grad_norm": 2.0206307640666026,
+      "learning_rate": 1.0691909152778096e-05,
+      "loss": 0.5918,
+      "step": 2973
+    },
+    {
+      "epoch": 1.5126200012715367,
+      "grad_norm": 1.7662150654506734,
+      "learning_rate": 1.0686315498657743e-05,
+      "loss": 0.5206,
+      "step": 2974
+    },
+    {
+      "epoch": 1.5131286159323543,
+      "grad_norm": 2.1564042409380697,
+      "learning_rate": 1.0680721628771742e-05,
+      "loss": 0.55,
+      "step": 2975
+    },
+    {
+      "epoch": 1.5136372305931718,
+      "grad_norm": 2.1012164543601237,
+      "learning_rate": 1.0675127544878713e-05,
+      "loss": 0.5136,
+      "step": 2976
+    },
+    {
+      "epoch": 1.5141458452539895,
+      "grad_norm": 1.8377464310838827,
+      "learning_rate": 1.0669533248737336e-05,
+      "loss": 0.5258,
+      "step": 2977
+    },
+    {
+      "epoch": 1.514654459914807,
+      "grad_norm": 1.908769911866418,
+      "learning_rate": 1.0663938742106363e-05,
+      "loss": 0.4371,
+      "step": 2978
+    },
+    {
+      "epoch": 1.5151630745756246,
+      "grad_norm": 1.9247313572091864,
+      "learning_rate": 1.065834402674461e-05,
+      "loss": 0.5559,
+      "step": 2979
+    },
+    {
+      "epoch": 1.5156716892364424,
+      "grad_norm": 2.024042105466719,
+      "learning_rate": 1.0652749104410954e-05,
+      "loss": 0.5006,
+      "step": 2980
+    },
+    {
+      "epoch": 1.5161803038972599,
+      "grad_norm": 2.113260716816259,
+      "learning_rate": 1.0647153976864341e-05,
+      "loss": 0.4328,
+      "step": 2981
+    },
+    {
+      "epoch": 1.5166889185580774,
+      "grad_norm": 2.074453451955144,
+      "learning_rate": 1.0641558645863786e-05,
+      "loss": 0.4855,
+      "step": 2982
+    },
+    {
+      "epoch": 1.5171975332188952,
+      "grad_norm": 2.0510721054905376,
+      "learning_rate": 1.0635963113168359e-05,
+      "loss": 0.55,
+      "step": 2983
+    },
+    {
+      "epoch": 1.5177061478797125,
+      "grad_norm": 1.929301729136306,
+      "learning_rate": 1.0630367380537206e-05,
+      "loss": 0.4666,
+      "step": 2984
+    },
+    {
+      "epoch": 1.5182147625405302,
+      "grad_norm": 1.982311707991859,
+      "learning_rate": 1.0624771449729517e-05,
+      "loss": 0.461,
+      "step": 2985
+    },
+    {
+      "epoch": 1.518723377201348,
+      "grad_norm": 2.1291737799578527,
+      "learning_rate": 1.061917532250456e-05,
+      "loss": 0.522,
+      "step": 2986
+    },
+    {
+      "epoch": 1.5192319918621653,
+      "grad_norm": 2.749023264043067,
+      "learning_rate": 1.0613579000621665e-05,
+      "loss": 0.5352,
+      "step": 2987
+    },
+    {
+      "epoch": 1.519740606522983,
+      "grad_norm": 2.2773927451043883,
+      "learning_rate": 1.0607982485840214e-05,
+      "loss": 0.5264,
+      "step": 2988
+    },
+    {
+      "epoch": 1.5202492211838006,
+      "grad_norm": 1.9361051029687983,
+      "learning_rate": 1.0602385779919653e-05,
+      "loss": 0.4702,
+      "step": 2989
+    },
+    {
+      "epoch": 1.520757835844618,
+      "grad_norm": 2.0998148972812434,
+      "learning_rate": 1.059678888461949e-05,
+      "loss": 0.6252,
+      "step": 2990
+    },
+    {
+      "epoch": 1.5212664505054359,
+      "grad_norm": 1.9848658637936767,
+      "learning_rate": 1.0591191801699288e-05,
+      "loss": 0.5405,
+      "step": 2991
+    },
+    {
+      "epoch": 1.5217750651662534,
+      "grad_norm": 2.0156863890769583,
+      "learning_rate": 1.0585594532918678e-05,
+      "loss": 0.504,
+      "step": 2992
+    },
+    {
+      "epoch": 1.522283679827071,
+      "grad_norm": 1.774846529706163,
+      "learning_rate": 1.0579997080037341e-05,
+      "loss": 0.527,
+      "step": 2993
+    },
+    {
+      "epoch": 1.5227922944878887,
+      "grad_norm": 2.135508192628558,
+      "learning_rate": 1.057439944481502e-05,
+      "loss": 0.542,
+      "step": 2994
+    },
+    {
+      "epoch": 1.5233009091487062,
+      "grad_norm": 1.9889143439865513,
+      "learning_rate": 1.0568801629011512e-05,
+      "loss": 0.4364,
+      "step": 2995
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 1.7538554772587374,
+      "learning_rate": 1.0563203634386673e-05,
+      "loss": 0.5031,
+      "step": 2996
+    },
+    {
+      "epoch": 1.5243181384703415,
+      "grad_norm": 2.31753353074064,
+      "learning_rate": 1.055760546270042e-05,
+      "loss": 0.4941,
+      "step": 2997
+    },
+    {
+      "epoch": 1.524826753131159,
+      "grad_norm": 1.899410798920718,
+      "learning_rate": 1.0552007115712714e-05,
+      "loss": 0.5477,
+      "step": 2998
+    },
+    {
+      "epoch": 1.5253353677919765,
+      "grad_norm": 1.8183475743944424,
+      "learning_rate": 1.0546408595183579e-05,
+      "loss": 0.4402,
+      "step": 2999
+    },
+    {
+      "epoch": 1.5258439824527943,
+      "grad_norm": 2.34515918329678,
+      "learning_rate": 1.0540809902873097e-05,
+      "loss": 0.4776,
+      "step": 3000
+    },
+    {
+      "epoch": 1.5258439824527943,
+      "eval_loss": 0.7014508247375488,
+      "eval_runtime": 169.7493,
+      "eval_samples_per_second": 19.511,
+      "eval_steps_per_second": 9.756,
+      "step": 3000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5898,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.5878170617616794e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}