diff --git "a/checkpoint-915840/trainer_state.json" "b/checkpoint-915840/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-915840/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_metric": 3.8417012691497803,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-domain/transformer/4/checkpoints/checkpoint-915840",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 915840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.9264,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8199,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1893,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9797,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.8175,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.7212,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.6053,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5389,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4515,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.4045,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3529,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3134,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2625,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2117,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1768,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1354,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.1058,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0837,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0543,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0221,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.0098,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 4.9704,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9542,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.93,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9181,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8909,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8728,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8517,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8332,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8185,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.8001,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7957,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.776,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 4.7707,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 4.7624,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 4.7444,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7346,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.7079,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7037,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6853,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6857,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6719,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6633,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6426,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6525,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6383,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6305,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6229,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5926,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5901,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5964,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5881,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.571,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5513,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 4.5545,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 4.5404,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 4.5657,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 4.5251,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 4.5367,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 4.5324,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5076,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5125,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 4.5017,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 4.4842,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 4.4854,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.5028,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4806,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 4.4706,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 4.4615,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4614,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4736,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4711,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.458,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4632,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4648,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4513,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4386,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.4299,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4293,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4222,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4163,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4192,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4198,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.416,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3902,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3907,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.398,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.4007,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3826,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.384,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3869,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3737,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922878764749932e-05,
+      "loss": 4.3686,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92204016999888e-05,
+      "loss": 4.3723,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921201575247828e-05,
+      "loss": 4.3633,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3676,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3596,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.344,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3532,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3539,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3485,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3486,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.3428,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.3362,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3393,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.314,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91114171399595e-05,
+      "loss": 4.3182,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3248,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.319,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.311,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3075,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.3041,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.3087,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3046,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.3048,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.2987,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.3041,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.2987,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.2962,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.2964,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.2962,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.284,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.2892,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.2783,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.2888,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2639,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2764,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2615,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2679,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.252,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2816,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.269,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.2556,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.2574,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.2608,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.2651,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.2624,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2485,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2569,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.244,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2561,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2414,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.243,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.24,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.2307,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.2379,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2342,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.2321,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.2344,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.226,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1915082931518555,
+      "eval_runtime": 320.2057,
+      "eval_samples_per_second": 1191.706,
+      "eval_steps_per_second": 37.242,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.2104,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.2084,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2254,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2118,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.2294,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.1993,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.2081,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.1915,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2069,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.201,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.2054,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.2021,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864198424621142e-05,
+      "loss": 4.1892,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86335982987009e-05,
+      "loss": 4.1876,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862521235119038e-05,
+      "loss": 4.1844,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.1774,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.187,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.1841,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.1778,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1982,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.1713,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.182,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.1768,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.1842,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541385633692644e-05,
+      "loss": 4.1608,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532999686182124e-05,
+      "loss": 4.169,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524613738671604e-05,
+      "loss": 4.164,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.1687,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1518,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.1555,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.1639,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.1597,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.1584,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.1617,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.1576,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.1579,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.1445,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.1495,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.1377,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841564555744604e-05,
+      "loss": 4.1449,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840725960993552e-05,
+      "loss": 4.1458,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.1455,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.1321,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382118146207687e-05,
+      "loss": 4.1413,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.1448,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.1368,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.1405,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834859073496934e-05,
+      "loss": 4.1127,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834020478745882e-05,
+      "loss": 4.119,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331818839948295e-05,
+      "loss": 4.1364,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.128,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831506332373099e-05,
+      "loss": 4.1225,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830667737622047e-05,
+      "loss": 4.1073,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298291428709944e-05,
+      "loss": 4.1132,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289905481199424e-05,
+      "loss": 4.1054,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828153591249264e-05,
+      "loss": 4.1301,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.1055,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.1168,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.1218,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248008501254296e-05,
+      "loss": 4.1006,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823962255374377e-05,
+      "loss": 4.1085,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823123660623325e-05,
+      "loss": 4.0999,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822285065872273e-05,
+      "loss": 4.0958,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.0961,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820609514250542e-05,
+      "loss": 4.1121,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81977091949949e-05,
+      "loss": 4.1028,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818932324748438e-05,
+      "loss": 4.0927,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.091,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172567731267074e-05,
+      "loss": 4.0917,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164181783756554e-05,
+      "loss": 4.1077,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155795836246034e-05,
+      "loss": 4.1082,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.1,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813904032002872e-05,
+      "loss": 4.1103,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81306543725182e-05,
+      "loss": 4.1093,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812226842500768e-05,
+      "loss": 4.1043,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0937,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810551290879037e-05,
+      "loss": 4.0918,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809712696127985e-05,
+      "loss": 4.0898,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808874101376933e-05,
+      "loss": 4.0884,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808035506625881e-05,
+      "loss": 4.084,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807198549755203e-05,
+      "loss": 4.0929,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806359955004151e-05,
+      "loss": 4.0971,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805521360253099e-05,
+      "loss": 4.0894,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804682765502047e-05,
+      "loss": 4.0731,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8038458086313677e-05,
+      "loss": 4.0736,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030072138803156e-05,
+      "loss": 4.0807,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021686191292636e-05,
+      "loss": 4.0887,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013300243782116e-05,
+      "loss": 4.0734,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8004930675075325e-05,
+      "loss": 4.0768,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996544727564805e-05,
+      "loss": 4.0818,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988158780054285e-05,
+      "loss": 4.0745,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979772832543765e-05,
+      "loss": 4.0687,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797140326383698e-05,
+      "loss": 4.0726,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796301731632646e-05,
+      "loss": 4.0736,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.0745,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.0754,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793787585259863e-05,
+      "loss": 4.0576,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792948990508811e-05,
+      "loss": 4.0725,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792110395757759e-05,
+      "loss": 4.0665,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791271801006707e-05,
+      "loss": 4.0739,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790433206255655e-05,
+      "loss": 4.0686,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789594611504603e-05,
+      "loss": 4.0677,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788757654633924e-05,
+      "loss": 4.0626,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787919059882872e-05,
+      "loss": 4.0687,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78708046513182e-05,
+      "loss": 4.0456,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786241870380768e-05,
+      "loss": 4.0517,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854049135100895e-05,
+      "loss": 4.0587,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845663187590375e-05,
+      "loss": 4.0534,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837277240079855e-05,
+      "loss": 4.048,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828891292569335e-05,
+      "loss": 4.047,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820505345058815e-05,
+      "loss": 4.0465,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812119397548295e-05,
+      "loss": 4.0539,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0482,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795363881330984e-05,
+      "loss": 4.0501,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786977933820464e-05,
+      "loss": 4.0524,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778591986309944e-05,
+      "loss": 4.0559,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0516,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776183647009263e-05,
+      "loss": 4.0479,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775345052258212e-05,
+      "loss": 4.0537,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450645750716e-05,
+      "loss": 4.0535,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773667862756108e-05,
+      "loss": 4.0406,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772830905885429e-05,
+      "loss": 4.0507,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771992311134377e-05,
+      "loss": 4.0402,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771153716383325e-05,
+      "loss": 4.0516,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770315121632272e-05,
+      "loss": 4.0275,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769478164761594e-05,
+      "loss": 4.0452,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768639570010542e-05,
+      "loss": 4.0274,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76780097525949e-05,
+      "loss": 4.0329,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766962380508437e-05,
+      "loss": 4.0235,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7661254236377586e-05,
+      "loss": 4.0526,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765286828886707e-05,
+      "loss": 4.042,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764448234135655e-05,
+      "loss": 4.034,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636096393846026e-05,
+      "loss": 4.0287,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762772682513924e-05,
+      "loss": 4.038,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761934087762872e-05,
+      "loss": 4.0431,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610954930118195e-05,
+      "loss": 4.042,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602568982607675e-05,
+      "loss": 4.0319,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759419941390089e-05,
+      "loss": 4.0369,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758581346639037e-05,
+      "loss": 4.0307,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577427518879844e-05,
+      "loss": 4.0443,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569041571369324e-05,
+      "loss": 4.0239,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756067200266254e-05,
+      "loss": 4.0287,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755228605515202e-05,
+      "loss": 4.0257,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75439001076415e-05,
+      "loss": 4.0216,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753551416013098e-05,
+      "loss": 4.0321,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527144591424196e-05,
+      "loss": 4.0255,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751875864391367e-05,
+      "loss": 4.0274,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751037269640315e-05,
+      "loss": 4.0254,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750198674889263e-05,
+      "loss": 4.0213,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.023377895355225,
+      "eval_runtime": 603.3722,
+      "eval_samples_per_second": 632.431,
+      "eval_steps_per_second": 19.764,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749360080138211e-05,
+      "loss": 4.0078,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748521485387159e-05,
+      "loss": 4.0078,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747682890636107e-05,
+      "loss": 4.0244,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746844295885055e-05,
+      "loss": 4.012,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746005701134003e-05,
+      "loss": 4.0348,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745167106382951e-05,
+      "loss": 4.0028,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744328511631899e-05,
+      "loss": 4.0128,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434899168808475e-05,
+      "loss": 3.9979,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426513221297955e-05,
+      "loss": 4.0145,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418127273787435e-05,
+      "loss": 4.01,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740974132627691e-05,
+      "loss": 4.0133,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401371757570124e-05,
+      "loss": 4.0092,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739300218886333e-05,
+      "loss": 3.9992,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738461624135281e-05,
+      "loss": 4.0038,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737623029384229e-05,
+      "loss": 4.0002,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736784434633177e-05,
+      "loss": 3.9943,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735945839882125e-05,
+      "loss": 3.9994,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735107245131073e-05,
+      "loss": 4.0027,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734268650380021e-05,
+      "loss": 3.9951,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733431693509343e-05,
+      "loss": 4.0166,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732593098758291e-05,
+      "loss": 3.9938,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 4.0012,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.9981,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730078952385508e-05,
+      "loss": 4.0073,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729240357634456e-05,
+      "loss": 3.9849,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728401762883403e-05,
+      "loss": 3.9983,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727563168132351e-05,
+      "loss": 3.9907,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726724573381299e-05,
+      "loss": 3.9946,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725885978630247e-05,
+      "loss": 3.9794,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725049021759568e-05,
+      "loss": 3.9839,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724210427008517e-05,
+      "loss": 3.9922,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233718322574647e-05,
+      "loss": 3.9925,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225332375064127e-05,
+      "loss": 3.992,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216962806357336e-05,
+      "loss": 3.9933,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208576858846816e-05,
+      "loss": 3.9882,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200190911336296e-05,
+      "loss": 3.9925,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191804963825775e-05,
+      "loss": 3.9785,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7183435395118985e-05,
+      "loss": 3.9911,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175049447608465e-05,
+      "loss": 3.9716,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166663500097945e-05,
+      "loss": 3.9816,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158293931391154e-05,
+      "loss": 3.9813,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149907983880634e-05,
+      "loss": 3.9877,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714152203637012e-05,
+      "loss": 3.9789,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71331360888596e-05,
+      "loss": 3.9774,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9861,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711638057264229e-05,
+      "loss": 3.9812,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710799462513177e-05,
+      "loss": 3.9815,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709960867762125e-05,
+      "loss": 3.9643,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709123910891446e-05,
+      "loss": 3.9604,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708285316140394e-05,
+      "loss": 3.9765,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707446721389342e-05,
+      "loss": 3.974,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70660812663829e-05,
+      "loss": 3.9725,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9524,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9602,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040939802655074e-05,
+      "loss": 3.9601,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032553855144554e-05,
+      "loss": 3.9716,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9562,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9692,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9707,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.9558,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699065687519941e-05,
+      "loss": 3.9599,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698227092768889e-05,
+      "loss": 3.9571,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697388498017837e-05,
+      "loss": 3.9475,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9566,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695712946396106e-05,
+      "loss": 3.9623,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694875989525428e-05,
+      "loss": 3.9668,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694037394774376e-05,
+      "loss": 3.9441,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693198800023324e-05,
+      "loss": 3.9502,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692360205272272e-05,
+      "loss": 3.9479,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69152161052122e-05,
+      "loss": 3.9729,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906830157701677e-05,
+      "loss": 3.9585,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9586,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9729,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9698,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.967,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9546,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.956,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9544,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9506,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683138938771446e-05,
+      "loss": 3.9522,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682300344020394e-05,
+      "loss": 3.9552,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681461749269342e-05,
+      "loss": 3.9622,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68062315451829e-05,
+      "loss": 3.9516,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9447,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9383,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9493,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677270413394455e-05,
+      "loss": 3.9551,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676431818643403e-05,
+      "loss": 3.9443,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675593223892351e-05,
+      "loss": 3.9409,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.954,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67391767227062e-05,
+      "loss": 3.9365,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673079077519568e-05,
+      "loss": 3.9439,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672240482768516e-05,
+      "loss": 3.9436,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671401888017464e-05,
+      "loss": 3.9484,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9484,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9427,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688877416446815e-05,
+      "loss": 3.9324,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668049146893629e-05,
+      "loss": 3.9465,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667210552142577e-05,
+      "loss": 3.9371,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666371957391525e-05,
+      "loss": 3.9525,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665535000520846e-05,
+      "loss": 3.9437,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664696405769794e-05,
+      "loss": 3.9442,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663857811018742e-05,
+      "loss": 3.9342,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66301921626769e-05,
+      "loss": 3.9491,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621806215166383e-05,
+      "loss": 3.9195,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661343664645959e-05,
+      "loss": 3.9287,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660505069894907e-05,
+      "loss": 3.9354,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659666475143855e-05,
+      "loss": 3.9347,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658827880392803e-05,
+      "loss": 3.924,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657989285641751e-05,
+      "loss": 3.9299,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657150690890699e-05,
+      "loss": 3.9243,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65631373402002e-05,
+      "loss": 3.9353,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655475139268968e-05,
+      "loss": 3.9274,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654636544517916e-05,
+      "loss": 3.9263,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653797949766864e-05,
+      "loss": 3.9371,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652960992896185e-05,
+      "loss": 3.9382,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652122398145134e-05,
+      "loss": 3.9361,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651283803394082e-05,
+      "loss": 3.9272,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65044520864303e-05,
+      "loss": 3.9372,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496082517723506e-05,
+      "loss": 3.9332,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487696570212986e-05,
+      "loss": 3.9246,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479310622702466e-05,
+      "loss": 3.9392,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470924675191946e-05,
+      "loss": 3.9204,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462538727681426e-05,
+      "loss": 3.9357,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454152780170906e-05,
+      "loss": 3.9155,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445766832660386e-05,
+      "loss": 3.9303,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437380885149866e-05,
+      "loss": 3.9119,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9191,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642064174773629e-05,
+      "loss": 3.9121,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641225580022577e-05,
+      "loss": 3.9368,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640386985271525e-05,
+      "loss": 3.9284,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639550028400846e-05,
+      "loss": 3.9244,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638711433649794e-05,
+      "loss": 3.9167,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637872838898742e-05,
+      "loss": 3.9232,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63703424414769e-05,
+      "loss": 3.9357,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636195649396638e-05,
+      "loss": 3.9267,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635358692525959e-05,
+      "loss": 3.9201,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634520097774907e-05,
+      "loss": 3.9302,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633681503023855e-05,
+      "loss": 3.9209,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632842908272803e-05,
+      "loss": 3.9294,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632004313521751e-05,
+      "loss": 3.9152,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.9203,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303287619000204e-05,
+      "loss": 3.9179,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294901671489684e-05,
+      "loss": 3.9158,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286515723979164e-05,
+      "loss": 3.9176,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9235,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.9141,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.917,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9161,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9497029781341553,
+      "eval_runtime": 306.2528,
+      "eval_samples_per_second": 1246.0,
+      "eval_steps_per_second": 38.938,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.9081,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9031,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9181,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9044,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621109133279568e-05,
+      "loss": 3.9323,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620270538528516e-05,
+      "loss": 3.8967,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619431943777464e-05,
+      "loss": 3.9087,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.893,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617754754275359e-05,
+      "loss": 3.9096,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616917797404681e-05,
+      "loss": 3.9102,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.9046,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9096,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614402013151524e-05,
+      "loss": 3.8943,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.9036,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.9004,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.8971,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8978,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610212315157011e-05,
+      "loss": 3.8988,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609373720405959e-05,
+      "loss": 3.897,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085351256549065e-05,
+      "loss": 3.9125,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.8939,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.9031,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606020979282124e-05,
+      "loss": 3.8999,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051823845310714e-05,
+      "loss": 3.9088,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043437897800194e-05,
+      "loss": 3.8871,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.9024,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602668238158289e-05,
+      "loss": 3.893,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601829643407237e-05,
+      "loss": 3.8947,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600991048656185e-05,
+      "loss": 3.8813,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.8925,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599315497034454e-05,
+      "loss": 3.8926,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598476902283402e-05,
+      "loss": 3.8979,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59763830753235e-05,
+      "loss": 3.8946,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.8997,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595962755910619e-05,
+      "loss": 3.8933,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595124161159567e-05,
+      "loss": 3.8977,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594285566408515e-05,
+      "loss": 3.8869,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593446971657463e-05,
+      "loss": 3.8937,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926100147867836e-05,
+      "loss": 3.8774,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591771420035732e-05,
+      "loss": 3.8873,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59093282528468e-05,
+      "loss": 3.8892,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590094230533628e-05,
+      "loss": 3.8947,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589257273662949e-05,
+      "loss": 3.8863,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588418678911897e-05,
+      "loss": 3.8857,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587580084160845e-05,
+      "loss": 3.893,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586741489409793e-05,
+      "loss": 3.8883,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585904532539114e-05,
+      "loss": 3.8887,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585065937788062e-05,
+      "loss": 3.8779,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58422734303701e-05,
+      "loss": 3.8634,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583388748285958e-05,
+      "loss": 3.8852,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582551791415279e-05,
+      "loss": 3.881,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581713196664228e-05,
+      "loss": 3.8789,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580874601913176e-05,
+      "loss": 3.8698,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580036007162124e-05,
+      "loss": 3.8694,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791990502914446e-05,
+      "loss": 3.8651,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783604555403926e-05,
+      "loss": 3.8832,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5775218607893406e-05,
+      "loss": 3.8702,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766832660382886e-05,
+      "loss": 3.8761,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758463091676095e-05,
+      "loss": 3.8851,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750077144165575e-05,
+      "loss": 3.8676,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741691196655055e-05,
+      "loss": 3.8698,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733305249144535e-05,
+      "loss": 3.8738,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724935680437744e-05,
+      "loss": 3.8556,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571654973292723e-05,
+      "loss": 3.8703,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570816378541671e-05,
+      "loss": 3.8774,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569977783790619e-05,
+      "loss": 3.8774,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56914082691994e-05,
+      "loss": 3.8595,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568302232168888e-05,
+      "loss": 3.8662,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567463637417836e-05,
+      "loss": 3.8606,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566625042666784e-05,
+      "loss": 3.8809,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565786447915732e-05,
+      "loss": 3.8741,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564949491045053e-05,
+      "loss": 3.8752,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564110896294001e-05,
+      "loss": 3.8855,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563272301542949e-05,
+      "loss": 3.8847,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56243534467227e-05,
+      "loss": 3.8848,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615967499212184e-05,
+      "loss": 3.8637,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607581551701664e-05,
+      "loss": 3.8733,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599195604191144e-05,
+      "loss": 3.8738,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559082603548435e-05,
+      "loss": 3.8652,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558244008797383e-05,
+      "loss": 3.8721,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557405414046331e-05,
+      "loss": 3.8695,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556566819295279e-05,
+      "loss": 3.8764,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557298624246e-05,
+      "loss": 3.8698,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554891267673548e-05,
+      "loss": 3.8548,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554052672922496e-05,
+      "loss": 3.8585,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553214078171444e-05,
+      "loss": 3.868,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552377121300765e-05,
+      "loss": 3.8705,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551538526549714e-05,
+      "loss": 3.8631,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550699931798662e-05,
+      "loss": 3.8572,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54986133704761e-05,
+      "loss": 3.8703,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549024380176931e-05,
+      "loss": 3.8569,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548185785425879e-05,
+      "loss": 3.8589,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547347190674827e-05,
+      "loss": 3.8656,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546508595923775e-05,
+      "loss": 3.8699,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545670001172723e-05,
+      "loss": 3.8696,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544831406421671e-05,
+      "loss": 3.8612,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543992811670619e-05,
+      "loss": 3.8472,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431542169195667e-05,
+      "loss": 3.8685,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423172600488876e-05,
+      "loss": 3.854,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414786652978356e-05,
+      "loss": 3.8743,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406400705467836e-05,
+      "loss": 3.8657,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539801475795732e-05,
+      "loss": 3.8635,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538964518925053e-05,
+      "loss": 3.8542,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538125924174001e-05,
+      "loss": 3.868,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537287329422949e-05,
+      "loss": 3.8445,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536448734671897e-05,
+      "loss": 3.8535,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535611777801218e-05,
+      "loss": 3.8507,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534773183050166e-05,
+      "loss": 3.8562,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533934588299114e-05,
+      "loss": 3.8469,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330959935480613e-05,
+      "loss": 3.8526,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532259036677383e-05,
+      "loss": 3.851,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531420441926331e-05,
+      "loss": 3.8496,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530581847175279e-05,
+      "loss": 3.8542,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529743252424227e-05,
+      "loss": 3.8457,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5289062955535485e-05,
+      "loss": 3.8644,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5280677008024965e-05,
+      "loss": 3.8601,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272291060514445e-05,
+      "loss": 3.8595,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526390511300392e-05,
+      "loss": 3.8515,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255535544297134e-05,
+      "loss": 3.8595,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247149596786614e-05,
+      "loss": 3.8601,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523876364927609e-05,
+      "loss": 3.8474,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523037770176557e-05,
+      "loss": 3.8607,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522200813305878e-05,
+      "loss": 3.8467,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521362218554826e-05,
+      "loss": 3.8579,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520523623803774e-05,
+      "loss": 3.843,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519685029052722e-05,
+      "loss": 3.8544,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518848072182044e-05,
+      "loss": 3.8353,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518009477430992e-05,
+      "loss": 3.8441,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517170882679939e-05,
+      "loss": 3.8409,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516332287928887e-05,
+      "loss": 3.8575,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515495331058209e-05,
+      "loss": 3.8497,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514656736307156e-05,
+      "loss": 3.8526,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513818141556104e-05,
+      "loss": 3.8433,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512979546805052e-05,
+      "loss": 3.8513,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512142589934374e-05,
+      "loss": 3.86,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511303995183322e-05,
+      "loss": 3.8519,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104654004322697e-05,
+      "loss": 3.8475,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5096268056812177e-05,
+      "loss": 3.8571,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508789848810539e-05,
+      "loss": 3.8467,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079512540594866e-05,
+      "loss": 3.8581,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5071126593084346e-05,
+      "loss": 3.8388,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062740645573826e-05,
+      "loss": 3.8495,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054371076867035e-05,
+      "loss": 3.8432,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045985129356515e-05,
+      "loss": 3.8431,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037599181845995e-05,
+      "loss": 3.849,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502922961313921e-05,
+      "loss": 3.8466,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020843665628684e-05,
+      "loss": 3.8452,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501245771811817e-05,
+      "loss": 3.8396,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500407177060765e-05,
+      "loss": 3.8489,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.908376932144165,
+      "eval_runtime": 305.9494,
+      "eval_samples_per_second": 1247.236,
+      "eval_steps_per_second": 38.977,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499568582309713e-05,
+      "loss": 3.8325,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498729987558661e-05,
+      "loss": 3.831,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497891392807609e-05,
+      "loss": 3.8471,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497052798056557e-05,
+      "loss": 3.8349,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496214203305505e-05,
+      "loss": 3.8602,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495375608554453e-05,
+      "loss": 3.8343,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494537013803401e-05,
+      "loss": 3.8348,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493698419052349e-05,
+      "loss": 3.8213,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49286146218167e-05,
+      "loss": 3.838,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492024505310991e-05,
+      "loss": 3.8449,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491185910559939e-05,
+      "loss": 3.8333,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490347315808887e-05,
+      "loss": 3.8347,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4895087210578355e-05,
+      "loss": 3.827,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886701263067835e-05,
+      "loss": 3.834,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878315315557315e-05,
+      "loss": 3.8319,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869929368046795e-05,
+      "loss": 3.8266,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861543420536275e-05,
+      "loss": 3.8281,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853173851829484e-05,
+      "loss": 3.8339,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844787904318964e-05,
+      "loss": 3.8263,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836401956808444e-05,
+      "loss": 3.8411,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4828016009297924e-05,
+      "loss": 3.8306,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481964644059113e-05,
+      "loss": 3.8368,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481126049308061e-05,
+      "loss": 3.8308,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480287454557009e-05,
+      "loss": 3.8388,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479448859805957e-05,
+      "loss": 3.8191,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478611902935279e-05,
+      "loss": 3.8316,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477773308184227e-05,
+      "loss": 3.8294,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476934713433175e-05,
+      "loss": 3.8246,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476096118682122e-05,
+      "loss": 3.8164,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475259161811444e-05,
+      "loss": 3.8227,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474420567060392e-05,
+      "loss": 3.8314,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47358197230934e-05,
+      "loss": 3.8302,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472743377558287e-05,
+      "loss": 3.8312,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4719064206876086e-05,
+      "loss": 3.8308,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710678259365566e-05,
+      "loss": 3.8256,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4702292311855046e-05,
+      "loss": 3.83,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693906364344526e-05,
+      "loss": 3.8247,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468553679563774e-05,
+      "loss": 3.8244,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467715084812722e-05,
+      "loss": 3.816,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668764900616695e-05,
+      "loss": 3.8219,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466039533190991e-05,
+      "loss": 3.8203,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465200938439939e-05,
+      "loss": 3.8301,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464362343688887e-05,
+      "loss": 3.8224,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635237489378344e-05,
+      "loss": 3.8217,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462686792067156e-05,
+      "loss": 3.8323,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461848197316104e-05,
+      "loss": 3.822,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461009602565052e-05,
+      "loss": 3.8228,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460171007814e-05,
+      "loss": 3.8129,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593340509433216e-05,
+      "loss": 3.799,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584954561922696e-05,
+      "loss": 3.8232,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457656861441217e-05,
+      "loss": 3.819,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456818266690165e-05,
+      "loss": 3.8176,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559813098194865e-05,
+      "loss": 3.802,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551427150684345e-05,
+      "loss": 3.81,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454304120317382e-05,
+      "loss": 3.798,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45346552556633e-05,
+      "loss": 3.8207,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526285686956514e-05,
+      "loss": 3.8083,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517899739445994e-05,
+      "loss": 3.8105,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450951379193547e-05,
+      "loss": 3.8227,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501127844424954e-05,
+      "loss": 3.8066,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449275827571816e-05,
+      "loss": 3.8084,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448437232820764e-05,
+      "loss": 3.8134,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447598638069712e-05,
+      "loss": 3.7935,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44676004331866e-05,
+      "loss": 3.8087,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445923086447982e-05,
+      "loss": 3.8111,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445084491696929e-05,
+      "loss": 3.8171,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444245896945877e-05,
+      "loss": 3.7971,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443407302194825e-05,
+      "loss": 3.8044,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442570345324147e-05,
+      "loss": 3.797,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441731750573094e-05,
+      "loss": 3.8212,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440893155822042e-05,
+      "loss": 3.8132,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44005456107099e-05,
+      "loss": 3.8107,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439219242080685e-05,
+      "loss": 3.8255,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438380647329633e-05,
+      "loss": 3.8175,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437542052578581e-05,
+      "loss": 3.8315,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436703457827529e-05,
+      "loss": 3.805,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358648630764765e-05,
+      "loss": 3.8113,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350262683254245e-05,
+      "loss": 3.8137,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341876735743725e-05,
+      "loss": 3.8019,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333490788233205e-05,
+      "loss": 3.8149,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325121219526414e-05,
+      "loss": 3.8084,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316735272015894e-05,
+      "loss": 3.8152,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308349324505374e-05,
+      "loss": 3.8098,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299963376994854e-05,
+      "loss": 3.7982,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429159380828807e-05,
+      "loss": 3.8054,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428320786077755e-05,
+      "loss": 3.8015,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427482191326703e-05,
+      "loss": 3.8137,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426643596575651e-05,
+      "loss": 3.8053,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425806639704972e-05,
+      "loss": 3.7959,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42496804495392e-05,
+      "loss": 3.809,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424131088083241e-05,
+      "loss": 3.8046,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423292493332189e-05,
+      "loss": 3.7943,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422453898581137e-05,
+      "loss": 3.8048,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421615303830085e-05,
+      "loss": 3.8126,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420776709079033e-05,
+      "loss": 3.8098,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419938114327981e-05,
+      "loss": 3.8048,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190995195769294e-05,
+      "loss": 3.7881,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182609248258774e-05,
+      "loss": 3.8075,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174239679551984e-05,
+      "loss": 3.7981,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165853732041464e-05,
+      "loss": 3.8118,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157467784530943e-05,
+      "loss": 3.8108,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4149081837020423e-05,
+      "loss": 3.8049,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414071226831363e-05,
+      "loss": 3.7968,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413234269960684e-05,
+      "loss": 3.8122,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412395675209632e-05,
+      "loss": 3.7829,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41155708045858e-05,
+      "loss": 3.7972,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410718485707528e-05,
+      "loss": 3.7906,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409879890956476e-05,
+      "loss": 3.8042,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409041296205425e-05,
+      "loss": 3.7857,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408202701454373e-05,
+      "loss": 3.7936,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407364106703321e-05,
+      "loss": 3.794,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406527149832642e-05,
+      "loss": 3.794,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40568855508159e-05,
+      "loss": 3.796,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404849960330538e-05,
+      "loss": 3.7925,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4040130034598586e-05,
+      "loss": 3.8068,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031744087088066e-05,
+      "loss": 3.8042,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4023358139577546e-05,
+      "loss": 3.7993,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014972192067026e-05,
+      "loss": 3.7926,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4006602623360235e-05,
+      "loss": 3.8049,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3998216675849715e-05,
+      "loss": 3.799,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39898307283392e-05,
+      "loss": 3.7905,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398144478082868e-05,
+      "loss": 3.8067,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397305883331816e-05,
+      "loss": 3.7887,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396467288580764e-05,
+      "loss": 3.8005,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395628693829712e-05,
+      "loss": 3.7876,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39479009907866e-05,
+      "loss": 3.7991,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393953142207981e-05,
+      "loss": 3.785,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393114547456929e-05,
+      "loss": 3.7864,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39227759058625e-05,
+      "loss": 3.789,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391438995835198e-05,
+      "loss": 3.8013,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390600401084146e-05,
+      "loss": 3.7942,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389761806333094e-05,
+      "loss": 3.7972,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388923211582042e-05,
+      "loss": 3.7855,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38808461683099e-05,
+      "loss": 3.7995,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387246022079938e-05,
+      "loss": 3.8043,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3864090652092595e-05,
+      "loss": 3.7992,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855704704582075e-05,
+      "loss": 3.7931,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384731875707155e-05,
+      "loss": 3.7996,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383893280956103e-05,
+      "loss": 3.7925,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830563240854244e-05,
+      "loss": 3.8022,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822193672147453e-05,
+      "loss": 3.788,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381380772463693e-05,
+      "loss": 3.7961,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380542177712641e-05,
+      "loss": 3.7893,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379703582961589e-05,
+      "loss": 3.7906,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378866626090911e-05,
+      "loss": 3.7927,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378028031339859e-05,
+      "loss": 3.791,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377189436588807e-05,
+      "loss": 3.7938,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376350841837754e-05,
+      "loss": 3.7819,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375512247086702e-05,
+      "loss": 3.7957,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.883833646774292,
+      "eval_runtime": 305.722,
+      "eval_samples_per_second": 1248.163,
+      "eval_steps_per_second": 39.006,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37467365233565e-05,
+      "loss": 3.7866,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373835057584598e-05,
+      "loss": 3.7759,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372998100713919e-05,
+      "loss": 3.795,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372159505962867e-05,
+      "loss": 3.7839,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371320911211815e-05,
+      "loss": 3.8029,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370482316460763e-05,
+      "loss": 3.7839,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369645359590085e-05,
+      "loss": 3.7754,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368806764839033e-05,
+      "loss": 3.7778,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367968170087981e-05,
+      "loss": 3.7812,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367129575336929e-05,
+      "loss": 3.7919,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366290980585877e-05,
+      "loss": 3.78,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365452385834825e-05,
+      "loss": 3.7857,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3646154289641456e-05,
+      "loss": 3.775,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637768342130936e-05,
+      "loss": 3.7777,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3629382394620416e-05,
+      "loss": 3.7831,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620996447109896e-05,
+      "loss": 3.7729,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612610499599376e-05,
+      "loss": 3.7731,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604224552088856e-05,
+      "loss": 3.7808,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595838604578336e-05,
+      "loss": 3.7773,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587452657067816e-05,
+      "loss": 3.788,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3579083088361025e-05,
+      "loss": 3.78,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357069714085051e-05,
+      "loss": 3.7849,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356231119333999e-05,
+      "loss": 3.7815,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355392524582947e-05,
+      "loss": 3.785,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354555567712268e-05,
+      "loss": 3.7667,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353716972961216e-05,
+      "loss": 3.7863,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352880016090537e-05,
+      "loss": 3.7739,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352041421339485e-05,
+      "loss": 3.7731,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351202826588433e-05,
+      "loss": 3.769,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350364231837381e-05,
+      "loss": 3.7707,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349525637086329e-05,
+      "loss": 3.7815,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34868868021565e-05,
+      "loss": 3.7763,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347850085464598e-05,
+      "loss": 3.7849,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470114907135465e-05,
+      "loss": 3.7828,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461728959624945e-05,
+      "loss": 3.7731,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453359390918154e-05,
+      "loss": 3.7796,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444973443407634e-05,
+      "loss": 3.7753,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436587495897114e-05,
+      "loss": 3.7732,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3428201548386594e-05,
+      "loss": 3.7673,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419815600876074e-05,
+      "loss": 3.768,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411429653365554e-05,
+      "loss": 3.7737,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340304370585503e-05,
+      "loss": 3.7807,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339465775834451e-05,
+      "loss": 3.7697,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338628818963772e-05,
+      "loss": 3.7731,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33779022421272e-05,
+      "loss": 3.7786,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336951629461668e-05,
+      "loss": 3.7759,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336113034710616e-05,
+      "loss": 3.7718,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335274439959564e-05,
+      "loss": 3.7658,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334435845208512e-05,
+      "loss": 3.7452,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33359725045746e-05,
+      "loss": 3.7787,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332758655706408e-05,
+      "loss": 3.7692,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331921698835729e-05,
+      "loss": 3.7708,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33108474196505e-05,
+      "loss": 3.7579,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330246147213998e-05,
+      "loss": 3.7532,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329407552462946e-05,
+      "loss": 3.7526,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328568957711894e-05,
+      "loss": 3.7669,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327730362960842e-05,
+      "loss": 3.7594,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689176820979e-05,
+      "loss": 3.7644,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260548113391116e-05,
+      "loss": 3.7721,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3252162165880596e-05,
+      "loss": 3.7596,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3243776218370076e-05,
+      "loss": 3.7611,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3235390270859556e-05,
+      "loss": 3.7674,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3227004323349036e-05,
+      "loss": 3.7471,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218618375838516e-05,
+      "loss": 3.7582,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3210232428327996e-05,
+      "loss": 3.7633,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201846480817476e-05,
+      "loss": 3.7694,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193476912110685e-05,
+      "loss": 3.7467,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185107343403894e-05,
+      "loss": 3.7578,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176721395893374e-05,
+      "loss": 3.7493,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168335448382854e-05,
+      "loss": 3.7702,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315994950087234e-05,
+      "loss": 3.7656,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315156355336182e-05,
+      "loss": 3.7623,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314319398465503e-05,
+      "loss": 3.7767,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313480803714451e-05,
+      "loss": 3.7724,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312642208963399e-05,
+      "loss": 3.7848,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311803614212347e-05,
+      "loss": 3.7622,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310965019461295e-05,
+      "loss": 3.7645,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310126424710243e-05,
+      "loss": 3.7621,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309287829959191e-05,
+      "loss": 3.7538,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7717,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30761227833746e-05,
+      "loss": 3.7613,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306773683586408e-05,
+      "loss": 3.7683,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305935088835356e-05,
+      "loss": 3.7622,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7511,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042595372136254e-05,
+      "loss": 3.759,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034209424625734e-05,
+      "loss": 3.7522,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025823477115214e-05,
+      "loss": 3.7689,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.7613,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3009067960897903e-05,
+      "loss": 3.7453,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000682013387383e-05,
+      "loss": 3.7676,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299229606587686e-05,
+      "loss": 3.7575,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.7497,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297554054965955e-05,
+      "loss": 3.7586,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296715460214903e-05,
+      "loss": 3.7669,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295876865463851e-05,
+      "loss": 3.7619,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7655,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7447,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293361081210695e-05,
+      "loss": 3.7606,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292524124340016e-05,
+      "loss": 3.7533,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7657,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.7667,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29000834008686e-05,
+      "loss": 3.7565,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289169745335808e-05,
+      "loss": 3.7556,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7639,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7421,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7513,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285817004211973e-05,
+      "loss": 3.7447,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849800473412946e-05,
+      "loss": 3.7573,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841414525902426e-05,
+      "loss": 3.7444,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833028578391906e-05,
+      "loss": 3.7509,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824642630881386e-05,
+      "loss": 3.7477,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816273062174595e-05,
+      "loss": 3.7499,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.753,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7463,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7605,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.758,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774359703425724e-05,
+      "loss": 3.7582,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.7477,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757587808404684e-05,
+      "loss": 3.7588,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749201860894164e-05,
+      "loss": 3.7587,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274083229218738e-05,
+      "loss": 3.7431,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273244634467686e-05,
+      "loss": 3.7658,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272406039716634e-05,
+      "loss": 3.7426,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271567444965582e-05,
+      "loss": 3.7582,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27072885021453e-05,
+      "loss": 3.7412,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269890255463478e-05,
+      "loss": 3.7582,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269051660712426e-05,
+      "loss": 3.7395,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268214703841747e-05,
+      "loss": 3.744,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267376109090695e-05,
+      "loss": 3.7414,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266537514339643e-05,
+      "loss": 3.764,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265698919588591e-05,
+      "loss": 3.747,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264861962717912e-05,
+      "loss": 3.7553,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2640233679668604e-05,
+      "loss": 3.7422,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631847732158084e-05,
+      "loss": 3.7578,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623461784647564e-05,
+      "loss": 3.7596,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2615075837137044e-05,
+      "loss": 3.7568,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260670626843025e-05,
+      "loss": 3.7531,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259832032091973e-05,
+      "loss": 3.7504,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258993437340921e-05,
+      "loss": 3.7504,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258154842589869e-05,
+      "loss": 3.761,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25731788571919e-05,
+      "loss": 3.7452,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256479290968138e-05,
+      "loss": 3.7524,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255640696217086e-05,
+      "loss": 3.7465,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254802101466034e-05,
+      "loss": 3.7491,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253965144595356e-05,
+      "loss": 3.75,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253126549844304e-05,
+      "loss": 3.744,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252287955093252e-05,
+      "loss": 3.7569,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514493603422e-05,
+      "loss": 3.7349,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250612403471521e-05,
+      "loss": 3.7572,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.867486000061035,
+      "eval_runtime": 310.0624,
+      "eval_samples_per_second": 1230.691,
+      "eval_steps_per_second": 38.46,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249773808720469e-05,
+      "loss": 3.743,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248935213969417e-05,
+      "loss": 3.7332,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248096619218364e-05,
+      "loss": 3.75,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247258024467312e-05,
+      "loss": 3.7438,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24641942971626e-05,
+      "loss": 3.7576,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455824728455816e-05,
+      "loss": 3.7422,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447438780945296e-05,
+      "loss": 3.7357,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2439052833434776e-05,
+      "loss": 3.7334,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430666885924255e-05,
+      "loss": 3.7415,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242229731721747e-05,
+      "loss": 3.7493,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413911369706945e-05,
+      "loss": 3.7385,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405525422196424e-05,
+      "loss": 3.7429,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397139474685904e-05,
+      "loss": 3.7406,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388769905979114e-05,
+      "loss": 3.7342,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380383958468593e-05,
+      "loss": 3.7406,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371998010958073e-05,
+      "loss": 3.7285,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363612063447553e-05,
+      "loss": 3.7339,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235522611593703e-05,
+      "loss": 3.7381,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234684016842651e-05,
+      "loss": 3.7356,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233845422091599e-05,
+      "loss": 3.745,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233006827340547e-05,
+      "loss": 3.745,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232171508350242e-05,
+      "loss": 3.7392,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23133291359919e-05,
+      "loss": 3.7382,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230494318848138e-05,
+      "loss": 3.7457,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229655724097086e-05,
+      "loss": 3.7228,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228818767226407e-05,
+      "loss": 3.749,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227980172475355e-05,
+      "loss": 3.7337,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227141577724303e-05,
+      "loss": 3.731,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226302982973251e-05,
+      "loss": 3.7263,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254660261025716e-05,
+      "loss": 3.7333,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2246274313515196e-05,
+      "loss": 3.7374,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223788836600468e-05,
+      "loss": 3.7399,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222950241849416e-05,
+      "loss": 3.7404,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222111647098364e-05,
+      "loss": 3.7435,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221273052347312e-05,
+      "loss": 3.7318,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220436095476633e-05,
+      "loss": 3.7411,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219597500725581e-05,
+      "loss": 3.7387,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218758905974529e-05,
+      "loss": 3.7293,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217920311223477e-05,
+      "loss": 3.7272,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217081716472425e-05,
+      "loss": 3.7315,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216244759601746e-05,
+      "loss": 3.7295,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215406164850694e-05,
+      "loss": 3.7402,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214567570099642e-05,
+      "loss": 3.7305,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21372897534859e-05,
+      "loss": 3.7303,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128920184779116e-05,
+      "loss": 3.743,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120534237268596e-05,
+      "loss": 3.7312,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112148289758076e-05,
+      "loss": 3.7358,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103762342247556e-05,
+      "loss": 3.7293,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095392773540765e-05,
+      "loss": 3.6986,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087006826030245e-05,
+      "loss": 3.7421,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078620878519725e-05,
+      "loss": 3.7288,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2070234931009205e-05,
+      "loss": 3.7312,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061865362302414e-05,
+      "loss": 3.7197,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2053479414791894e-05,
+      "loss": 3.7115,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2045093467281374e-05,
+      "loss": 3.7166,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036707519770854e-05,
+      "loss": 3.7272,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2028321572260334e-05,
+      "loss": 3.7226,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201993562474982e-05,
+      "loss": 3.7229,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201156605604303e-05,
+      "loss": 3.7357,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200318010853251e-05,
+      "loss": 3.7199,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199479416102199e-05,
+      "loss": 3.7187,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198640821351147e-05,
+      "loss": 3.7317,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197802226600095e-05,
+      "loss": 3.7058,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196965269729416e-05,
+      "loss": 3.7204,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196126674978364e-05,
+      "loss": 3.7245,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195288080227312e-05,
+      "loss": 3.7323,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19444948547626e-05,
+      "loss": 3.7065,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193612528605581e-05,
+      "loss": 3.7217,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192773933854529e-05,
+      "loss": 3.7095,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919353391034775e-05,
+      "loss": 3.7317,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191096744352425e-05,
+      "loss": 3.7296,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902597874817464e-05,
+      "loss": 3.7265,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1894211927306944e-05,
+      "loss": 3.7322,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885825979796424e-05,
+      "loss": 3.7364,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18774400322859e-05,
+      "loss": 3.747,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186907046357911e-05,
+      "loss": 3.7233,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186068451606859e-05,
+      "loss": 3.7267,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185229856855807e-05,
+      "loss": 3.7244,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843912621047546e-05,
+      "loss": 3.7143,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183554305234076e-05,
+      "loss": 3.7289,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182715710483024e-05,
+      "loss": 3.7255,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181877115731972e-05,
+      "loss": 3.7284,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18103852098092e-05,
+      "loss": 3.732,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180201564110242e-05,
+      "loss": 3.713,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17936296935919e-05,
+      "loss": 3.715,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178524374608137e-05,
+      "loss": 3.7183,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177685779857085e-05,
+      "loss": 3.7337,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176847185106033e-05,
+      "loss": 3.7215,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760102282353546e-05,
+      "loss": 3.7125,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175171633484302e-05,
+      "loss": 3.7177,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17433303873325e-05,
+      "loss": 3.727,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173494443982198e-05,
+      "loss": 3.7131,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726574871115195e-05,
+      "loss": 3.72,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1718188923604675e-05,
+      "loss": 3.733,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709802976094155e-05,
+      "loss": 3.7238,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1701417028583635e-05,
+      "loss": 3.726,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1693031081073115e-05,
+      "loss": 3.7084,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684661512366324e-05,
+      "loss": 3.7224,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1676275564855804e-05,
+      "loss": 3.7158,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667889617345284e-05,
+      "loss": 3.7253,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1659503669834764e-05,
+      "loss": 3.7288,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651117722324244e-05,
+      "loss": 3.726,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164274815361745e-05,
+      "loss": 3.7179,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163436220610693e-05,
+      "loss": 3.7271,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162597625859642e-05,
+      "loss": 3.7067,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16175903110859e-05,
+      "loss": 3.7169,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160922074237911e-05,
+      "loss": 3.7043,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160083479486859e-05,
+      "loss": 3.7244,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159244884735807e-05,
+      "loss": 3.7081,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158406289984755e-05,
+      "loss": 3.7121,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157567695233703e-05,
+      "loss": 3.7095,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156730738363024e-05,
+      "loss": 3.7126,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155892143611972e-05,
+      "loss": 3.7173,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15505354886092e-05,
+      "loss": 3.7102,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154214954109868e-05,
+      "loss": 3.7214,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153377997239189e-05,
+      "loss": 3.7275,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152539402488137e-05,
+      "loss": 3.7157,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151700807737085e-05,
+      "loss": 3.7192,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150862212986033e-05,
+      "loss": 3.7186,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150025256115354e-05,
+      "loss": 3.722,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149186661364302e-05,
+      "loss": 3.7137,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14834806661325e-05,
+      "loss": 3.7225,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147509471862198e-05,
+      "loss": 3.7064,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146670877111146e-05,
+      "loss": 3.7219,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145833920240467e-05,
+      "loss": 3.709,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144995325489415e-05,
+      "loss": 3.7159,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144156730738363e-05,
+      "loss": 3.7099,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143318135987311e-05,
+      "loss": 3.7091,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142481179116632e-05,
+      "loss": 3.7053,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141642584365581e-05,
+      "loss": 3.7225,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140803989614529e-05,
+      "loss": 3.7137,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139965394863477e-05,
+      "loss": 3.7202,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1391284379927976e-05,
+      "loss": 3.7028,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382898432417456e-05,
+      "loss": 3.7226,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374512484906936e-05,
+      "loss": 3.7211,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1366126537396416e-05,
+      "loss": 3.7207,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357756968689625e-05,
+      "loss": 3.7165,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349371021179105e-05,
+      "loss": 3.714,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340985073668585e-05,
+      "loss": 3.716,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332599126158065e-05,
+      "loss": 3.729,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1324213178647545e-05,
+      "loss": 3.7068,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131584360994076e-05,
+      "loss": 3.7151,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130745766243024e-05,
+      "loss": 3.7157,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129907171491972e-05,
+      "loss": 3.7149,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12906857674092e-05,
+      "loss": 3.7057,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128231619870241e-05,
+      "loss": 3.7083,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127393025119189e-05,
+      "loss": 3.7238,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126554430368137e-05,
+      "loss": 3.7017,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125715835617085e-05,
+      "loss": 3.7211,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.857250690460205,
+      "eval_runtime": 305.8615,
+      "eval_samples_per_second": 1247.594,
+      "eval_steps_per_second": 38.988,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124877240866032e-05,
+      "loss": 3.7086,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124040283995354e-05,
+      "loss": 3.6963,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123201689244302e-05,
+      "loss": 3.7163,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12236309449325e-05,
+      "loss": 3.7035,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121524499742198e-05,
+      "loss": 3.7194,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206875428715194e-05,
+      "loss": 3.7094,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1198489481204674e-05,
+      "loss": 3.7045,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1190103533694154e-05,
+      "loss": 3.7007,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118171758618363e-05,
+      "loss": 3.7079,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117334801747684e-05,
+      "loss": 3.7127,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116496206996632e-05,
+      "loss": 3.7027,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156576122455796e-05,
+      "loss": 3.7056,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148190174945276e-05,
+      "loss": 3.7027,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113982060623849e-05,
+      "loss": 3.7016,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113143465872797e-05,
+      "loss": 3.7065,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112304871121745e-05,
+      "loss": 3.6972,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111466276370693e-05,
+      "loss": 3.697,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110627681619641e-05,
+      "loss": 3.7039,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109790724748963e-05,
+      "loss": 3.7026,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10895212999791e-05,
+      "loss": 3.7082,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108113535246858e-05,
+      "loss": 3.7127,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107274940495806e-05,
+      "loss": 3.7077,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106437983625127e-05,
+      "loss": 3.7051,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105599388874075e-05,
+      "loss": 3.7065,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104760794123023e-05,
+      "loss": 3.6936,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1039238372523446e-05,
+      "loss": 3.7139,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103085242501292e-05,
+      "loss": 3.6976,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1022466477502406e-05,
+      "loss": 3.702,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1014080529991886e-05,
+      "loss": 3.6956,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005694582481366e-05,
+      "loss": 3.698,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0997325013774575e-05,
+      "loss": 3.7018,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988939066264055e-05,
+      "loss": 3.7076,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0980553118753535e-05,
+      "loss": 3.7049,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0972167171243015e-05,
+      "loss": 3.7073,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963781223732495e-05,
+      "loss": 3.6967,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955411655025704e-05,
+      "loss": 3.71,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0947025707515184e-05,
+      "loss": 3.7078,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938639760004664e-05,
+      "loss": 3.6933,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0930253812494144e-05,
+      "loss": 3.6976,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092188424378736e-05,
+      "loss": 3.697,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091349829627684e-05,
+      "loss": 3.698,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090511234876632e-05,
+      "loss": 3.7052,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08967264012558e-05,
+      "loss": 3.6982,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088835683254901e-05,
+      "loss": 3.6975,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087997088503849e-05,
+      "loss": 3.708,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087158493752797e-05,
+      "loss": 3.6991,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086319899001745e-05,
+      "loss": 3.7004,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085481304250693e-05,
+      "loss": 3.6986,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084644347380014e-05,
+      "loss": 3.6655,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083805752628962e-05,
+      "loss": 3.7101,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08296715787791e-05,
+      "loss": 3.6945,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082128563126858e-05,
+      "loss": 3.7026,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081291606256179e-05,
+      "loss": 3.6829,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080453011505127e-05,
+      "loss": 3.6817,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079614416754075e-05,
+      "loss": 3.6796,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078775822003023e-05,
+      "loss": 3.6926,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077938865132344e-05,
+      "loss": 3.6938,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077100270381292e-05,
+      "loss": 3.6865,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07626167563024e-05,
+      "loss": 3.7073,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075423080879188e-05,
+      "loss": 3.6876,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074586124008509e-05,
+      "loss": 3.6878,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073747529257457e-05,
+      "loss": 3.6991,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072908934506405e-05,
+      "loss": 3.6735,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072070339755353e-05,
+      "loss": 3.6853,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071233382884675e-05,
+      "loss": 3.6922,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070394788133623e-05,
+      "loss": 3.7004,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069556193382571e-05,
+      "loss": 3.6746,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068717598631519e-05,
+      "loss": 3.689,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678806417608396e-05,
+      "loss": 3.6764,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0670420470097876e-05,
+      "loss": 3.7007,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0662034522587356e-05,
+      "loss": 3.6932,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653648575076836e-05,
+      "loss": 3.695,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645279006370045e-05,
+      "loss": 3.6984,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636893058859525e-05,
+      "loss": 3.7035,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0628507111349005e-05,
+      "loss": 3.714,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0620121163838485e-05,
+      "loss": 3.6928,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611735216327965e-05,
+      "loss": 3.6946,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060336564762118e-05,
+      "loss": 3.6968,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059497970011066e-05,
+      "loss": 3.6798,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058659375260014e-05,
+      "loss": 3.6944,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057820780508962e-05,
+      "loss": 3.6971,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056983823638283e-05,
+      "loss": 3.6974,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056145228887231e-05,
+      "loss": 3.6976,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055306634136179e-05,
+      "loss": 3.6843,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054468039385127e-05,
+      "loss": 3.6796,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053631082514448e-05,
+      "loss": 3.6859,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052792487763396e-05,
+      "loss": 3.703,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051953893012344e-05,
+      "loss": 3.6895,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051115298261292e-05,
+      "loss": 3.6808,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502783413906134e-05,
+      "loss": 3.687,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0494397466395614e-05,
+      "loss": 3.6953,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0486011518885094e-05,
+      "loss": 3.6847,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477625571374574e-05,
+      "loss": 3.6876,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046925600266778e-05,
+      "loss": 3.7,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046087005515726e-05,
+      "loss": 3.6918,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045248410764674e-05,
+      "loss": 3.6925,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044409816013622e-05,
+      "loss": 3.6822,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043572859142943e-05,
+      "loss": 3.6872,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042734264391891e-05,
+      "loss": 3.6837,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041895669640839e-05,
+      "loss": 3.6971,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041057074889787e-05,
+      "loss": 3.6927,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040218480138735e-05,
+      "loss": 3.6963,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039381523268057e-05,
+      "loss": 3.689,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038542928517005e-05,
+      "loss": 3.6924,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037704333765953e-05,
+      "loss": 3.6798,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368657390149e-05,
+      "loss": 3.681,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036028782144222e-05,
+      "loss": 3.677,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03519018739317e-05,
+      "loss": 3.6934,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0343515926421177e-05,
+      "loss": 3.6787,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033512997891065e-05,
+      "loss": 3.6777,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326760410203866e-05,
+      "loss": 3.6762,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0318374462693346e-05,
+      "loss": 3.6858,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309988515182826e-05,
+      "loss": 3.6858,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0301602567672305e-05,
+      "loss": 3.6761,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029323299896552e-05,
+      "loss": 3.6954,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284847051455e-05,
+      "loss": 3.6914,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0276461103944475e-05,
+      "loss": 3.6846,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0268075156433954e-05,
+      "loss": 3.6901,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025970558772717e-05,
+      "loss": 3.6884,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025131964021665e-05,
+      "loss": 3.6862,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242933692706123e-05,
+      "loss": 3.6872,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0234547745195603e-05,
+      "loss": 3.689,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022617817648882e-05,
+      "loss": 3.678,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02177922289783e-05,
+      "loss": 3.6901,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020940628146778e-05,
+      "loss": 3.682,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020102033395726e-05,
+      "loss": 3.6855,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019263438644674e-05,
+      "loss": 3.6767,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018426481773995e-05,
+      "loss": 3.6761,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017587887022943e-05,
+      "loss": 3.6775,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016749292271891e-05,
+      "loss": 3.6896,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015910697520839e-05,
+      "loss": 3.6828,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01507374065016e-05,
+      "loss": 3.6916,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014235145899108e-05,
+      "loss": 3.6705,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013396551148056e-05,
+      "loss": 3.697,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012557956397004e-05,
+      "loss": 3.6863,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011720999526325e-05,
+      "loss": 3.6914,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010882404775273e-05,
+      "loss": 3.6876,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010043810024221e-05,
+      "loss": 3.682,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009205215273169e-05,
+      "loss": 3.6854,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00836825840249e-05,
+      "loss": 3.6956,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007529663651438e-05,
+      "loss": 3.6774,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006691068900386e-05,
+      "loss": 3.6872,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005852474149334e-05,
+      "loss": 3.6854,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005015517278655e-05,
+      "loss": 3.6802,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004176922527603e-05,
+      "loss": 3.6774,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003338327776551e-05,
+      "loss": 3.6808,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002499733025499e-05,
+      "loss": 3.6903,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001661138274447e-05,
+      "loss": 3.6757,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0008241814037687e-05,
+      "loss": 3.6912,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8499464988708496,
+      "eval_runtime": 303.8172,
+      "eval_samples_per_second": 1255.989,
+      "eval_steps_per_second": 39.251,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9999855866527166e-05,
+      "loss": 3.6754,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9991469919016646e-05,
+      "loss": 3.671,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9983083971506126e-05,
+      "loss": 3.6799,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9974698023995606e-05,
+      "loss": 3.679,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9966328455288815e-05,
+      "loss": 3.6891,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957942507778295e-05,
+      "loss": 3.6785,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9949556560267775e-05,
+      "loss": 3.6763,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9941170612757255e-05,
+      "loss": 3.672,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932801044050464e-05,
+      "loss": 3.6732,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9924415096539944e-05,
+      "loss": 3.6851,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9916029149029424e-05,
+      "loss": 3.6727,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9907643201518904e-05,
+      "loss": 3.6731,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9899257254008384e-05,
+      "loss": 3.6827,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98908876853016e-05,
+      "loss": 3.6646,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988250173779108e-05,
+      "loss": 3.6749,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987411579028056e-05,
+      "loss": 3.6702,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986572984277004e-05,
+      "loss": 3.6673,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985736027406325e-05,
+      "loss": 3.6774,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984897432655273e-05,
+      "loss": 3.6739,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984058837904221e-05,
+      "loss": 3.6761,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983220243153169e-05,
+      "loss": 3.6848,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98238328628249e-05,
+      "loss": 3.679,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981544691531438e-05,
+      "loss": 3.6787,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980706096780386e-05,
+      "loss": 3.6733,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979867502029334e-05,
+      "loss": 3.6699,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9790305451586554e-05,
+      "loss": 3.6813,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781919504076034e-05,
+      "loss": 3.67,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9773533556565514e-05,
+      "loss": 3.6754,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9765147609054994e-05,
+      "loss": 3.661,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97567780403482e-05,
+      "loss": 3.6712,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974839209283768e-05,
+      "loss": 3.6692,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974000614532716e-05,
+      "loss": 3.677,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973162019781664e-05,
+      "loss": 3.6804,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972325062910985e-05,
+      "loss": 3.6803,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971486468159933e-05,
+      "loss": 3.6696,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970647873408881e-05,
+      "loss": 3.6748,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969809278657829e-05,
+      "loss": 3.6805,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968972321787151e-05,
+      "loss": 3.6687,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968133727036099e-05,
+      "loss": 3.6678,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967295132285047e-05,
+      "loss": 3.6666,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966456537533995e-05,
+      "loss": 3.6712,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965617942782943e-05,
+      "loss": 3.6701,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647809859122636e-05,
+      "loss": 3.676,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9639423911612116e-05,
+      "loss": 3.6704,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9631037964101596e-05,
+      "loss": 3.676,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9622652016591076e-05,
+      "loss": 3.6732,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9614282447884285e-05,
+      "loss": 3.6705,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605896500373765e-05,
+      "loss": 3.6691,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9597510552863245e-05,
+      "loss": 3.635,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958912460535273e-05,
+      "loss": 3.6812,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958075503664594e-05,
+      "loss": 3.6642,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957236908913542e-05,
+      "loss": 3.6749,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95639831416249e-05,
+      "loss": 3.6597,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9555597194114374e-05,
+      "loss": 3.653,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954722762540759e-05,
+      "loss": 3.6492,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953884167789707e-05,
+      "loss": 3.6617,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953045573038655e-05,
+      "loss": 3.6664,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952206978287602e-05,
+      "loss": 3.6606,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951370021416924e-05,
+      "loss": 3.6736,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950531426665872e-05,
+      "loss": 3.6625,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94969283191482e-05,
+      "loss": 3.6578,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948854237163768e-05,
+      "loss": 3.6696,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948015642412716e-05,
+      "loss": 3.6492,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9471786855420375e-05,
+      "loss": 3.6545,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946340090790985e-05,
+      "loss": 3.6666,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945501496039933e-05,
+      "loss": 3.6699,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944662901288881e-05,
+      "loss": 3.6505,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9438259444182024e-05,
+      "loss": 3.6581,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94298734966715e-05,
+      "loss": 3.6486,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942148754916098e-05,
+      "loss": 3.6707,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941310160165046e-05,
+      "loss": 3.6685,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940473203294367e-05,
+      "loss": 3.6647,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939634608543315e-05,
+      "loss": 3.6716,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938796013792263e-05,
+      "loss": 3.673,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937957419041211e-05,
+      "loss": 3.6876,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937120462170532e-05,
+      "loss": 3.6669,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93628186741948e-05,
+      "loss": 3.6679,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935443272668428e-05,
+      "loss": 3.664,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934604677917376e-05,
+      "loss": 3.6587,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933767721046697e-05,
+      "loss": 3.6642,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932929126295645e-05,
+      "loss": 3.6689,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932090531544593e-05,
+      "loss": 3.669,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931251936793541e-05,
+      "loss": 3.6728,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304149799228626e-05,
+      "loss": 3.6537,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295763851718106e-05,
+      "loss": 3.6538,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287377904207586e-05,
+      "loss": 3.6563,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278991956697066e-05,
+      "loss": 3.6751,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9270622387990275e-05,
+      "loss": 3.6656,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9262236440479755e-05,
+      "loss": 3.6512,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253850492969235e-05,
+      "loss": 3.6611,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9245464545458715e-05,
+      "loss": 3.6656,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9237094976751924e-05,
+      "loss": 3.6586,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9228709029241404e-05,
+      "loss": 3.6567,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9220323081730884e-05,
+      "loss": 3.6733,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211937134220364e-05,
+      "loss": 3.6652,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920356756551358e-05,
+      "loss": 3.6658,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919518161800306e-05,
+      "loss": 3.6549,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918679567049254e-05,
+      "loss": 3.6584,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917840972298202e-05,
+      "loss": 3.6587,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917004015427523e-05,
+      "loss": 3.6691,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916165420676471e-05,
+      "loss": 3.6655,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915326825925419e-05,
+      "loss": 3.6677,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914488231174367e-05,
+      "loss": 3.6658,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913651274303688e-05,
+      "loss": 3.6603,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912812679552636e-05,
+      "loss": 3.6598,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911974084801584e-05,
+      "loss": 3.6505,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911135490050532e-05,
+      "loss": 3.6468,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91029689529948e-05,
+      "loss": 3.6682,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9094599384288014e-05,
+      "loss": 3.6552,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9086213436777494e-05,
+      "loss": 3.648,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077827489266973e-05,
+      "loss": 3.6501,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9069441541756453e-05,
+      "loss": 3.6509,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906107197304966e-05,
+      "loss": 3.6628,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905268602553914e-05,
+      "loss": 3.6531,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904430007802862e-05,
+      "loss": 3.6648,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903593050932183e-05,
+      "loss": 3.6651,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902754456181131e-05,
+      "loss": 3.6612,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901915861430079e-05,
+      "loss": 3.6623,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901077266679027e-05,
+      "loss": 3.6596,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900240309808349e-05,
+      "loss": 3.66,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899401715057297e-05,
+      "loss": 3.6602,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898563120306245e-05,
+      "loss": 3.6613,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897724525555193e-05,
+      "loss": 3.6551,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8968875686845136e-05,
+      "loss": 3.6634,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8960489739334616e-05,
+      "loss": 3.6565,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8952103791824096e-05,
+      "loss": 3.6561,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8943717844313576e-05,
+      "loss": 3.651,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935348275606785e-05,
+      "loss": 3.6504,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926962328096265e-05,
+      "loss": 3.6496,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918576380585745e-05,
+      "loss": 3.6632,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8910190433075225e-05,
+      "loss": 3.6543,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890182086436844e-05,
+      "loss": 3.6696,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889343491685792e-05,
+      "loss": 3.6444,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88850489693474e-05,
+      "loss": 3.6663,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887666302183688e-05,
+      "loss": 3.6597,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886829345313009e-05,
+      "loss": 3.6649,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885990750561957e-05,
+      "loss": 3.6619,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885152155810905e-05,
+      "loss": 3.6554,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884313561059853e-05,
+      "loss": 3.6596,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883476604189174e-05,
+      "loss": 3.6708,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882638009438122e-05,
+      "loss": 3.652,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88179941468707e-05,
+      "loss": 3.6617,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880960819936018e-05,
+      "loss": 3.6578,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8801238630653395e-05,
+      "loss": 3.6538,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8792852683142875e-05,
+      "loss": 3.6534,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8784466735632355e-05,
+      "loss": 3.6533,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8776080788121834e-05,
+      "loss": 3.6627,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8767694840611314e-05,
+      "loss": 3.6534,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8759325271904524e-05,
+      "loss": 3.6594,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8459384441375732,
+      "eval_runtime": 304.3449,
+      "eval_samples_per_second": 1253.811,
+      "eval_steps_per_second": 39.183,
+      "step": 686880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8750939324394004e-05,
+      "loss": 3.6442,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8742553376883483e-05,
+      "loss": 3.6462,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8734167429372963e-05,
+      "loss": 3.6517,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872578148186244e-05,
+      "loss": 3.6532,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871741191315565e-05,
+      "loss": 3.6659,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870902596564513e-05,
+      "loss": 3.653,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870064001813461e-05,
+      "loss": 3.6512,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869225407062409e-05,
+      "loss": 3.6478,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.868388450191731e-05,
+      "loss": 3.6463,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.867549855440679e-05,
+      "loss": 3.6601,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.866711260689627e-05,
+      "loss": 3.6466,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865872665938575e-05,
+      "loss": 3.6476,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.865035709067896e-05,
+      "loss": 3.6567,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.864197114316844e-05,
+      "loss": 3.6406,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.863358519565792e-05,
+      "loss": 3.648,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.86251992481474e-05,
+      "loss": 3.6411,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.861681330063687e-05,
+      "loss": 3.6445,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8608443731930086e-05,
+      "loss": 3.6425,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8600057784419566e-05,
+      "loss": 3.6506,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8591671836909046e-05,
+      "loss": 3.6512,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858330226820226e-05,
+      "loss": 3.6592,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857491632069174e-05,
+      "loss": 3.6518,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856653037318122e-05,
+      "loss": 3.654,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8558144425670695e-05,
+      "loss": 3.6464,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8549758478160175e-05,
+      "loss": 3.6482,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854138890945339e-05,
+      "loss": 3.6558,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853300296194287e-05,
+      "loss": 3.6434,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8524617014432344e-05,
+      "loss": 3.6518,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8516231066921824e-05,
+      "loss": 3.6362,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850786149821504e-05,
+      "loss": 3.6444,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.849947555070452e-05,
+      "loss": 3.6437,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8491089603194e-05,
+      "loss": 3.6492,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.848270365568348e-05,
+      "loss": 3.6556,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8474334086976695e-05,
+      "loss": 3.6551,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846594813946617e-05,
+      "loss": 3.6448,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.845756219195565e-05,
+      "loss": 3.6447,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844917624444513e-05,
+      "loss": 3.6597,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8440806675738344e-05,
+      "loss": 3.644,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843242072822782e-05,
+      "loss": 3.6418,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84240347807173e-05,
+      "loss": 3.6411,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.841564883320678e-05,
+      "loss": 3.6481,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8407279264499993e-05,
+      "loss": 3.6445,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839889331698947e-05,
+      "loss": 3.6478,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839050736947895e-05,
+      "loss": 3.647,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838212142196843e-05,
+      "loss": 3.6482,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837375185326164e-05,
+      "loss": 3.6494,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836536590575112e-05,
+      "loss": 3.6457,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83569799582406e-05,
+      "loss": 3.6478,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834859401073008e-05,
+      "loss": 3.611,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.834022444202329e-05,
+      "loss": 3.6535,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833183849451277e-05,
+      "loss": 3.6386,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832345254700225e-05,
+      "loss": 3.6539,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831506659949173e-05,
+      "loss": 3.6333,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830669703078495e-05,
+      "loss": 3.6287,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829831108327443e-05,
+      "loss": 3.6241,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828992513576391e-05,
+      "loss": 3.6359,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828153918825339e-05,
+      "loss": 3.6482,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8273169619546596e-05,
+      "loss": 3.6299,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8264783672036076e-05,
+      "loss": 3.6511,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8256397724525556e-05,
+      "loss": 3.6411,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8248011777015036e-05,
+      "loss": 3.6272,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8239642208308245e-05,
+      "loss": 3.6454,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8231256260797725e-05,
+      "loss": 3.6295,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8222870313287205e-05,
+      "loss": 3.6276,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8214484365776685e-05,
+      "loss": 3.6406,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82061147970699e-05,
+      "loss": 3.6476,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.819772884955938e-05,
+      "loss": 3.6278,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818934290204886e-05,
+      "loss": 3.6327,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.818095695453834e-05,
+      "loss": 3.6266,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817257100702782e-05,
+      "loss": 3.6444,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.816420143832103e-05,
+      "loss": 3.6437,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.815581549081051e-05,
+      "loss": 3.6415,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.814742954329999e-05,
+      "loss": 3.646,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.81390599745932e-05,
+      "loss": 3.6489,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.813067402708268e-05,
+      "loss": 3.6601,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812228807957216e-05,
+      "loss": 3.6427,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.811390213206164e-05,
+      "loss": 3.6456,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.810551618455112e-05,
+      "loss": 3.6368,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8097146615844334e-05,
+      "loss": 3.6365,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8088760668333814e-05,
+      "loss": 3.6382,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8080374720823294e-05,
+      "loss": 3.6462,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8071988773312774e-05,
+      "loss": 3.6422,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.806361920460598e-05,
+      "loss": 3.6517,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805523325709546e-05,
+      "loss": 3.6276,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804684730958494e-05,
+      "loss": 3.6288,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803846136207442e-05,
+      "loss": 3.6364,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803009179336763e-05,
+      "loss": 3.6462,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802170584585711e-05,
+      "loss": 3.6418,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801331989834659e-05,
+      "loss": 3.63,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800493395083607e-05,
+      "loss": 3.6357,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799656438212929e-05,
+      "loss": 3.641,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798817843461877e-05,
+      "loss": 3.6318,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797979248710825e-05,
+      "loss": 3.6309,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797140653959773e-05,
+      "loss": 3.6494,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796303697089094e-05,
+      "loss": 3.641,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795465102338042e-05,
+      "loss": 3.6426,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79462650758699e-05,
+      "loss": 3.6317,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793787912835938e-05,
+      "loss": 3.6353,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7929509559652586e-05,
+      "loss": 3.6315,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7921123612142066e-05,
+      "loss": 3.6438,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7912737664631546e-05,
+      "loss": 3.6449,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7904351717121026e-05,
+      "loss": 3.6393,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789598214841424e-05,
+      "loss": 3.6421,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788759620090372e-05,
+      "loss": 3.6369,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78792102533932e-05,
+      "loss": 3.6378,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787082430588268e-05,
+      "loss": 3.6258,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786245473717589e-05,
+      "loss": 3.6255,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.785406878966537e-05,
+      "loss": 3.6365,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784568284215485e-05,
+      "loss": 3.6374,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783729689464433e-05,
+      "loss": 3.6196,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782892732593754e-05,
+      "loss": 3.6281,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782054137842702e-05,
+      "loss": 3.6273,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78121554309165e-05,
+      "loss": 3.6385,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780376948340598e-05,
+      "loss": 3.6273,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7795399914699195e-05,
+      "loss": 3.6426,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7787013967188675e-05,
+      "loss": 3.6422,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7778628019678155e-05,
+      "loss": 3.6399,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7770242072167635e-05,
+      "loss": 3.6332,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7761872503460844e-05,
+      "loss": 3.6418,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7753486555950324e-05,
+      "loss": 3.6357,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7745100608439804e-05,
+      "loss": 3.6378,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7736714660929284e-05,
+      "loss": 3.6371,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772834509222249e-05,
+      "loss": 3.6291,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771995914471197e-05,
+      "loss": 3.6384,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.771157319720145e-05,
+      "loss": 3.636,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.770318724969093e-05,
+      "loss": 3.6284,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.769481768098415e-05,
+      "loss": 3.6306,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768643173347363e-05,
+      "loss": 3.628,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.767804578596311e-05,
+      "loss": 3.6216,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.766965983845259e-05,
+      "loss": 3.6405,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76612902697458e-05,
+      "loss": 3.6313,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.765290432223528e-05,
+      "loss": 3.6458,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.764451837472476e-05,
+      "loss": 3.6215,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.763613242721424e-05,
+      "loss": 3.642,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.762776285850745e-05,
+      "loss": 3.6384,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761937691099693e-05,
+      "loss": 3.6425,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761099096348641e-05,
+      "loss": 3.6365,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760260501597589e-05,
+      "loss": 3.6295,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75942354472691e-05,
+      "loss": 3.6414,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758584949975858e-05,
+      "loss": 3.6429,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757746355224806e-05,
+      "loss": 3.6309,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756907760473754e-05,
+      "loss": 3.6366,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.756070803603075e-05,
+      "loss": 3.6355,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755232208852023e-05,
+      "loss": 3.6292,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754393614100971e-05,
+      "loss": 3.6294,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753555019349919e-05,
+      "loss": 3.6298,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75271806247924e-05,
+      "loss": 3.6371,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751879467728188e-05,
+      "loss": 3.6312,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751040872977136e-05,
+      "loss": 3.6379,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8439977169036865,
+      "eval_runtime": 307.6583,
+      "eval_samples_per_second": 1240.308,
+      "eval_steps_per_second": 38.761,
+      "step": 763200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750202278226084e-05,
+      "loss": 3.6242,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7493653213554056e-05,
+      "loss": 3.622,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7485267266043536e-05,
+      "loss": 3.6277,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7476881318533016e-05,
+      "loss": 3.6303,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7468495371022496e-05,
+      "loss": 3.6438,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7460125802315705e-05,
+      "loss": 3.6329,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7451739854805185e-05,
+      "loss": 3.6263,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7443353907294665e-05,
+      "loss": 3.6262,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.743496795978414e-05,
+      "loss": 3.6226,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7426598391077354e-05,
+      "loss": 3.6326,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7418212443566834e-05,
+      "loss": 3.6273,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7409826496056314e-05,
+      "loss": 3.6294,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7401440548545794e-05,
+      "loss": 3.634,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7393070979839e-05,
+      "loss": 3.614,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.738468503232849e-05,
+      "loss": 3.6268,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.737629908481796e-05,
+      "loss": 3.6152,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.736791313730744e-05,
+      "loss": 3.6275,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735954356860066e-05,
+      "loss": 3.6222,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.735115762109014e-05,
+      "loss": 3.6263,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734277167357961e-05,
+      "loss": 3.625,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733438572606909e-05,
+      "loss": 3.6359,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.732601615736231e-05,
+      "loss": 3.6258,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.731763020985179e-05,
+      "loss": 3.6336,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730924426234126e-05,
+      "loss": 3.6263,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.730085831483074e-05,
+      "loss": 3.6242,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729248874612396e-05,
+      "loss": 3.6319,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.728410279861344e-05,
+      "loss": 3.6192,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.727571685110292e-05,
+      "loss": 3.6276,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72673309035924e-05,
+      "loss": 3.6167,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.725896133488561e-05,
+      "loss": 3.6197,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7250575387375086e-05,
+      "loss": 3.6249,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7242189439864566e-05,
+      "loss": 3.628,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7233803492354046e-05,
+      "loss": 3.627,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722543392364726e-05,
+      "loss": 3.6387,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7217047976136735e-05,
+      "loss": 3.6214,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208662028626215e-05,
+      "loss": 3.6229,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7200276081115695e-05,
+      "loss": 3.6386,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719190651240891e-05,
+      "loss": 3.6213,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718352056489839e-05,
+      "loss": 3.6197,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717513461738787e-05,
+      "loss": 3.6222,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716674866987735e-05,
+      "loss": 3.6185,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715837910117056e-05,
+      "loss": 3.6238,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714999315366004e-05,
+      "loss": 3.6254,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714160720614952e-05,
+      "loss": 3.628,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7133221258639e-05,
+      "loss": 3.6209,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712485168993221e-05,
+      "loss": 3.6294,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.711646574242169e-05,
+      "loss": 3.6215,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.710807979491117e-05,
+      "loss": 3.6284,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709969384740065e-05,
+      "loss": 3.5918,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7091324278693864e-05,
+      "loss": 3.6322,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7082938331183344e-05,
+      "loss": 3.6124,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7074552383672824e-05,
+      "loss": 3.6289,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7066166436162304e-05,
+      "loss": 3.6136,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705779686745551e-05,
+      "loss": 3.6088,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704941091994499e-05,
+      "loss": 3.6033,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704102497243447e-05,
+      "loss": 3.6133,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703263902492395e-05,
+      "loss": 3.6252,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702426945621716e-05,
+      "loss": 3.6077,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701588350870664e-05,
+      "loss": 3.6293,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700749756119612e-05,
+      "loss": 3.6212,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69991116136856e-05,
+      "loss": 3.6011,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699074204497882e-05,
+      "loss": 3.6228,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69823560974683e-05,
+      "loss": 3.6075,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.697397014995778e-05,
+      "loss": 3.6063,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.696558420244726e-05,
+      "loss": 3.6192,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.695721463374047e-05,
+      "loss": 3.6227,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694882868622995e-05,
+      "loss": 3.603,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694044273871943e-05,
+      "loss": 3.6135,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.693205679120891e-05,
+      "loss": 3.6047,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692367084369839e-05,
+      "loss": 3.6176,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6915301274991596e-05,
+      "loss": 3.6235,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6906915327481076e-05,
+      "loss": 3.6189,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6898529379970556e-05,
+      "loss": 3.6257,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.689014343246004e-05,
+      "loss": 3.6244,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.688177386375325e-05,
+      "loss": 3.6396,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.687338791624273e-05,
+      "loss": 3.6219,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686500196873221e-05,
+      "loss": 3.6273,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685663240002542e-05,
+      "loss": 3.6113,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68482464525149e-05,
+      "loss": 3.6159,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683986050500438e-05,
+      "loss": 3.6175,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683147455749386e-05,
+      "loss": 3.6221,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682310498878707e-05,
+      "loss": 3.6262,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681471904127655e-05,
+      "loss": 3.6282,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.680633309376603e-05,
+      "loss": 3.6058,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679794714625551e-05,
+      "loss": 3.6107,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6789577577548725e-05,
+      "loss": 3.6121,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6781191630038205e-05,
+      "loss": 3.6255,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6772805682527685e-05,
+      "loss": 3.6184,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6764419735017165e-05,
+      "loss": 3.6104,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6756050166310374e-05,
+      "loss": 3.6174,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6747664218799854e-05,
+      "loss": 3.6167,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6739278271289334e-05,
+      "loss": 3.6124,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6730892323778814e-05,
+      "loss": 3.6102,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.672252275507202e-05,
+      "loss": 3.6261,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.67141368075615e-05,
+      "loss": 3.6236,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.670575086005098e-05,
+      "loss": 3.6193,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.669736491254046e-05,
+      "loss": 3.6062,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668899534383368e-05,
+      "loss": 3.6133,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.668060939632316e-05,
+      "loss": 3.6127,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667222344881264e-05,
+      "loss": 3.6195,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666383750130212e-05,
+      "loss": 3.6247,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.665546793259533e-05,
+      "loss": 3.6169,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664708198508481e-05,
+      "loss": 3.6196,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663869603757429e-05,
+      "loss": 3.6162,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663031009006377e-05,
+      "loss": 3.6162,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662194052135698e-05,
+      "loss": 3.6058,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661355457384646e-05,
+      "loss": 3.6049,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660516862633594e-05,
+      "loss": 3.6154,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659678267882542e-05,
+      "loss": 3.6147,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658841311011863e-05,
+      "loss": 3.6014,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658002716260811e-05,
+      "loss": 3.6088,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.657164121509759e-05,
+      "loss": 3.5995,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.656325526758707e-05,
+      "loss": 3.6164,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.655488569888028e-05,
+      "loss": 3.6091,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.654649975136976e-05,
+      "loss": 3.6197,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653811380385924e-05,
+      "loss": 3.6198,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652972785634872e-05,
+      "loss": 3.616,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.652135828764193e-05,
+      "loss": 3.6141,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.651297234013141e-05,
+      "loss": 3.624,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650458639262089e-05,
+      "loss": 3.6135,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649620044511037e-05,
+      "loss": 3.6196,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6487830876403586e-05,
+      "loss": 3.6135,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6479444928893066e-05,
+      "loss": 3.6113,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6471058981382546e-05,
+      "loss": 3.6158,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6462673033872026e-05,
+      "loss": 3.6148,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6454287086361506e-05,
+      "loss": 3.608,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6445917517654715e-05,
+      "loss": 3.6098,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6437531570144195e-05,
+      "loss": 3.603,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6429145622633675e-05,
+      "loss": 3.6037,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6420759675123155e-05,
+      "loss": 3.6197,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6412390106416364e-05,
+      "loss": 3.611,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6404004158905844e-05,
+      "loss": 3.6202,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6395618211395324e-05,
+      "loss": 3.6041,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6387232263884804e-05,
+      "loss": 3.6185,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637886269517802e-05,
+      "loss": 3.6161,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63704767476675e-05,
+      "loss": 3.6169,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636209080015698e-05,
+      "loss": 3.6157,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.635370485264646e-05,
+      "loss": 3.6091,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.634533528393967e-05,
+      "loss": 3.6224,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.633694933642915e-05,
+      "loss": 3.6229,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632856338891863e-05,
+      "loss": 3.6067,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.632017744140811e-05,
+      "loss": 3.6157,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.631180787270132e-05,
+      "loss": 3.6157,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.63034219251908e-05,
+      "loss": 3.6092,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.629503597768028e-05,
+      "loss": 3.607,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.628665003016976e-05,
+      "loss": 3.6122,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6278280461462974e-05,
+      "loss": 3.6146,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6269894513952454e-05,
+      "loss": 3.6161,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6261508566441933e-05,
+      "loss": 3.6134,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8427672386169434,
+      "eval_runtime": 306.8547,
+      "eval_samples_per_second": 1243.556,
+      "eval_steps_per_second": 38.862,
+      "step": 839520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625312261893141e-05,
+      "loss": 3.6029,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6244736671420887e-05,
+      "loss": 3.6001,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6236350723910367e-05,
+      "loss": 3.6074,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6227964776399847e-05,
+      "loss": 3.6092,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6219595207693056e-05,
+      "loss": 3.6203,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6211209260182536e-05,
+      "loss": 3.6109,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6202823312672016e-05,
+      "loss": 3.6099,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6194437365161495e-05,
+      "loss": 3.6075,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.618606779645471e-05,
+      "loss": 3.5986,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.617768184894419e-05,
+      "loss": 3.6062,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616929590143367e-05,
+      "loss": 3.612,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.616090995392315e-05,
+      "loss": 3.6077,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.615252400641263e-05,
+      "loss": 3.612,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.614415443770584e-05,
+      "loss": 3.596,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613576849019532e-05,
+      "loss": 3.6056,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.61273825426848e-05,
+      "loss": 3.5917,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611899659517428e-05,
+      "loss": 3.6084,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611062702646749e-05,
+      "loss": 3.6035,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610224107895697e-05,
+      "loss": 3.6047,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.609385513144645e-05,
+      "loss": 3.6043,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6085469183935936e-05,
+      "loss": 3.6199,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6077099615229145e-05,
+      "loss": 3.6044,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6068713667718625e-05,
+      "loss": 3.6134,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6060327720208105e-05,
+      "loss": 3.6027,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6051941772697585e-05,
+      "loss": 3.6046,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6043572203990794e-05,
+      "loss": 3.6125,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6035186256480274e-05,
+      "loss": 3.5977,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6026800308969754e-05,
+      "loss": 3.6073,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6018414361459234e-05,
+      "loss": 3.5969,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601004479275244e-05,
+      "loss": 3.5982,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600165884524192e-05,
+      "loss": 3.6059,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59932728977314e-05,
+      "loss": 3.6084,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.598488695022089e-05,
+      "loss": 3.6006,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59765173815141e-05,
+      "loss": 3.6222,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596813143400358e-05,
+      "loss": 3.6038,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595974548649306e-05,
+      "loss": 3.6022,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595135953898254e-05,
+      "loss": 3.6164,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.594298997027575e-05,
+      "loss": 3.6008,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.593460402276523e-05,
+      "loss": 3.5991,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592621807525471e-05,
+      "loss": 3.6047,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591783212774419e-05,
+      "loss": 3.5981,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5909462559037397e-05,
+      "loss": 3.6029,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5901076611526877e-05,
+      "loss": 3.6095,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5892690664016356e-05,
+      "loss": 3.6036,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588430471650584e-05,
+      "loss": 3.6063,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.587593514779905e-05,
+      "loss": 3.6049,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586754920028853e-05,
+      "loss": 3.6027,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585916325277801e-05,
+      "loss": 3.6064,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585077730526749e-05,
+      "loss": 3.5724,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58424077365607e-05,
+      "loss": 3.6077,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583402178905018e-05,
+      "loss": 3.5928,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582563584153966e-05,
+      "loss": 3.6091,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581724989402914e-05,
+      "loss": 3.5938,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580888032532235e-05,
+      "loss": 3.5849,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580049437781183e-05,
+      "loss": 3.5862,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579210843030131e-05,
+      "loss": 3.5921,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.57837224827908e-05,
+      "loss": 3.6044,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5775352914084006e-05,
+      "loss": 3.5841,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5766966966573486e-05,
+      "loss": 3.6096,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5758581019062966e-05,
+      "loss": 3.6032,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5750195071552446e-05,
+      "loss": 3.5795,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5741825502845655e-05,
+      "loss": 3.6022,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5733439555335135e-05,
+      "loss": 3.5871,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5725053607824615e-05,
+      "loss": 3.5869,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5716667660314095e-05,
+      "loss": 3.5933,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5708298091607304e-05,
+      "loss": 3.6071,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5699912144096784e-05,
+      "loss": 3.5835,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5691526196586264e-05,
+      "loss": 3.5978,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5683140249075744e-05,
+      "loss": 3.5807,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.567475430156523e-05,
+      "loss": 3.599,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.566638473285844e-05,
+      "loss": 3.6012,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.565799878534792e-05,
+      "loss": 3.6045,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56496128378374e-05,
+      "loss": 3.603,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564124326913061e-05,
+      "loss": 3.6067,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563285732162009e-05,
+      "loss": 3.6209,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562447137410957e-05,
+      "loss": 3.6032,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561608542659905e-05,
+      "loss": 3.6082,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560769947908853e-05,
+      "loss": 3.5902,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559932991038174e-05,
+      "loss": 3.5952,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559094396287122e-05,
+      "loss": 3.5978,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55825580153607e-05,
+      "loss": 3.5992,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5574172067850184e-05,
+      "loss": 3.6058,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556580249914339e-05,
+      "loss": 3.6077,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555741655163287e-05,
+      "loss": 3.5892,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554903060412235e-05,
+      "loss": 3.5904,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554064465661183e-05,
+      "loss": 3.5896,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.553227508790504e-05,
+      "loss": 3.6097,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.552388914039452e-05,
+      "loss": 3.5982,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5515503192884e-05,
+      "loss": 3.5882,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.550711724537348e-05,
+      "loss": 3.5999,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549874767666669e-05,
+      "loss": 3.595,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549036172915617e-05,
+      "loss": 3.592,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548197578164565e-05,
+      "loss": 3.5908,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547358983413514e-05,
+      "loss": 3.6062,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546522026542835e-05,
+      "loss": 3.6015,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545683431791783e-05,
+      "loss": 3.6025,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544844837040731e-05,
+      "loss": 3.5868,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544006242289678e-05,
+      "loss": 3.5935,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5431692854189996e-05,
+      "loss": 3.5948,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5423306906679476e-05,
+      "loss": 3.602,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5414920959168956e-05,
+      "loss": 3.6024,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.540653501165843e-05,
+      "loss": 3.601,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5398165442951645e-05,
+      "loss": 3.6009,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5389779495441125e-05,
+      "loss": 3.5974,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5381393547930605e-05,
+      "loss": 3.596,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5373007600420085e-05,
+      "loss": 3.586,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53646380317133e-05,
+      "loss": 3.5897,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.535625208420278e-05,
+      "loss": 3.5899,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5347866136692254e-05,
+      "loss": 3.5954,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5339480189181734e-05,
+      "loss": 3.5821,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.533111062047495e-05,
+      "loss": 3.5859,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.532272467296443e-05,
+      "loss": 3.5844,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.53143387254539e-05,
+      "loss": 3.5929,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.530595277794338e-05,
+      "loss": 3.5925,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52975832092366e-05,
+      "loss": 3.596,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528919726172608e-05,
+      "loss": 3.5995,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.528081131421556e-05,
+      "loss": 3.5995,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.527242536670504e-05,
+      "loss": 3.5924,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5264055797998254e-05,
+      "loss": 3.6026,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525566985048773e-05,
+      "loss": 3.595,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.524728390297721e-05,
+      "loss": 3.6006,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523889795546669e-05,
+      "loss": 3.5934,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52305283867599e-05,
+      "loss": 3.5924,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5222142439249376e-05,
+      "loss": 3.5932,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5213756491738856e-05,
+      "loss": 3.5968,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5205370544228336e-05,
+      "loss": 3.5883,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519700097552155e-05,
+      "loss": 3.5918,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518861502801103e-05,
+      "loss": 3.5789,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518022908050051e-05,
+      "loss": 3.587,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517184313298999e-05,
+      "loss": 3.5996,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.51634735642832e-05,
+      "loss": 3.5922,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515508761677268e-05,
+      "loss": 3.6045,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514670166926216e-05,
+      "loss": 3.5793,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.513831572175164e-05,
+      "loss": 3.6032,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512994615304485e-05,
+      "loss": 3.5971,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.512156020553433e-05,
+      "loss": 3.6001,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.511317425802381e-05,
+      "loss": 3.5882,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.510478831051329e-05,
+      "loss": 3.5935,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5096418741806506e-05,
+      "loss": 3.6047,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5088032794295986e-05,
+      "loss": 3.6033,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5079646846785466e-05,
+      "loss": 3.5916,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5071260899274946e-05,
+      "loss": 3.5911,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5062891330568155e-05,
+      "loss": 3.5962,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5054505383057635e-05,
+      "loss": 3.5931,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5046119435547115e-05,
+      "loss": 3.5852,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5037733488036595e-05,
+      "loss": 3.5968,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5029347540526075e-05,
+      "loss": 3.5907,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5020977971819284e-05,
+      "loss": 3.5946,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5012592024308764e-05,
+      "loss": 3.5959,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8417012691497803,
+      "eval_runtime": 302.8328,
+      "eval_samples_per_second": 1260.072,
+      "eval_steps_per_second": 39.378,
+      "step": 915840
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.369302133115085e+17,
+  "trial_name": null,
+  "trial_params": null
+}