diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 4.083654880523682,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/existential-there-quantifier/lstm/1/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "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": 10.8197,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5496,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0587,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 7.0003,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9467,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9226,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7712,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6663,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5618,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4812,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4138,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3469,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2679,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.2028,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1428,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0893,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0413,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.008,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9596,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9256,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.893,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.844,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8126,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7802,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7629,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979036769104073e-05,
+      "loss": 5.7265,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978198174353021e-05,
+      "loss": 5.6971,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6779,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6502,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6223,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5997,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5903,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5673,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5576,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5353,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.516,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5113,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4703,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4597,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4415,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4355,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4107,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4177,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3821,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3815,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3764,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.368,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3586,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3361,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3157,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.328,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3083,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.2947,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2771,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2829,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2633,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2768,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2462,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2388,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2396,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2243,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2019,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2062,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1845,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1856,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.1951,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1871,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1688,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1493,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1433,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1539,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1439,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1386,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1373,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1358,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1271,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.0987,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1043,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0974,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.0837,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1039,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0823,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0872,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0723,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0478,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0536,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0566,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0512,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0458,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0322,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0309,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0197,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0126,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0105,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0034,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0118,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9931,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9869,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9804,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9826,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9811,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9709,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9625,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.947,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9555,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9424,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9377,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9391,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9372,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9334,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9291,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9181,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9194,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9087,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9033,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9125,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.903,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9011,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.8915,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.8998,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.8955,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8764,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8924,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8791,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8805,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8589,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8543,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8493,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8538,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8411,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8599,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8563,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8386,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8425,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8379,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.839,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8388,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8401,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8208,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.8245,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8175,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8087,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8209,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8079,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.7971,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8133,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.795,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.7923,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7955,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7869,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.752319812774658,
+      "eval_runtime": 371.8701,
+      "eval_samples_per_second": 1026.141,
+      "eval_steps_per_second": 32.068,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7731,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7828,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7988,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7743,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7864,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7606,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7765,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7533,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.7644,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7637,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7656,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865035381491821e-05,
+      "loss": 4.7624,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7452,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.7406,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7364,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7408,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7538,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7361,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7426,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.7542,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857489666612726e-05,
+      "loss": 4.7285,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7256,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7253,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7377,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7161,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.719,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7158,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7143,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.698,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7018,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.6985,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848266762231527e-05,
+      "loss": 4.7072,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847428167480475e-05,
+      "loss": 4.712,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846589572729423e-05,
+      "loss": 4.6989,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.6939,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.7055,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.6778,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8432368316055884e-05,
+      "loss": 4.6842,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8423982368545363e-05,
+      "loss": 4.6729,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8415596421034843e-05,
+      "loss": 4.6804,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840721047352432e-05,
+      "loss": 4.6722,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6871,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6686,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838206900979649e-05,
+      "loss": 4.6755,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837368306228597e-05,
+      "loss": 4.6698,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836529711477545e-05,
+      "loss": 4.675,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835691116726493e-05,
+      "loss": 4.6698,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834852521975441e-05,
+      "loss": 4.6601,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834013927224389e-05,
+      "loss": 4.6475,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.671,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.6575,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831499780851606e-05,
+      "loss": 4.6509,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830661186100554e-05,
+      "loss": 4.6451,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829822591349502e-05,
+      "loss": 4.6524,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82898399659845e-05,
+      "loss": 4.6422,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828147039727771e-05,
+      "loss": 4.6603,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827308444976719e-05,
+      "loss": 4.6417,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826469850225667e-05,
+      "loss": 4.6397,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825631255474615e-05,
+      "loss": 4.6487,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.6292,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.623,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6348,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6162,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6231,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6331,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.6369,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.622,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180855405955206e-05,
+      "loss": 4.615,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172469458444685e-05,
+      "loss": 4.6064,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164099889737895e-05,
+      "loss": 4.627,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155713942227375e-05,
+      "loss": 4.6201,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147327994716854e-05,
+      "loss": 4.6211,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138942047206334e-05,
+      "loss": 4.6187,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.6278,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.6229,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.5999,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6113,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6103,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808867549855441e-05,
+      "loss": 4.5921,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808028955104389e-05,
+      "loss": 4.622,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807190360353337e-05,
+      "loss": 4.6067,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806351765602285e-05,
+      "loss": 4.6119,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805513170851233e-05,
+      "loss": 4.6084,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804674576100181e-05,
+      "loss": 4.5826,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803835981349129e-05,
+      "loss": 4.598,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802997386598077e-05,
+      "loss": 4.5967,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802158791847025e-05,
+      "loss": 4.5988,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.5967,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800481602344921e-05,
+      "loss": 4.592,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799643007593869e-05,
+      "loss": 4.5934,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798804412842817e-05,
+      "loss": 4.5889,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797967455972138e-05,
+      "loss": 4.574,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.5811,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.5837,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795451671718982e-05,
+      "loss": 4.5868,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79461307696793e-05,
+      "loss": 4.5754,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793774482216878e-05,
+      "loss": 4.5675,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929375253461986e-05,
+      "loss": 4.5695,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.5793,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5778,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.571,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789584784222364e-05,
+      "loss": 4.5621,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788746189471312e-05,
+      "loss": 4.5552,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78790759472026e-05,
+      "loss": 4.5664,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787068999969208e-05,
+      "loss": 4.5527,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786230405218156e-05,
+      "loss": 4.5563,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785393448347477e-05,
+      "loss": 4.5582,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784554853596425e-05,
+      "loss": 4.5593,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783716258845373e-05,
+      "loss": 4.5562,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782877664094321e-05,
+      "loss": 4.5601,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782039069343269e-05,
+      "loss": 4.5469,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.5504,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5472,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795265608508595e-05,
+      "loss": 4.5405,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786879660998075e-05,
+      "loss": 4.5567,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778493713487555e-05,
+      "loss": 4.5508,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770107765977035e-05,
+      "loss": 4.5493,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761721818466515e-05,
+      "loss": 4.5423,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753335870955995e-05,
+      "loss": 4.549,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7744949923445475e-05,
+      "loss": 4.5466,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736563975934955e-05,
+      "loss": 4.5323,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728178028424435e-05,
+      "loss": 4.5536,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771979208091391e-05,
+      "loss": 4.5395,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771140613340339e-05,
+      "loss": 4.5421,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703020185892875e-05,
+      "loss": 4.5258,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694650617186084e-05,
+      "loss": 4.5267,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686264669675564e-05,
+      "loss": 4.5204,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767789510096878e-05,
+      "loss": 4.5246,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766950915345826e-05,
+      "loss": 4.5187,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766112320594773e-05,
+      "loss": 4.5364,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765273725843721e-05,
+      "loss": 4.5376,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764435131092669e-05,
+      "loss": 4.5252,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763596536341617e-05,
+      "loss": 4.5201,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762757941590565e-05,
+      "loss": 4.5254,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761919346839513e-05,
+      "loss": 4.5269,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5338,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760245433098156e-05,
+      "loss": 4.5285,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5198,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758568243596052e-05,
+      "loss": 4.5134,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757729648845e-05,
+      "loss": 4.5203,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756891054093948e-05,
+      "loss": 4.5134,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756052459342896e-05,
+      "loss": 4.5177,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755213864591844e-05,
+      "loss": 4.5143,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754375269840792e-05,
+      "loss": 4.507,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535383129701126e-05,
+      "loss": 4.5215,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7526997182190606e-05,
+      "loss": 4.5078,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5013,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5067,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.5021,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.473540782928467,
+      "eval_runtime": 372.6137,
+      "eval_samples_per_second": 1024.093,
+      "eval_steps_per_second": 32.004,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.4882,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.5017,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.518,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.4959,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5079,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.4907,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5037,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.4786,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.4985,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.4961,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.4961,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.5026,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739285477962975e-05,
+      "loss": 4.4785,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738446883211923e-05,
+      "loss": 4.4801,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737608288460871e-05,
+      "loss": 4.4763,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4851,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4926,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4825,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.485,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.5019,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.479,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4731,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730901168332828e-05,
+      "loss": 4.4801,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730062573581776e-05,
+      "loss": 4.4901,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729223978830724e-05,
+      "loss": 4.473,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728385384079672e-05,
+      "loss": 4.4736,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72754678932862e-05,
+      "loss": 4.4716,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726708194577568e-05,
+      "loss": 4.4723,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.4573,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250326429558375e-05,
+      "loss": 4.4644,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241940482047855e-05,
+      "loss": 4.4662,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233554534537335e-05,
+      "loss": 4.4666,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722516858702681e-05,
+      "loss": 4.4798,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721678263951629e-05,
+      "loss": 4.4599,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208413070809504e-05,
+      "loss": 4.4636,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200027123298984e-05,
+      "loss": 4.4719,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719164117578846e-05,
+      "loss": 4.4526,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718325522827794e-05,
+      "loss": 4.4516,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717486928076742e-05,
+      "loss": 4.4487,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716649971206063e-05,
+      "loss": 4.4554,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715811376455011e-05,
+      "loss": 4.4449,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4644,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714134186952907e-05,
+      "loss": 4.4432,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713295592201855e-05,
+      "loss": 4.46,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712456997450803e-05,
+      "loss": 4.4503,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711618402699751e-05,
+      "loss": 4.4502,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710781445829072e-05,
+      "loss": 4.4547,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70994285107802e-05,
+      "loss": 4.4407,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4324,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4561,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4449,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.4358,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705751515203134e-05,
+      "loss": 4.4378,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704912920452082e-05,
+      "loss": 4.4452,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70407432570103e-05,
+      "loss": 4.4263,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703235730949978e-05,
+      "loss": 4.4495,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.4342,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4316,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007215845771946e-05,
+      "loss": 4.445,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998829898261426e-05,
+      "loss": 4.4247,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.421,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.4362,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973688434533595e-05,
+      "loss": 4.4178,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965302487023075e-05,
+      "loss": 4.4203,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4329,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.4387,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940144644491515e-05,
+      "loss": 4.421,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931758696980995e-05,
+      "loss": 4.4218,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4079,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4322,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906600854449435e-05,
+      "loss": 4.425,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689824766454638e-05,
+      "loss": 4.4304,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688986171703586e-05,
+      "loss": 4.4248,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688147576952534e-05,
+      "loss": 4.434,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687308982201481e-05,
+      "loss": 4.4371,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686470387450429e-05,
+      "loss": 4.41,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685631792699377e-05,
+      "loss": 4.4202,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4252,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.4052,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.4345,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.421,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4267,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.4259,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679763267322387e-05,
+      "loss": 4.4005,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678924672571335e-05,
+      "loss": 4.4169,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678086077820283e-05,
+      "loss": 4.4126,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677249120949604e-05,
+      "loss": 4.4193,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676410526198552e-05,
+      "loss": 4.4117,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755719314475e-05,
+      "loss": 4.415,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674733336696448e-05,
+      "loss": 4.4136,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673894741945396e-05,
+      "loss": 4.4094,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.3976,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672219190323665e-05,
+      "loss": 4.403,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.4091,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670542000821561e-05,
+      "loss": 4.4127,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669703406070509e-05,
+      "loss": 4.405,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668864811319457e-05,
+      "loss": 4.3924,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668026216568405e-05,
+      "loss": 4.3974,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667187621817353e-05,
+      "loss": 4.4066,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663490270663006e-05,
+      "loss": 4.4032,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655104323152486e-05,
+      "loss": 4.399,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6646718375641966e-05,
+      "loss": 4.3941,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638332428131446e-05,
+      "loss": 4.3844,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662996285942466e-05,
+      "loss": 4.4016,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662157691191414e-05,
+      "loss": 4.3798,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.3924,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.3844,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.4007,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.3861,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657966355316527e-05,
+      "loss": 4.397,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657127760565475e-05,
+      "loss": 4.3847,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656289165814423e-05,
+      "loss": 4.3831,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655450571063371e-05,
+      "loss": 4.3897,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654613614192692e-05,
+      "loss": 4.376,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65377501944164e-05,
+      "loss": 4.395,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652936424690588e-05,
+      "loss": 4.3929,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6520994678199095e-05,
+      "loss": 4.3905,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.3808,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.3869,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495836835667535e-05,
+      "loss": 4.3904,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487467266960744e-05,
+      "loss": 4.3705,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479081319450224e-05,
+      "loss": 4.395,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470695371939704e-05,
+      "loss": 4.3841,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646232580323291e-05,
+      "loss": 4.3839,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645393985572239e-05,
+      "loss": 4.3737,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644555390821187e-05,
+      "loss": 4.3753,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643716796070135e-05,
+      "loss": 4.361,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642878201319083e-05,
+      "loss": 4.3724,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642041244448405e-05,
+      "loss": 4.366,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641202649697353e-05,
+      "loss": 4.3818,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640364054946301e-05,
+      "loss": 4.3851,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639525460195249e-05,
+      "loss": 4.3763,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638686865444197e-05,
+      "loss": 4.3707,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637848270693145e-05,
+      "loss": 4.3733,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637009675942093e-05,
+      "loss": 4.3774,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636171081191041e-05,
+      "loss": 4.3824,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635332486439989e-05,
+      "loss": 4.378,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634493891688936e-05,
+      "loss": 4.3664,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.3736,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632819977947579e-05,
+      "loss": 4.3755,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631981383196527e-05,
+      "loss": 4.3623,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631142788445475e-05,
+      "loss": 4.3728,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630304193694423e-05,
+      "loss": 4.3667,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629465598943371e-05,
+      "loss": 4.3624,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628627004192319e-05,
+      "loss": 4.3761,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277884094412666e-05,
+      "loss": 4.3648,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269498146902146e-05,
+      "loss": 4.3581,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261112199391626e-05,
+      "loss": 4.3637,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252726251881106e-05,
+      "loss": 4.3565,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.338195323944092,
+      "eval_runtime": 303.1267,
+      "eval_samples_per_second": 1258.85,
+      "eval_steps_per_second": 39.34,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244356683174315e-05,
+      "loss": 4.3524,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6235970735663795e-05,
+      "loss": 4.3569,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227584788153275e-05,
+      "loss": 4.3773,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219198840642755e-05,
+      "loss": 4.3568,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210812893132235e-05,
+      "loss": 4.3736,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202426945621715e-05,
+      "loss": 4.3468,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619405737691493e-05,
+      "loss": 4.3649,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618567142940441e-05,
+      "loss": 4.3421,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617728548189389e-05,
+      "loss": 4.3605,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61689159131871e-05,
+      "loss": 4.3607,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616052996567658e-05,
+      "loss": 4.3577,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615214401816606e-05,
+      "loss": 4.3621,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614375807065554e-05,
+      "loss": 4.3436,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613537212314502e-05,
+      "loss": 4.3437,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61269861756345e-05,
+      "loss": 4.3427,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611860022812398e-05,
+      "loss": 4.3464,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611021428061346e-05,
+      "loss": 4.3562,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610182833310294e-05,
+      "loss": 4.353,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609344238559242e-05,
+      "loss": 4.349,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085072816885636e-05,
+      "loss": 4.3667,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076686869375116e-05,
+      "loss": 4.3492,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068300921864595e-05,
+      "loss": 4.3469,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059914974354075e-05,
+      "loss": 4.3476,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051529026843555e-05,
+      "loss": 4.3547,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604314307933303e-05,
+      "loss": 4.3408,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603475713182251e-05,
+      "loss": 4.3457,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602637118431199e-05,
+      "loss": 4.3439,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60180016156052e-05,
+      "loss": 4.3388,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600961566809468e-05,
+      "loss": 4.332,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600122972058416e-05,
+      "loss": 4.3344,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599286015187737e-05,
+      "loss": 4.3406,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598447420436685e-05,
+      "loss": 4.3397,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597608825685633e-05,
+      "loss": 4.3518,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596770230934581e-05,
+      "loss": 4.3397,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595933274063903e-05,
+      "loss": 4.3311,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59509467931285e-05,
+      "loss": 4.3415,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594256084561798e-05,
+      "loss": 4.3336,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593417489810746e-05,
+      "loss": 4.3237,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592578895059694e-05,
+      "loss": 4.3271,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591740300308642e-05,
+      "loss": 4.3278,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59090170555759e-05,
+      "loss": 4.324,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590063110806538e-05,
+      "loss": 4.3407,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589226153935859e-05,
+      "loss": 4.3237,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588387559184808e-05,
+      "loss": 4.3359,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587548964433756e-05,
+      "loss": 4.3315,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586710369682704e-05,
+      "loss": 4.3263,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585873412812025e-05,
+      "loss": 4.3359,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585034818060973e-05,
+      "loss": 4.3168,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584196223309921e-05,
+      "loss": 4.3061,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583357628558869e-05,
+      "loss": 4.3388,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825206716881896e-05,
+      "loss": 4.3255,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816820769371376e-05,
+      "loss": 4.3205,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808434821860856e-05,
+      "loss": 4.3174,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800048874350336e-05,
+      "loss": 4.3235,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579169568444728e-05,
+      "loss": 4.3113,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578330973693676e-05,
+      "loss": 4.3263,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577492378942624e-05,
+      "loss": 4.3204,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3173,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3258,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574978232569841e-05,
+      "loss": 4.3077,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3046,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3226,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572464086197058e-05,
+      "loss": 4.2977,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571625491446006e-05,
+      "loss": 4.3021,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570786896694954e-05,
+      "loss": 4.3209,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3241,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.304,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3134,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.2925,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665939229396945e-05,
+      "loss": 4.3188,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657553281886425e-05,
+      "loss": 4.3114,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3201,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.309,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.3217,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56240422494518e-05,
+      "loss": 4.3271,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561565630194128e-05,
+      "loss": 4.2962,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560727035443076e-05,
+      "loss": 4.314,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559888440692024e-05,
+      "loss": 4.3162,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.2898,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821125118992e-05,
+      "loss": 4.3247,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557372656438868e-05,
+      "loss": 4.3129,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556534061687816e-05,
+      "loss": 4.3129,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555697104817138e-05,
+      "loss": 4.3206,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554858510066086e-05,
+      "loss": 4.2941,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554019915315033e-05,
+      "loss": 4.3013,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553181320563981e-05,
+      "loss": 4.3069,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552342725812929e-05,
+      "loss": 4.311,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551505768942251e-05,
+      "loss": 4.3036,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550668812071572e-05,
+      "loss": 4.3083,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54983021732052e-05,
+      "loss": 4.3009,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548991622569468e-05,
+      "loss": 4.3107,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481530278184157e-05,
+      "loss": 4.2903,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473144330673637e-05,
+      "loss": 4.2989,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464758383163116e-05,
+      "loss": 4.3039,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456372435652596e-05,
+      "loss": 4.3072,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5447986488142076e-05,
+      "loss": 4.3015,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439600540631556e-05,
+      "loss": 4.2873,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431230971924765e-05,
+      "loss": 4.2904,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422845024414245e-05,
+      "loss": 4.2994,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414459076903725e-05,
+      "loss": 4.2993,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406073129393205e-05,
+      "loss": 4.2976,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397687181882685e-05,
+      "loss": 4.2895,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389301234372165e-05,
+      "loss": 4.2838,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380915286861645e-05,
+      "loss": 4.2981,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372529339351125e-05,
+      "loss": 4.2788,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536415977064434e-05,
+      "loss": 4.2875,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535579020193755e-05,
+      "loss": 4.281,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534740425442703e-05,
+      "loss": 4.2964,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533901830691651e-05,
+      "loss": 4.2871,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533063235940599e-05,
+      "loss": 4.2998,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532224641189547e-05,
+      "loss": 4.2791,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531386046438495e-05,
+      "loss": 4.2839,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.2873,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.2814,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.2909,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.2977,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527194710563608e-05,
+      "loss": 4.291,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.2839,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255191589418775e-05,
+      "loss": 4.2876,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2902,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.2721,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.2975,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221664178180424e-05,
+      "loss": 4.2823,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213278230669904e-05,
+      "loss": 4.2888,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5204892283159384e-05,
+      "loss": 4.281,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196506335648863e-05,
+      "loss": 4.2794,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188120388138343e-05,
+      "loss": 4.2639,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179734440627817e-05,
+      "loss": 4.2818,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171348493117297e-05,
+      "loss": 4.2703,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5162962545606777e-05,
+      "loss": 4.2818,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154576598096256e-05,
+      "loss": 4.2878,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514619065058574e-05,
+      "loss": 4.2865,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513782108187895e-05,
+      "loss": 4.2698,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.2816,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512104918685791e-05,
+      "loss": 4.2854,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511266323934739e-05,
+      "loss": 4.2858,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510427729183687e-05,
+      "loss": 4.2834,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509590772313008e-05,
+      "loss": 4.276,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508752177561956e-05,
+      "loss": 4.2756,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507913582810904e-05,
+      "loss": 4.284,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507074988059852e-05,
+      "loss": 4.2704,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062363933088e-05,
+      "loss": 4.2821,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505397798557748e-05,
+      "loss": 4.2692,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504559203806696e-05,
+      "loss": 4.2758,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503720609055644e-05,
+      "loss": 4.2853,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502883652184966e-05,
+      "loss": 4.2731,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020466953142866e-05,
+      "loss": 4.2654,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.2683,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2703,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.254990100860596,
+      "eval_runtime": 302.7181,
+      "eval_samples_per_second": 1260.549,
+      "eval_steps_per_second": 39.393,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.2566,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.2674,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2829,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2673,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2823,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2638,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2701,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2581,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2665,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2716,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.2712,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2747,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489471049809253e-05,
+      "loss": 4.2532,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886324550582003e-05,
+      "loss": 4.2575,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877938603071483e-05,
+      "loss": 4.2572,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486955265556096e-05,
+      "loss": 4.2566,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486118308685418e-05,
+      "loss": 4.2682,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485279713934365e-05,
+      "loss": 4.2627,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484441119183313e-05,
+      "loss": 4.2657,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483602524432262e-05,
+      "loss": 4.2795,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48276392968121e-05,
+      "loss": 4.267,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481925334930158e-05,
+      "loss": 4.2598,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481086740179106e-05,
+      "loss": 4.2607,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480248145428054e-05,
+      "loss": 4.2667,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.2549,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2618,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.2563,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2552,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476056809553167e-05,
+      "loss": 4.2507,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2506,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2532,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2575,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.2661,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718654736782804e-05,
+      "loss": 4.2557,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471028516807601e-05,
+      "loss": 4.2465,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470191559936922e-05,
+      "loss": 4.2587,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46935296518587e-05,
+      "loss": 4.2554,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468514370434818e-05,
+      "loss": 4.237,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467675775683766e-05,
+      "loss": 4.2426,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466837180932714e-05,
+      "loss": 4.2483,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465998586181662e-05,
+      "loss": 4.24,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.46515999143061e-05,
+      "loss": 4.2609,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464321396679558e-05,
+      "loss": 4.2416,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463482801928506e-05,
+      "loss": 4.2498,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462645845057827e-05,
+      "loss": 4.2513,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461807250306775e-05,
+      "loss": 4.2498,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460968655555724e-05,
+      "loss": 4.2531,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460130060804672e-05,
+      "loss": 4.2398,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459291466053619e-05,
+      "loss": 4.2213,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458452871302567e-05,
+      "loss": 4.2635,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576159144318886e-05,
+      "loss": 4.2413,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567773196808366e-05,
+      "loss": 4.2436,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455938724929784e-05,
+      "loss": 4.2419,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455100130178732e-05,
+      "loss": 4.237,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.2356,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2443,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452585983805949e-05,
+      "loss": 4.2421,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517473890548975e-05,
+      "loss": 4.238,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2513,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500701995527935e-05,
+      "loss": 4.2263,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2258,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.2485,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.2165,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.223,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458788636779064e-05,
+      "loss": 4.2454,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445041906807227e-05,
+      "loss": 4.2461,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444203312056175e-05,
+      "loss": 4.2287,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443364717305123e-05,
+      "loss": 4.2377,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442526122554071e-05,
+      "loss": 4.218,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441687527803019e-05,
+      "loss": 4.2401,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440848933051967e-05,
+      "loss": 4.2335,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440010338300916e-05,
+      "loss": 4.2427,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439171743549864e-05,
+      "loss": 4.2338,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2455,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2511,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2245,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435819002426029e-05,
+      "loss": 4.2318,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434980407674977e-05,
+      "loss": 4.2425,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434143450804298e-05,
+      "loss": 4.2166,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2454,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.2415,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316293044315146e-05,
+      "loss": 4.2406,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307907096804626e-05,
+      "loss": 4.2444,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429952114929411e-05,
+      "loss": 4.2226,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429115158058732e-05,
+      "loss": 4.2243,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42827656330768e-05,
+      "loss": 4.2328,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427437968556628e-05,
+      "loss": 4.2401,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426599373805576e-05,
+      "loss": 4.2324,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425760779054524e-05,
+      "loss": 4.2317,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424922184303472e-05,
+      "loss": 4.2235,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42408358955242e-05,
+      "loss": 4.2385,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2221,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2219,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2282,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207292105482115e-05,
+      "loss": 4.2364,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198906157971595e-05,
+      "loss": 4.2275,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419052021046108e-05,
+      "loss": 4.2171,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2194,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.2263,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416537874673325e-05,
+      "loss": 4.2314,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415699279922273e-05,
+      "loss": 4.2237,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414860685171221e-05,
+      "loss": 4.2217,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414022090420169e-05,
+      "loss": 4.2127,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41318513354949e-05,
+      "loss": 4.221,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.2089,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2225,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2032,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409830754545282e-05,
+      "loss": 4.2327,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2162,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2229,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2128,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406476375541074e-05,
+      "loss": 4.2134,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2177,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2129,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039622291682915e-05,
+      "loss": 4.2229,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403123634417239e-05,
+      "loss": 4.2258,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402285039666187e-05,
+      "loss": 4.2221,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401446444915135e-05,
+      "loss": 4.2164,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400607850164083e-05,
+      "loss": 4.2178,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399769255413031e-05,
+      "loss": 4.2263,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398933936422725e-05,
+      "loss": 4.1997,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398095341671673e-05,
+      "loss": 4.2282,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397256746920621e-05,
+      "loss": 4.2128,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396418152169569e-05,
+      "loss": 4.2235,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395579557418517e-05,
+      "loss": 4.2116,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394740962667465e-05,
+      "loss": 4.2125,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393902367916413e-05,
+      "loss": 4.1985,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.212,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.207,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39138822154363e-05,
+      "loss": 4.2134,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390549626792578e-05,
+      "loss": 4.2169,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389711032041526e-05,
+      "loss": 4.2211,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388872437290474e-05,
+      "loss": 4.2027,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.2144,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387196885668744e-05,
+      "loss": 4.221,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386358290917692e-05,
+      "loss": 4.2193,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38551969616664e-05,
+      "loss": 4.2127,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384681101415588e-05,
+      "loss": 4.2159,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383842506664536e-05,
+      "loss": 4.2079,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830055497938566e-05,
+      "loss": 4.2178,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821669550428046e-05,
+      "loss": 4.2021,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813283602917526e-05,
+      "loss": 4.216,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804897655407006e-05,
+      "loss": 4.2064,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796511707896486e-05,
+      "loss": 4.2077,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788125760385966e-05,
+      "loss": 4.2191,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779739812875446e-05,
+      "loss": 4.21,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771370244168655e-05,
+      "loss": 4.2027,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376300067546187e-05,
+      "loss": 4.2002,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375461472795135e-05,
+      "loss": 4.207,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.198840618133545,
+      "eval_runtime": 296.194,
+      "eval_samples_per_second": 1288.315,
+      "eval_steps_per_second": 40.261,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374622878044083e-05,
+      "loss": 4.2011,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373784283293031e-05,
+      "loss": 4.2004,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372945688541979e-05,
+      "loss": 4.2184,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721087316713e-05,
+      "loss": 4.2047,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371270136920248e-05,
+      "loss": 4.2165,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370431542169196e-05,
+      "loss": 4.2023,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369592947418144e-05,
+      "loss": 4.2015,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368754352667092e-05,
+      "loss": 4.1967,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367915757916039e-05,
+      "loss": 4.1978,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367077163164987e-05,
+      "loss": 4.2105,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366238568413936e-05,
+      "loss": 4.2105,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654016115432576e-05,
+      "loss": 4.2068,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645646546725785e-05,
+      "loss": 4.1965,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637260599215265e-05,
+      "loss": 4.1937,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3628874651704745e-05,
+      "loss": 4.1933,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620488704194225e-05,
+      "loss": 4.1917,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36121027566837e-05,
+      "loss": 4.2047,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360371680917318e-05,
+      "loss": 4.2009,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359533086166266e-05,
+      "loss": 4.2042,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358696129295587e-05,
+      "loss": 4.2148,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357857534544535e-05,
+      "loss": 4.2092,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357018939793483e-05,
+      "loss": 4.1935,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356180345042431e-05,
+      "loss": 4.1992,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355341750291379e-05,
+      "loss": 4.2028,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354503155540327e-05,
+      "loss": 4.1951,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353664560789275e-05,
+      "loss": 4.2009,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352825966038223e-05,
+      "loss": 4.1955,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351987371287171e-05,
+      "loss": 4.1955,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351150414416492e-05,
+      "loss": 4.1932,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35031181966544e-05,
+      "loss": 4.1865,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349473224914388e-05,
+      "loss": 4.1929,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348634630163336e-05,
+      "loss": 4.1953,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347796035412284e-05,
+      "loss": 4.2032,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346957440661232e-05,
+      "loss": 4.1974,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34611884591018e-05,
+      "loss": 4.1853,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345281889039501e-05,
+      "loss": 4.201,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344443294288449e-05,
+      "loss": 4.1946,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343604699537398e-05,
+      "loss": 4.1749,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342766104786346e-05,
+      "loss": 4.1893,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341927510035294e-05,
+      "loss": 4.1864,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341088915284241e-05,
+      "loss": 4.1821,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340250320533189e-05,
+      "loss": 4.1995,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339411725782137e-05,
+      "loss": 4.1843,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338574768911458e-05,
+      "loss": 4.1892,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.1907,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.1894,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360606225386756e-05,
+      "loss": 4.1985,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335222027787623e-05,
+      "loss": 4.1794,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343834330365716e-05,
+      "loss": 4.1593,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333546476165893e-05,
+      "loss": 4.2063,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332707881414841e-05,
+      "loss": 4.1821,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318692866637885e-05,
+      "loss": 4.1904,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.179,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.1799,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.1828,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285165455399534e-05,
+      "loss": 4.1812,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276779507889013e-05,
+      "loss": 4.1862,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.1762,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.1948,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.1727,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324325209665067e-05,
+      "loss": 4.1652,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323486614914015e-05,
+      "loss": 4.1905,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322648020162963e-05,
+      "loss": 4.1631,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321809425411911e-05,
+      "loss": 4.1641,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320970830660859e-05,
+      "loss": 4.1871,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320132235909807e-05,
+      "loss": 4.1861,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319293641158755e-05,
+      "loss": 4.1748,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318455046407703e-05,
+      "loss": 4.1809,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317616451656651e-05,
+      "loss": 4.1609,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316779494785972e-05,
+      "loss": 4.1814,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31594090003492e-05,
+      "loss": 4.1759,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315103943164241e-05,
+      "loss": 4.1885,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314265348413189e-05,
+      "loss": 4.1789,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313426753662137e-05,
+      "loss": 4.1853,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3125881589110854e-05,
+      "loss": 4.1945,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311751202040406e-05,
+      "loss": 4.1671,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.18,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.1844,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30923541778725e-05,
+      "loss": 4.1637,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308396823036198e-05,
+      "loss": 4.1872,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307558228285146e-05,
+      "loss": 4.1885,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30672290929484e-05,
+      "loss": 4.1828,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.196,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.1667,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.1699,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.1774,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30252993553958e-05,
+      "loss": 4.18,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301691340788529e-05,
+      "loss": 4.1822,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300852746037477e-05,
+      "loss": 4.1755,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300014151286425e-05,
+      "loss": 4.1664,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299175556535372e-05,
+      "loss": 4.1836,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29833696178432e-05,
+      "loss": 4.1691,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297498367033268e-05,
+      "loss": 4.1661,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296661410162589e-05,
+      "loss": 4.1739,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295822815411537e-05,
+      "loss": 4.186,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949858585408585e-05,
+      "loss": 4.1735,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941472637898065e-05,
+      "loss": 4.1613,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.1658,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.173,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.1753,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2907928847855985e-05,
+      "loss": 4.173,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899542900345465e-05,
+      "loss": 4.1676,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891156952834945e-05,
+      "loss": 4.1583,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2882787384128154e-05,
+      "loss": 4.166,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874401436617634e-05,
+      "loss": 4.1599,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866015489107114e-05,
+      "loss": 4.167,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285764592040032e-05,
+      "loss": 4.1509,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28492599728898e-05,
+      "loss": 4.1817,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284087402537928e-05,
+      "loss": 4.1633,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283248807786876e-05,
+      "loss": 4.1685,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282410213035824e-05,
+      "loss": 4.1565,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281571618284772e-05,
+      "loss": 4.1681,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280736299294467e-05,
+      "loss": 4.1635,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279897704543415e-05,
+      "loss": 4.1587,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279059109792363e-05,
+      "loss": 4.1704,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278220515041311e-05,
+      "loss": 4.1779,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277381920290259e-05,
+      "loss": 4.1641,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276543325539207e-05,
+      "loss": 4.1721,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275704730788155e-05,
+      "loss": 4.1578,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274866136037103e-05,
+      "loss": 4.1735,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274027541286051e-05,
+      "loss": 4.1546,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273188946534999e-05,
+      "loss": 4.1755,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723519896643196e-05,
+      "loss": 4.1592,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715133949132676e-05,
+      "loss": 4.1711,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270676438042589e-05,
+      "loss": 4.1641,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269837843291537e-05,
+      "loss": 4.1618,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268999248540485e-05,
+      "loss": 4.1451,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268160653789433e-05,
+      "loss": 4.1613,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.1558,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.1661,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265644869536277e-05,
+      "loss": 4.1594,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.1715,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263969317914546e-05,
+      "loss": 4.1546,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263130723163494e-05,
+      "loss": 4.1652,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262292128412442e-05,
+      "loss": 4.1696,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26145353366139e-05,
+      "loss": 4.1704,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260614938910338e-05,
+      "loss": 4.163,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259776344159286e-05,
+      "loss": 4.1643,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258937749408234e-05,
+      "loss": 4.1559,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258099154657182e-05,
+      "loss": 4.1728,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257262197786503e-05,
+      "loss": 4.1508,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256423603035451e-05,
+      "loss": 4.1673,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255585008284399e-05,
+      "loss": 4.1537,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254746413533347e-05,
+      "loss": 4.1599,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253907818782295e-05,
+      "loss": 4.1657,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253070861911616e-05,
+      "loss": 4.1565,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252232267160564e-05,
+      "loss": 4.1602,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2513953102898855e-05,
+      "loss": 4.1472,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505567155388335e-05,
+      "loss": 4.1597,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.158144474029541,
+      "eval_runtime": 294.0203,
+      "eval_samples_per_second": 1297.839,
+      "eval_steps_per_second": 40.558,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497181207877815e-05,
+      "loss": 4.1508,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488795260367295e-05,
+      "loss": 4.146,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480409312856774e-05,
+      "loss": 4.1684,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472023365346254e-05,
+      "loss": 4.1581,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463637417835734e-05,
+      "loss": 4.1651,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455251470325214e-05,
+      "loss": 4.153,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446865522814694e-05,
+      "loss": 4.1518,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2438479575304174e-05,
+      "loss": 4.1494,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430093627793654e-05,
+      "loss": 4.1512,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2421707680283134e-05,
+      "loss": 4.1617,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413321732772614e-05,
+      "loss": 4.1571,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240493578526209e-05,
+      "loss": 4.1564,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23965662165553e-05,
+      "loss": 4.1493,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238818026904478e-05,
+      "loss": 4.1453,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237979432153426e-05,
+      "loss": 4.1466,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237140837402374e-05,
+      "loss": 4.142,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236302242651322e-05,
+      "loss": 4.1585,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23546364790027e-05,
+      "loss": 4.1549,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234625053149218e-05,
+      "loss": 4.1518,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233786458398166e-05,
+      "loss": 4.1631,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232949501527487e-05,
+      "loss": 4.1669,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232110906776435e-05,
+      "loss": 4.1483,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231272312025383e-05,
+      "loss": 4.1479,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230433717274331e-05,
+      "loss": 4.1565,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229595122523279e-05,
+      "loss": 4.1506,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228756527772227e-05,
+      "loss": 4.1518,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227917933021175e-05,
+      "loss": 4.1507,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227079338270123e-05,
+      "loss": 4.1478,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226240743519072e-05,
+      "loss": 4.1453,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254054245287657e-05,
+      "loss": 4.142,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245668297777137e-05,
+      "loss": 4.1397,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237282350266617e-05,
+      "loss": 4.1468,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228896402756096e-05,
+      "loss": 4.1561,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220510455245576e-05,
+      "loss": 4.1537,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212124507735056e-05,
+      "loss": 4.1402,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203754939028265e-05,
+      "loss": 4.1504,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195368991517745e-05,
+      "loss": 4.1496,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2186983044007225e-05,
+      "loss": 4.1297,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178597096496705e-05,
+      "loss": 4.1434,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217022752778992e-05,
+      "loss": 4.1399,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21618415802794e-05,
+      "loss": 4.1318,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215345563276888e-05,
+      "loss": 4.1518,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214506968525836e-05,
+      "loss": 4.1381,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213668373774784e-05,
+      "loss": 4.1469,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212829779023732e-05,
+      "loss": 4.1444,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211992822153053e-05,
+      "loss": 4.1423,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211154227402001e-05,
+      "loss": 4.1525,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210315632650949e-05,
+      "loss": 4.1365,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209477037899897e-05,
+      "loss": 4.1108,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208638443148845e-05,
+      "loss": 4.1577,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207799848397792e-05,
+      "loss": 4.1374,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206961253646741e-05,
+      "loss": 4.146,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206122658895689e-05,
+      "loss": 4.1347,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20528570202501e-05,
+      "loss": 4.1349,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044487451543315e-05,
+      "loss": 4.1316,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036101504032795e-05,
+      "loss": 4.1364,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027715556522275e-05,
+      "loss": 4.1431,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201932960901175e-05,
+      "loss": 4.1292,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201094366150123e-05,
+      "loss": 4.151,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200255771399071e-05,
+      "loss": 4.1313,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199417176648019e-05,
+      "loss": 4.1206,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198578581896967e-05,
+      "loss": 4.1392,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197741625026288e-05,
+      "loss": 4.1221,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969030302752363e-05,
+      "loss": 4.1179,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960644355241843e-05,
+      "loss": 4.1455,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952258407731323e-05,
+      "loss": 4.1416,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19438724602208e-05,
+      "loss": 4.1312,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193548651271028e-05,
+      "loss": 4.1338,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192710056519976e-05,
+      "loss": 4.1192,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191871461768924e-05,
+      "loss": 4.1361,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191034504898245e-05,
+      "loss": 4.1368,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190197548027566e-05,
+      "loss": 4.1402,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189358953276514e-05,
+      "loss": 4.1343,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188520358525462e-05,
+      "loss": 4.1383,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768176377441e-05,
+      "loss": 4.1522,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186843169023358e-05,
+      "loss": 4.1238,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186004574272306e-05,
+      "loss": 4.1396,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185165979521254e-05,
+      "loss": 4.1345,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184327384770203e-05,
+      "loss": 4.1239,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183490427899524e-05,
+      "loss": 4.1405,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182651833148472e-05,
+      "loss": 4.1471,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818148762777926e-05,
+      "loss": 4.1368,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809762815267406e-05,
+      "loss": 4.1528,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801376867756886e-05,
+      "loss": 4.1233,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1792990920246366e-05,
+      "loss": 4.1249,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784604972735846e-05,
+      "loss": 4.1339,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776219025225326e-05,
+      "loss": 4.1402,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767849456518535e-05,
+      "loss": 4.1319,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759463509008015e-05,
+      "loss": 4.1363,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751077561497495e-05,
+      "loss": 4.1264,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174269161398698e-05,
+      "loss": 4.1395,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173430566647646e-05,
+      "loss": 4.1261,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725919718965935e-05,
+      "loss": 4.1243,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717533771455415e-05,
+      "loss": 4.1328,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170916420274863e-05,
+      "loss": 4.1429,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170077825523811e-05,
+      "loss": 4.1304,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692392307727584e-05,
+      "loss": 4.1171,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1684006360217064e-05,
+      "loss": 4.1283,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675620412706544e-05,
+      "loss": 4.1268,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667234465196024e-05,
+      "loss": 4.1354,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165886489648923e-05,
+      "loss": 4.1295,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165047894897872e-05,
+      "loss": 4.1241,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16420930014682e-05,
+      "loss": 4.1206,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163370705395768e-05,
+      "loss": 4.1193,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162532110644716e-05,
+      "loss": 4.1275,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161693515893664e-05,
+      "loss": 4.1229,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160854921142612e-05,
+      "loss": 4.1104,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160017964271933e-05,
+      "loss": 4.1384,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159179369520881e-05,
+      "loss": 4.1212,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158340774769829e-05,
+      "loss": 4.1235,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157502180018777e-05,
+      "loss": 4.1154,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156663585267725e-05,
+      "loss": 4.1245,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155826628397046e-05,
+      "loss": 4.1239,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154988033645994e-05,
+      "loss": 4.1182,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154149438894942e-05,
+      "loss": 4.1309,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533108441438904e-05,
+      "loss": 4.1317,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152473887273211e-05,
+      "loss": 4.1266,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151635292522159e-05,
+      "loss": 4.13,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150796697771107e-05,
+      "loss": 4.1176,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149958103020055e-05,
+      "loss": 4.1307,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149119508269003e-05,
+      "loss": 4.1169,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148282551398324e-05,
+      "loss": 4.1318,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147443956647272e-05,
+      "loss": 4.1185,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14660536189622e-05,
+      "loss": 4.1289,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145766767145168e-05,
+      "loss": 4.1283,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144928172394116e-05,
+      "loss": 4.1198,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144091215523437e-05,
+      "loss": 4.1082,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143252620772386e-05,
+      "loss": 4.1194,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142414026021334e-05,
+      "loss": 4.1131,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141575431270282e-05,
+      "loss": 4.1273,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14073683651923e-05,
+      "loss": 4.1179,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1398998796485506e-05,
+      "loss": 4.1311,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390612848974986e-05,
+      "loss": 4.1115,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382226901464466e-05,
+      "loss": 4.1267,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373840953953946e-05,
+      "loss": 4.1265,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136545500644342e-05,
+      "loss": 4.1305,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13570690589329e-05,
+      "loss": 4.1246,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134868311142238e-05,
+      "loss": 4.1245,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340313542715595e-05,
+      "loss": 4.1165,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331927595205075e-05,
+      "loss": 4.1325,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323541647694555e-05,
+      "loss": 4.1105,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315155700184035e-05,
+      "loss": 4.127,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306769752673515e-05,
+      "loss": 4.1154,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298383805162995e-05,
+      "loss": 4.1267,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1289997857652475e-05,
+      "loss": 4.1204,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281611910141955e-05,
+      "loss": 4.1172,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1273242341435164e-05,
+      "loss": 4.1198,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126487277272837e-05,
+      "loss": 4.1099,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125648682521785e-05,
+      "loss": 4.1214,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.126465320587158,
+      "eval_runtime": 297.8635,
+      "eval_samples_per_second": 1281.093,
+      "eval_steps_per_second": 40.035,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124811725651106e-05,
+      "loss": 4.1132,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123973130900054e-05,
+      "loss": 4.1062,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123134536149003e-05,
+      "loss": 4.1251,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122295941397951e-05,
+      "loss": 4.1189,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121457346646899e-05,
+      "loss": 4.1267,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120618751895847e-05,
+      "loss": 4.1169,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.1151,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1113,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1111,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117264372891639e-05,
+      "loss": 4.123,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116425778140587e-05,
+      "loss": 4.1155,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.117,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114751864399229e-05,
+      "loss": 4.116,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113913269648177e-05,
+      "loss": 4.1073,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1130746748971247e-05,
+      "loss": 4.1021,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122360801460727e-05,
+      "loss": 4.104,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111397485395021e-05,
+      "loss": 4.1224,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110558890643969e-05,
+      "loss": 4.1166,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1109,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1274,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108044744271186e-05,
+      "loss": 4.126,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107206149520134e-05,
+      "loss": 4.1138,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106369192649455e-05,
+      "loss": 4.1151,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105530597898403e-05,
+      "loss": 4.1087,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104693641027724e-05,
+      "loss": 4.1178,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103855046276672e-05,
+      "loss": 4.1136,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10301645152562e-05,
+      "loss": 4.1113,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102177856774568e-05,
+      "loss": 4.1131,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101339262023517e-05,
+      "loss": 4.1049,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100500667272465e-05,
+      "loss": 4.1045,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099662072521413e-05,
+      "loss": 4.0987,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098823477770361e-05,
+      "loss": 4.1133,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097984883019308e-05,
+      "loss": 4.1188,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097146288268256e-05,
+      "loss": 4.1163,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096307693517204e-05,
+      "loss": 4.1076,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095469098766152e-05,
+      "loss": 4.1047,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094632141895473e-05,
+      "loss": 4.1144,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093793547144421e-05,
+      "loss": 4.1,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092954952393369e-05,
+      "loss": 4.1003,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092116357642317e-05,
+      "loss": 4.1054,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091277762891265e-05,
+      "loss": 4.0978,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904391681402136e-05,
+      "loss": 4.11,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896005733891615e-05,
+      "loss": 4.1029,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0887619786381095e-05,
+      "loss": 4.1109,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879250217674305e-05,
+      "loss": 4.1041,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0870864270163784e-05,
+      "loss": 4.1078,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862494701456994e-05,
+      "loss": 4.1163,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854108753946474e-05,
+      "loss": 4.1011,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845722806435953e-05,
+      "loss": 4.0748,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083735323772916e-05,
+      "loss": 4.1202,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082896729021864e-05,
+      "loss": 4.1014,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082058134270812e-05,
+      "loss": 4.1101,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08121953951976e-05,
+      "loss": 4.0999,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080380944768708e-05,
+      "loss": 4.1007,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07954398789803e-05,
+      "loss": 4.0909,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078705393146978e-05,
+      "loss": 4.1019,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077866798395926e-05,
+      "loss": 4.1117,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077029841525247e-05,
+      "loss": 4.0918,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076191246774195e-05,
+      "loss": 4.1171,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075352652023143e-05,
+      "loss": 4.0975,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074514057272091e-05,
+      "loss": 4.0848,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073675462521039e-05,
+      "loss": 4.1001,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072836867769987e-05,
+      "loss": 4.0884,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071998273018935e-05,
+      "loss": 4.0838,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071159678267883e-05,
+      "loss": 4.1076,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070321083516831e-05,
+      "loss": 4.1044,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069482488765779e-05,
+      "loss": 4.0959,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068643894014727e-05,
+      "loss": 4.1028,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067805299263675e-05,
+      "loss": 4.0824,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066968342392996e-05,
+      "loss": 4.1031,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066129747641944e-05,
+      "loss": 4.0956,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0652911528908916e-05,
+      "loss": 4.1021,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644525581398396e-05,
+      "loss": 4.1051,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636139633887876e-05,
+      "loss": 4.1006,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627753686377356e-05,
+      "loss": 4.1201,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619367738866836e-05,
+      "loss": 4.0888,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0610981791356316e-05,
+      "loss": 4.1042,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602595843845796e-05,
+      "loss": 4.0982,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594209896335276e-05,
+      "loss": 4.0909,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058584032762849e-05,
+      "loss": 4.1057,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057745438011797e-05,
+      "loss": 4.1115,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056906843260745e-05,
+      "loss": 4.1027,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056068248509693e-05,
+      "loss": 4.1182,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055229653758641e-05,
+      "loss": 4.0895,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054391059007589e-05,
+      "loss": 4.0901,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053552464256537e-05,
+      "loss": 4.0965,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052715507385858e-05,
+      "loss": 4.1081,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051876912634806e-05,
+      "loss": 4.095,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051039955764127e-05,
+      "loss": 4.1039,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050201361013075e-05,
+      "loss": 4.092,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049362766262023e-05,
+      "loss": 4.1033,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048524171510971e-05,
+      "loss": 4.0945,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047685576759919e-05,
+      "loss": 4.0863,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468469820088676e-05,
+      "loss": 4.0984,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460083872578156e-05,
+      "loss": 4.1088,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045169792506763e-05,
+      "loss": 4.0975,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044331197755711e-05,
+      "loss": 4.0805,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043492603004659e-05,
+      "loss": 4.0924,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04265564613398e-05,
+      "loss": 4.0929,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041817051382928e-05,
+      "loss": 4.1024,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040978456631876e-05,
+      "loss": 4.0988,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040139861880824e-05,
+      "loss": 4.089,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039302905010145e-05,
+      "loss": 4.0881,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038464310259093e-05,
+      "loss": 4.083,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376257155080414e-05,
+      "loss": 4.0919,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0367871207569894e-05,
+      "loss": 4.0878,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359485260059374e-05,
+      "loss": 4.0784,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351099312548853e-05,
+      "loss": 4.1024,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342713365038333e-05,
+      "loss": 4.0922,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033432741752781e-05,
+      "loss": 4.0887,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032594147001729e-05,
+      "loss": 4.0841,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031755552250677e-05,
+      "loss": 4.0889,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030918595379998e-05,
+      "loss": 4.0925,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030080000628946e-05,
+      "loss": 4.0849,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029243043758267e-05,
+      "loss": 4.101,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028404449007215e-05,
+      "loss": 4.1006,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.0905,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026728897385485e-05,
+      "loss": 4.0976,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025890302634433e-05,
+      "loss": 4.085,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025051707883381e-05,
+      "loss": 4.1,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024213113132329e-05,
+      "loss": 4.0856,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023374518381277e-05,
+      "loss": 4.0947,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022535923630225e-05,
+      "loss": 4.0889,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021697328879173e-05,
+      "loss": 4.098,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020858734128121e-05,
+      "loss": 4.0961,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020020139377069e-05,
+      "loss": 4.0848,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019181544626016e-05,
+      "loss": 4.0693,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0183445877553376e-05,
+      "loss": 4.0917,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175059930042856e-05,
+      "loss": 4.0807,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0166673982532336e-05,
+      "loss": 4.0921,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0158288035021816e-05,
+      "loss": 4.087,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0149902087511296e-05,
+      "loss": 4.0974,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0141516140000776e-05,
+      "loss": 4.0776,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133130192490256e-05,
+      "loss": 4.0967,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0124744244979736e-05,
+      "loss": 4.0918,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0116374676272945e-05,
+      "loss": 4.1014,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0107988728762425e-05,
+      "loss": 4.0918,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0099602781251905e-05,
+      "loss": 4.0882,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091216833741385e-05,
+      "loss": 4.087,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0082830886230865e-05,
+      "loss": 4.0995,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007447769632781e-05,
+      "loss": 4.0764,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006609174881728e-05,
+      "loss": 4.0935,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005770580130677e-05,
+      "loss": 4.089,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0049336232599985e-05,
+      "loss": 4.0898,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0040950285089465e-05,
+      "loss": 4.0875,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003256433757894e-05,
+      "loss": 4.0879,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002417839006842e-05,
+      "loss": 4.0869,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00157924425579e-05,
+      "loss": 4.0796,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000740649504738e-05,
+      "loss": 4.087,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.102380275726318,
+      "eval_runtime": 296.6137,
+      "eval_samples_per_second": 1286.492,
+      "eval_steps_per_second": 40.204,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999902054753686e-05,
+      "loss": 4.0779,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999063460002634e-05,
+      "loss": 4.0767,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998224865251582e-05,
+      "loss": 4.0917,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99738627050053e-05,
+      "loss": 4.0867,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996547675749478e-05,
+      "loss": 4.1002,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995709080998426e-05,
+      "loss": 4.0823,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994870486247374e-05,
+      "loss": 4.0859,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994031891496322e-05,
+      "loss": 4.0823,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99319329674527e-05,
+      "loss": 4.0774,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992354701994218e-05,
+      "loss": 4.0945,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991516107243166e-05,
+      "loss": 4.0822,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990677512492114e-05,
+      "loss": 4.0845,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989840555621435e-05,
+      "loss": 4.0921,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989001960870383e-05,
+      "loss": 4.0691,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988163366119331e-05,
+      "loss": 4.0722,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987324771368279e-05,
+      "loss": 4.0712,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986486176617227e-05,
+      "loss": 4.0948,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985647581866175e-05,
+      "loss": 4.0822,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984808987115123e-05,
+      "loss": 4.0818,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839703923640707e-05,
+      "loss": 4.0942,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831317976130187e-05,
+      "loss": 4.0943,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9822932028619666e-05,
+      "loss": 4.0808,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981456245991288e-05,
+      "loss": 4.0858,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980617651240236e-05,
+      "loss": 4.0806,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979779056489184e-05,
+      "loss": 4.0853,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978940461738132e-05,
+      "loss": 4.0804,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97810186698708e-05,
+      "loss": 4.0774,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977263272236028e-05,
+      "loss": 4.0867,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976426315365349e-05,
+      "loss": 4.0736,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975587720614297e-05,
+      "loss": 4.0735,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974749125863245e-05,
+      "loss": 4.0717,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973912168992566e-05,
+      "loss": 4.0754,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973073574241514e-05,
+      "loss": 4.084,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972234979490462e-05,
+      "loss": 4.0919,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97139638473941e-05,
+      "loss": 4.0755,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705594278687316e-05,
+      "loss": 4.077,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697208331176796e-05,
+      "loss": 4.0869,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688822383666276e-05,
+      "loss": 4.0636,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680436436155756e-05,
+      "loss": 4.0708,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672050488645236e-05,
+      "loss": 4.0789,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966366454113471e-05,
+      "loss": 4.0653,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655294972427925e-05,
+      "loss": 4.0769,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9646909024917405e-05,
+      "loss": 4.0746,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638523077406885e-05,
+      "loss": 4.0823,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963013712989636e-05,
+      "loss": 4.0722,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962175118238584e-05,
+      "loss": 4.0803,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613365234875325e-05,
+      "loss": 4.0835,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9604979287364805e-05,
+      "loss": 4.0728,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596593339854285e-05,
+      "loss": 4.0471,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588207392343765e-05,
+      "loss": 4.0866,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9579821444833244e-05,
+      "loss": 4.0708,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571435497322724e-05,
+      "loss": 4.0801,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9563049549812204e-05,
+      "loss": 4.0708,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554663602301684e-05,
+      "loss": 4.067,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9546294033594893e-05,
+      "loss": 4.0676,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95379244648881e-05,
+      "loss": 4.0689,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952953851737758e-05,
+      "loss": 4.0837,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952115256986706e-05,
+      "loss": 4.0579,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951276662235654e-05,
+      "loss": 4.0898,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950438067484602e-05,
+      "loss": 4.0693,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949599472733551e-05,
+      "loss": 4.0529,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948760877982499e-05,
+      "loss": 4.0711,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947922283231447e-05,
+      "loss": 4.0598,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947085326360768e-05,
+      "loss": 4.0568,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946246731609716e-05,
+      "loss": 4.0751,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945408136858664e-05,
+      "loss": 4.0767,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944569542107612e-05,
+      "loss": 4.0639,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437342231173056e-05,
+      "loss": 4.0761,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9428956283662536e-05,
+      "loss": 4.0509,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420570336152016e-05,
+      "loss": 4.0724,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412184388641496e-05,
+      "loss": 4.0722,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403798441130976e-05,
+      "loss": 4.0707,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939542887242419e-05,
+      "loss": 4.0761,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938704292491367e-05,
+      "loss": 4.0696,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937865697740315e-05,
+      "loss": 4.0929,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937027102989263e-05,
+      "loss": 4.0594,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936188508238211e-05,
+      "loss": 4.0752,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935349913487159e-05,
+      "loss": 4.0672,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93451295661648e-05,
+      "loss": 4.0633,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933674361865428e-05,
+      "loss": 4.0769,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932835767114376e-05,
+      "loss": 4.077,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931997172363324e-05,
+      "loss": 4.0774,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931158577612272e-05,
+      "loss": 4.0891,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930321620741593e-05,
+      "loss": 4.0635,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929483025990541e-05,
+      "loss": 4.0598,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286444312394896e-05,
+      "loss": 4.0676,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927805836488437e-05,
+      "loss": 4.0798,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926967241737385e-05,
+      "loss": 4.0642,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926128646986333e-05,
+      "loss": 4.0729,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925290052235281e-05,
+      "loss": 4.0695,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924451457484229e-05,
+      "loss": 4.0708,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92361450061355e-05,
+      "loss": 4.0698,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922775905862498e-05,
+      "loss": 4.0561,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921937311111446e-05,
+      "loss": 4.0718,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921098716360394e-05,
+      "loss": 4.0781,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920261759489715e-05,
+      "loss": 4.0713,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194231647386634e-05,
+      "loss": 4.0561,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9185845699876114e-05,
+      "loss": 4.0639,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9177459752365594e-05,
+      "loss": 4.0633,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169073804855074e-05,
+      "loss": 4.0717,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9160687857344554e-05,
+      "loss": 4.0732,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152301909834034e-05,
+      "loss": 4.0584,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9143915962323514e-05,
+      "loss": 4.0628,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913554639361672e-05,
+      "loss": 4.0556,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91271604461062e-05,
+      "loss": 4.0633,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911877449859568e-05,
+      "loss": 4.0616,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911038855108516e-05,
+      "loss": 4.0544,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910201898237837e-05,
+      "loss": 4.0712,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909363303486785e-05,
+      "loss": 4.0623,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908526346616107e-05,
+      "loss": 4.0624,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907687751865055e-05,
+      "loss": 4.0529,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906849157114003e-05,
+      "loss": 4.0615,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906010562362951e-05,
+      "loss": 4.0661,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905171967611899e-05,
+      "loss": 4.0546,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904333372860847e-05,
+      "loss": 4.072,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903494778109795e-05,
+      "loss": 4.07,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902656183358743e-05,
+      "loss": 4.0627,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018208643684366e-05,
+      "loss": 4.0687,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009822696173846e-05,
+      "loss": 4.0559,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001436748663326e-05,
+      "loss": 4.071,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8993050801152806e-05,
+      "loss": 4.0612,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8984664853642286e-05,
+      "loss": 4.0664,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89762952849355e-05,
+      "loss": 4.0611,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896792571622871e-05,
+      "loss": 4.0694,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895953976871819e-05,
+      "loss": 4.0714,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895115382120767e-05,
+      "loss": 4.0549,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894276787369715e-05,
+      "loss": 4.0469,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893438192618663e-05,
+      "loss": 4.058,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892599597867611e-05,
+      "loss": 4.0556,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891761003116559e-05,
+      "loss": 4.0653,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890922408365507e-05,
+      "loss": 4.0585,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890083813614455e-05,
+      "loss": 4.068,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889246856743776e-05,
+      "loss": 4.0529,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888408261992724e-05,
+      "loss": 4.0773,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875696672416726e-05,
+      "loss": 4.0627,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867310724906206e-05,
+      "loss": 4.0702,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885892477739568e-05,
+      "loss": 4.0657,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885053882988516e-05,
+      "loss": 4.0601,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884215288237464e-05,
+      "loss": 4.0603,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883376693486412e-05,
+      "loss": 4.0721,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882539736615733e-05,
+      "loss": 4.0504,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817027797450544e-05,
+      "loss": 4.0669,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8808641849940024e-05,
+      "loss": 4.0606,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88002559024295e-05,
+      "loss": 4.064,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879186995491898e-05,
+      "loss": 4.0591,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783484007408464e-05,
+      "loss": 4.0594,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775098059897944e-05,
+      "loss": 4.0594,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876672849119115e-05,
+      "loss": 4.0601,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875834254368063e-05,
+      "loss": 4.0579,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.083654880523682,
+      "eval_runtime": 296.0701,
+      "eval_samples_per_second": 1288.854,
+      "eval_steps_per_second": 40.278,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.8260667339428506e+17,
+  "trial_name": null,
+  "trial_params": null
+}