End of training

README.md

@@ -4,6 +4,7 @@ license: apache-2.0
 base_model: Qwen/Qwen2.5-7B-Instruct
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: b2_code_askllm_10k
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # b2_code_askllm_10k
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/b2_code_askllm_10k dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.992,
+    "total_flos": 1.9434542678430188e+18,
+    "train_loss": 0.5044988004060892,
+    "train_runtime": 62031.5499,
+    "train_samples_per_second": 0.806,
+    "train_steps_per_second": 0.006
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.992,
+    "total_flos": 1.9434542678430188e+18,
+    "train_loss": 0.5044988004060892,
+    "train_runtime": 62031.5499,
+    "train_samples_per_second": 0.806,
+    "train_steps_per_second": 0.006
+}

trainer_state.json

@@ -0,0 +1,2772 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.992,
+  "eval_steps": 500,
+  "global_step": 390,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0128,
+      "grad_norm": 6.440505180668678,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.0338,
+      "step": 1
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 6.360134627939838,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.0112,
+      "step": 2
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.268412021121898,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0222,
+      "step": 3
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 5.816218990274612,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.0173,
+      "step": 4
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.284063377996134,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.9658,
+      "step": 5
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.6899860465527485,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9648,
+      "step": 6
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.496586497267657,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.9224,
+      "step": 7
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 4.3506016896731605,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.9467,
+      "step": 8
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 4.1490515738724145,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.9192,
+      "step": 9
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 3.948953116030095,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.8814,
+      "step": 10
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 3.5834213792492458,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.8726,
+      "step": 11
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 2.374696347338207,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.8524,
+      "step": 12
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.7200259478397855,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8208,
+      "step": 13
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 1.6523655711025134,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.7982,
+      "step": 14
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.3117464345799783,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.776,
+      "step": 15
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.0872725354418016,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.7718,
+      "step": 16
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.2630060897661322,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.7649,
+      "step": 17
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.1285511450226429,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.7453,
+      "step": 18
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.8713404568210159,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.7418,
+      "step": 19
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.1022623507071383,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.7295,
+      "step": 20
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.1636613532977724,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.719,
+      "step": 21
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.9325043449885487,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.7254,
+      "step": 22
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.090542169821071,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.7139,
+      "step": 23
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.842549769847692,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.7238,
+      "step": 24
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7114948348319531,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.705,
+      "step": 25
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.8921820771837474,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7078,
+      "step": 26
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.7921749112393623,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.7068,
+      "step": 27
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.7644141503561558,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.6914,
+      "step": 28
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.9144509766520436,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.6792,
+      "step": 29
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.1055713469684516,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6912,
+      "step": 30
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.7849781574172187,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.6774,
+      "step": 31
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 1.1009807222432384,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.6574,
+      "step": 32
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 1.1657586549517602,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6836,
+      "step": 33
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 1.5020689377881666,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.6653,
+      "step": 34
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.8086964944354174,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.6723,
+      "step": 35
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 2.087558717206756,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.665,
+      "step": 36
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 1.3739438121951246,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.6789,
+      "step": 37
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 1.9589960994697235,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6778,
+      "step": 38
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 1.7197772072895912,
+      "learning_rate": 4e-05,
+      "loss": 0.6754,
+      "step": 39
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.239516859482234,
+      "learning_rate": 3.9999198907597046e-05,
+      "loss": 0.6605,
+      "step": 40
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 1.3222263492108368,
+      "learning_rate": 3.9996795694563096e-05,
+      "loss": 0.665,
+      "step": 41
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.8974468069466353,
+      "learning_rate": 3.999279055341771e-05,
+      "loss": 0.6615,
+      "step": 42
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 1.2420342798281085,
+      "learning_rate": 3.998718380500971e-05,
+      "loss": 0.6587,
+      "step": 43
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 1.226235999609397,
+      "learning_rate": 3.997997589849145e-05,
+      "loss": 0.6604,
+      "step": 44
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.1133987875860607,
+      "learning_rate": 3.9971167411282835e-05,
+      "loss": 0.6563,
+      "step": 45
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.9041885925772853,
+      "learning_rate": 3.99607590490251e-05,
+      "loss": 0.6514,
+      "step": 46
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.8949480429440133,
+      "learning_rate": 3.9948751645524235e-05,
+      "loss": 0.6626,
+      "step": 47
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.318063035845643,
+      "learning_rate": 3.9935146162684206e-05,
+      "loss": 0.6654,
+      "step": 48
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.8046633597729084,
+      "learning_rate": 3.9919943690429906e-05,
+      "loss": 0.655,
+      "step": 49
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6058287067824193,
+      "learning_rate": 3.9903145446619837e-05,
+      "loss": 0.6569,
+      "step": 50
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.9533893187645094,
+      "learning_rate": 3.9884752776948564e-05,
+      "loss": 0.6478,
+      "step": 51
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 1.7417364834918179,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6635,
+      "step": 52
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 1.2249766521333794,
+      "learning_rate": 3.9843190181323744e-05,
+      "loss": 0.6484,
+      "step": 53
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 1.5359824771240203,
+      "learning_rate": 3.982002358491817e-05,
+      "loss": 0.6655,
+      "step": 54
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.3606419207755613,
+      "learning_rate": 3.979526922148058e-05,
+      "loss": 0.6576,
+      "step": 55
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 1.133037884656615,
+      "learning_rate": 3.9768929074064206e-05,
+      "loss": 0.6355,
+      "step": 56
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 1.2625950366285374,
+      "learning_rate": 3.9741005252758255e-05,
+      "loss": 0.6432,
+      "step": 57
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.9892824359096428,
+      "learning_rate": 3.971149999451886e-05,
+      "loss": 0.6465,
+      "step": 58
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 1.3290509959496604,
+      "learning_rate": 3.9680415662989806e-05,
+      "loss": 0.6469,
+      "step": 59
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.9544676550042052,
+      "learning_rate": 3.9647754748313294e-05,
+      "loss": 0.6261,
+      "step": 60
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 1.3783376727212315,
+      "learning_rate": 3.96135198669304e-05,
+      "loss": 0.6357,
+      "step": 61
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.9440358615212426,
+      "learning_rate": 3.957771376137144e-05,
+      "loss": 0.6338,
+      "step": 62
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.9803709039155389,
+      "learning_rate": 3.954033930003634e-05,
+      "loss": 0.6286,
+      "step": 63
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 1.0995516160551828,
+      "learning_rate": 3.9501399476964806e-05,
+      "loss": 0.6408,
+      "step": 64
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.06604572631698,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6317,
+      "step": 65
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.9553425734859959,
+      "learning_rate": 3.9418836348521045e-05,
+      "loss": 0.6345,
+      "step": 66
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.9344995799040006,
+      "learning_rate": 3.937521965721831e-05,
+      "loss": 0.6347,
+      "step": 67
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.7717967370748647,
+      "learning_rate": 3.933005083178828e-05,
+      "loss": 0.6277,
+      "step": 68
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.8845532815632384,
+      "learning_rate": 3.928333349067125e-05,
+      "loss": 0.6333,
+      "step": 69
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.7811593435671148,
+      "learning_rate": 3.923507137635792e-05,
+      "loss": 0.6385,
+      "step": 70
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.6722600324453193,
+      "learning_rate": 3.9185268355089606e-05,
+      "loss": 0.6363,
+      "step": 71
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.7013843628196784,
+      "learning_rate": 3.913392841654851e-05,
+      "loss": 0.6182,
+      "step": 72
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.5241271397222523,
+      "learning_rate": 3.9081055673538093e-05,
+      "loss": 0.6286,
+      "step": 73
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.8378711686250688,
+      "learning_rate": 3.902665436165364e-05,
+      "loss": 0.6262,
+      "step": 74
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5700616630534898,
+      "learning_rate": 3.897072883894291e-05,
+      "loss": 0.6247,
+      "step": 75
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.707717496346406,
+      "learning_rate": 3.8913283585557054e-05,
+      "loss": 0.6331,
+      "step": 76
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.4854149491841238,
+      "learning_rate": 3.885432320339167e-05,
+      "loss": 0.622,
+      "step": 77
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.7025692117424397,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6358,
+      "step": 78
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 0.7476411190441912,
+      "learning_rate": 3.873187606680543e-05,
+      "loss": 0.5851,
+      "step": 79
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.5354881944955312,
+      "learning_rate": 3.866839912153168e-05,
+      "loss": 0.5616,
+      "step": 80
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.72979259702871,
+      "learning_rate": 3.860342666498677e-05,
+      "loss": 0.5687,
+      "step": 81
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.9107316631547568,
+      "learning_rate": 3.853696390206484e-05,
+      "loss": 0.5701,
+      "step": 82
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.8142918955825038,
+      "learning_rate": 3.846901615704734e-05,
+      "loss": 0.5783,
+      "step": 83
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.9013571744948835,
+      "learning_rate": 3.839958887317649e-05,
+      "loss": 0.5834,
+      "step": 84
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 1.0062069408220184,
+      "learning_rate": 3.832868761221926e-05,
+      "loss": 0.5719,
+      "step": 85
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.8228424564002687,
+      "learning_rate": 3.825631805402182e-05,
+      "loss": 0.5642,
+      "step": 86
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.8592927492803276,
+      "learning_rate": 3.818248599605448e-05,
+      "loss": 0.5876,
+      "step": 87
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.9406399419576136,
+      "learning_rate": 3.810719735294731e-05,
+      "loss": 0.5723,
+      "step": 88
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.6830298227783856,
+      "learning_rate": 3.8030458156016326e-05,
+      "loss": 0.5637,
+      "step": 89
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.7339719088666201,
+      "learning_rate": 3.795227455278029e-05,
+      "loss": 0.5847,
+      "step": 90
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.6629114611083571,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5819,
+      "step": 91
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.5708712660332336,
+      "learning_rate": 3.7791599295517825e-05,
+      "loss": 0.5704,
+      "step": 92
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.6220784047435565,
+      "learning_rate": 3.7709120513064196e-05,
+      "loss": 0.5719,
+      "step": 93
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.7301067427856471,
+      "learning_rate": 3.762522306641998e-05,
+      "loss": 0.582,
+      "step": 94
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.833057664299409,
+      "learning_rate": 3.7539913676545874e-05,
+      "loss": 0.5842,
+      "step": 95
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.9379034745119356,
+      "learning_rate": 3.745319917751229e-05,
+      "loss": 0.5704,
+      "step": 96
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.8127571638061729,
+      "learning_rate": 3.736508651595188e-05,
+      "loss": 0.5653,
+      "step": 97
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.6766945990885098,
+      "learning_rate": 3.727558275050301e-05,
+      "loss": 0.5633,
+      "step": 98
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.6430573461612847,
+      "learning_rate": 3.718469505124434e-05,
+      "loss": 0.5606,
+      "step": 99
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6753115266892358,
+      "learning_rate": 3.709243069912041e-05,
+      "loss": 0.5661,
+      "step": 100
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.8345931599971936,
+      "learning_rate": 3.699879708535838e-05,
+      "loss": 0.5702,
+      "step": 101
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.617803588936545,
+      "learning_rate": 3.69038017108759e-05,
+      "loss": 0.5608,
+      "step": 102
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.5088020067540242,
+      "learning_rate": 3.680745218568026e-05,
+      "loss": 0.565,
+      "step": 103
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.7276172035472214,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5633,
+      "step": 104
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.7328563107909479,
+      "learning_rate": 3.6610721664960236e-05,
+      "loss": 0.5469,
+      "step": 105
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.6484190996796542,
+      "learning_rate": 3.65103564293684e-05,
+      "loss": 0.5624,
+      "step": 106
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.5845316546941463,
+      "learning_rate": 3.640866856166601e-05,
+      "loss": 0.5712,
+      "step": 107
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.7680563077284489,
+      "learning_rate": 3.6305666207990886e-05,
+      "loss": 0.5611,
+      "step": 108
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.5730659339685035,
+      "learning_rate": 3.6201357619783336e-05,
+      "loss": 0.577,
+      "step": 109
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.4851095532004884,
+      "learning_rate": 3.609575115312511e-05,
+      "loss": 0.5556,
+      "step": 110
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.7826442580287045,
+      "learning_rate": 3.598885526807003e-05,
+      "loss": 0.578,
+      "step": 111
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.5403543360554025,
+      "learning_rate": 3.5880678527966224e-05,
+      "loss": 0.5543,
+      "step": 112
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.520450735369304,
+      "learning_rate": 3.577122959877017e-05,
+      "loss": 0.5428,
+      "step": 113
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.6730691244877828,
+      "learning_rate": 3.566051724835245e-05,
+      "loss": 0.5707,
+      "step": 114
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.4482581465131313,
+      "learning_rate": 3.554855034579532e-05,
+      "loss": 0.5685,
+      "step": 115
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.4674208810680886,
+      "learning_rate": 3.5435337860682304e-05,
+      "loss": 0.5642,
+      "step": 116
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.45402044921252604,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5645,
+      "step": 117
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.4927930274515102,
+      "learning_rate": 3.520521251930941e-05,
+      "loss": 0.5542,
+      "step": 118
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.4501428271386669,
+      "learning_rate": 3.5088318098215805e-05,
+      "loss": 0.568,
+      "step": 119
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.4896780844891411,
+      "learning_rate": 3.497021496342203e-05,
+      "loss": 0.5517,
+      "step": 120
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.4949783954604816,
+      "learning_rate": 3.485091257608047e-05,
+      "loss": 0.5609,
+      "step": 121
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.43184064258446964,
+      "learning_rate": 3.473042049341474e-05,
+      "loss": 0.5671,
+      "step": 122
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.46773313003774414,
+      "learning_rate": 3.4608748367954064e-05,
+      "loss": 0.5571,
+      "step": 123
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.4801033155074372,
+      "learning_rate": 3.4485905946759965e-05,
+      "loss": 0.5585,
+      "step": 124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.48761484953567685,
+      "learning_rate": 3.4361903070645484e-05,
+      "loss": 0.578,
+      "step": 125
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.5183931155891501,
+      "learning_rate": 3.423674967338681e-05,
+      "loss": 0.5709,
+      "step": 126
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.5027240920465532,
+      "learning_rate": 3.411045578092754e-05,
+      "loss": 0.5458,
+      "step": 127
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.3668021582009626,
+      "learning_rate": 3.398303151057543e-05,
+      "loss": 0.5604,
+      "step": 128
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.5293360802932268,
+      "learning_rate": 3.385448707019199e-05,
+      "loss": 0.5706,
+      "step": 129
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.5067359991418402,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5613,
+      "step": 130
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.43956211682669255,
+      "learning_rate": 3.359407895863199e-05,
+      "loss": 0.5426,
+      "step": 131
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.43321820890624846,
+      "learning_rate": 3.34622361485514e-05,
+      "loss": 0.5714,
+      "step": 132
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.5506855614494184,
+      "learning_rate": 3.332931488896029e-05,
+      "loss": 0.5663,
+      "step": 133
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.40674075771239687,
+      "learning_rate": 3.319532582807977e-05,
+      "loss": 0.5553,
+      "step": 134
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.5942652931118136,
+      "learning_rate": 3.30602796996717e-05,
+      "loss": 0.5581,
+      "step": 135
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.5165956517595004,
+      "learning_rate": 3.2924187322178865e-05,
+      "loss": 0.5573,
+      "step": 136
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.46259566037191663,
+      "learning_rate": 3.278705959785821e-05,
+      "loss": 0.5543,
+      "step": 137
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.5222546408351808,
+      "learning_rate": 3.2648907511907544e-05,
+      "loss": 0.5644,
+      "step": 138
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.377430226314278,
+      "learning_rate": 3.250974213158555e-05,
+      "loss": 0.5534,
+      "step": 139
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.557254631898166,
+      "learning_rate": 3.23695746053251e-05,
+      "loss": 0.562,
+      "step": 140
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.4428053343917969,
+      "learning_rate": 3.222841616184025e-05,
+      "loss": 0.5634,
+      "step": 141
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.507424711580751,
+      "learning_rate": 3.208627810922665e-05,
+      "loss": 0.5615,
+      "step": 142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.48740719950313904,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5622,
+      "step": 143
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.5520350720623239,
+      "learning_rate": 3.1799108800462466e-05,
+      "loss": 0.5644,
+      "step": 144
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.4470006220654049,
+      "learning_rate": 3.1654100549227024e-05,
+      "loss": 0.5569,
+      "step": 145
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.5049011825096722,
+      "learning_rate": 3.1508158696850275e-05,
+      "loss": 0.5583,
+      "step": 146
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.5024884916073381,
+      "learning_rate": 3.136129493462312e-05,
+      "loss": 0.5604,
+      "step": 147
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.42965423511043344,
+      "learning_rate": 3.121352102768998e-05,
+      "loss": 0.5681,
+      "step": 148
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.43628178509966303,
+      "learning_rate": 3.106484881410628e-05,
+      "loss": 0.5561,
+      "step": 149
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.516672384401874,
+      "learning_rate": 3.091529020389009e-05,
+      "loss": 0.5612,
+      "step": 150
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.3971572803988885,
+      "learning_rate": 3.076485717806808e-05,
+      "loss": 0.5479,
+      "step": 151
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.5219460456119351,
+      "learning_rate": 3.061356178771564e-05,
+      "loss": 0.5511,
+      "step": 152
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.5441247836401949,
+      "learning_rate": 3.0461416152991555e-05,
+      "loss": 0.5478,
+      "step": 153
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.42158079374127294,
+      "learning_rate": 3.0308432462167045e-05,
+      "loss": 0.556,
+      "step": 154
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.3710818016637592,
+      "learning_rate": 3.015462297064936e-05,
+      "loss": 0.5622,
+      "step": 155
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.39287859604255465,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5552,
+      "step": 156
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 0.4810957487215277,
+      "learning_rate": 2.98445759369477e-05,
+      "loss": 0.5019,
+      "step": 157
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.4696753127217347,
+      "learning_rate": 2.9688363232396056e-05,
+      "loss": 0.4898,
+      "step": 158
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.559677168390859,
+      "learning_rate": 2.9531374400426158e-05,
+      "loss": 0.4841,
+      "step": 159
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.6879445150427815,
+      "learning_rate": 2.9373622017294075e-05,
+      "loss": 0.4749,
+      "step": 160
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.5741929875240579,
+      "learning_rate": 2.9215118720423375e-05,
+      "loss": 0.4689,
+      "step": 161
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.5702999528092169,
+      "learning_rate": 2.9055877207392752e-05,
+      "loss": 0.4915,
+      "step": 162
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.5786948955417902,
+      "learning_rate": 2.8895910234918828e-05,
+      "loss": 0.4726,
+      "step": 163
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.5927881966274303,
+      "learning_rate": 2.873523061783426e-05,
+      "loss": 0.4683,
+      "step": 164
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.5356273330838037,
+      "learning_rate": 2.8573851228061084e-05,
+      "loss": 0.4553,
+      "step": 165
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.5196550141498971,
+      "learning_rate": 2.8411784993579633e-05,
+      "loss": 0.4763,
+      "step": 166
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.5484054422196126,
+      "learning_rate": 2.8249044897392814e-05,
+      "loss": 0.4885,
+      "step": 167
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.627010389991889,
+      "learning_rate": 2.80856439764861e-05,
+      "loss": 0.474,
+      "step": 168
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.5943064505781659,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4745,
+      "step": 169
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.5078520524905042,
+      "learning_rate": 2.77569120720971e-05,
+      "loss": 0.474,
+      "step": 170
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.5247617124108166,
+      "learning_rate": 2.7591607423077932e-05,
+      "loss": 0.4857,
+      "step": 171
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.4548412049154901,
+      "learning_rate": 2.7425694616155474e-05,
+      "loss": 0.477,
+      "step": 172
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.464235021826899,
+      "learning_rate": 2.7259186942478656e-05,
+      "loss": 0.4702,
+      "step": 173
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.536156033395784,
+      "learning_rate": 2.7092097740850712e-05,
+      "loss": 0.4765,
+      "step": 174
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.455714036052608,
+      "learning_rate": 2.692444039666066e-05,
+      "loss": 0.484,
+      "step": 175
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.4397770750259298,
+      "learning_rate": 2.6756228340810946e-05,
+      "loss": 0.4805,
+      "step": 176
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.4012588163930126,
+      "learning_rate": 2.6587475048641596e-05,
+      "loss": 0.4841,
+      "step": 177
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.44509143203605256,
+      "learning_rate": 2.6418194038850634e-05,
+      "loss": 0.4754,
+      "step": 178
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.3610143090508979,
+      "learning_rate": 2.624839887241115e-05,
+      "loss": 0.4822,
+      "step": 179
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.580899767456742,
+      "learning_rate": 2.607810315148494e-05,
+      "loss": 0.4665,
+      "step": 180
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.36906325472102786,
+      "learning_rate": 2.5907320518332827e-05,
+      "loss": 0.481,
+      "step": 181
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.49840724291179755,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.474,
+      "step": 182
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.33340363432650977,
+      "learning_rate": 2.5564349278329056e-05,
+      "loss": 0.4787,
+      "step": 183
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.4446905727425129,
+      "learning_rate": 2.539218814664288e-05,
+      "loss": 0.4748,
+      "step": 184
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.3456318247132697,
+      "learning_rate": 2.521959505086075e-05,
+      "loss": 0.4863,
+      "step": 185
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.3908786848365938,
+      "learning_rate": 2.5046583817284437e-05,
+      "loss": 0.4743,
+      "step": 186
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.37827442700520575,
+      "learning_rate": 2.487316830571244e-05,
+      "loss": 0.4881,
+      "step": 187
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.36474956731721675,
+      "learning_rate": 2.4699362408329646e-05,
+      "loss": 0.4755,
+      "step": 188
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 0.46148927737950757,
+      "learning_rate": 2.4525180048594452e-05,
+      "loss": 0.4833,
+      "step": 189
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.32648315431106817,
+      "learning_rate": 2.435063518012335e-05,
+      "loss": 0.4821,
+      "step": 190
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.43546951278327345,
+      "learning_rate": 2.4175741785573177e-05,
+      "loss": 0.4882,
+      "step": 191
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.3750762365022394,
+      "learning_rate": 2.4000513875520892e-05,
+      "loss": 0.488,
+      "step": 192
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.3568417525583788,
+      "learning_rate": 2.3824965487341247e-05,
+      "loss": 0.4744,
+      "step": 193
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.37221601465303467,
+      "learning_rate": 2.3649110684082258e-05,
+      "loss": 0.4804,
+      "step": 194
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.3204035329648721,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4864,
+      "step": 195
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.31257417198288145,
+      "learning_rate": 2.3296538206123134e-05,
+      "loss": 0.4644,
+      "step": 196
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.3121211904729013,
+      "learning_rate": 2.311984877573636e-05,
+      "loss": 0.4757,
+      "step": 197
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.36782619196771305,
+      "learning_rate": 2.2942909416634326e-05,
+      "loss": 0.4867,
+      "step": 198
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.3375621503827436,
+      "learning_rate": 2.2765734303294666e-05,
+      "loss": 0.471,
+      "step": 199
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3328945194503403,
+      "learning_rate": 2.2588337629081107e-05,
+      "loss": 0.466,
+      "step": 200
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.3474306821675776,
+      "learning_rate": 2.2410733605106462e-05,
+      "loss": 0.4871,
+      "step": 201
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.3082815992096003,
+      "learning_rate": 2.2232936459094158e-05,
+      "loss": 0.4832,
+      "step": 202
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.35202613044348846,
+      "learning_rate": 2.205496043423849e-05,
+      "loss": 0.4761,
+      "step": 203
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.3214647520518177,
+      "learning_rate": 2.1876819788063586e-05,
+      "loss": 0.4705,
+      "step": 204
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.36274690635166024,
+      "learning_rate": 2.16985287912813e-05,
+      "loss": 0.4784,
+      "step": 205
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.3812928113876961,
+      "learning_rate": 2.1520101726647922e-05,
+      "loss": 0.4778,
+      "step": 206
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.3671138052949733,
+      "learning_rate": 2.1341552887820048e-05,
+      "loss": 0.4959,
+      "step": 207
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.4081573643432729,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4887,
+      "step": 208
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.35479419412733626,
+      "learning_rate": 2.0984147109837564e-05,
+      "loss": 0.4793,
+      "step": 209
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.359314772773895,
+      "learning_rate": 2.0805318802188307e-05,
+      "loss": 0.4771,
+      "step": 210
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.32255143450933743,
+      "learning_rate": 2.0626425981061608e-05,
+      "loss": 0.4863,
+      "step": 211
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.36887422582218393,
+      "learning_rate": 2.0447482977425465e-05,
+      "loss": 0.4557,
+      "step": 212
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.3433586828942609,
+      "learning_rate": 2.0268504126267952e-05,
+      "loss": 0.463,
+      "step": 213
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.3725158685333721,
+      "learning_rate": 2.008950376544887e-05,
+      "loss": 0.4876,
+      "step": 214
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.3724465983324709,
+      "learning_rate": 1.9910496234551132e-05,
+      "loss": 0.4704,
+      "step": 215
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.3635263348421871,
+      "learning_rate": 1.9731495873732055e-05,
+      "loss": 0.4886,
+      "step": 216
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.34375654841143083,
+      "learning_rate": 1.9552517022574542e-05,
+      "loss": 0.4873,
+      "step": 217
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.40213447796532403,
+      "learning_rate": 1.93735740189384e-05,
+      "loss": 0.4664,
+      "step": 218
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.3191206042145243,
+      "learning_rate": 1.9194681197811703e-05,
+      "loss": 0.4828,
+      "step": 219
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.3586806902127841,
+      "learning_rate": 1.901585289016244e-05,
+      "loss": 0.4797,
+      "step": 220
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.3373025892719824,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4821,
+      "step": 221
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.3803205603470629,
+      "learning_rate": 1.8658447112179952e-05,
+      "loss": 0.4664,
+      "step": 222
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.3371932795198666,
+      "learning_rate": 1.8479898273352084e-05,
+      "loss": 0.4879,
+      "step": 223
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.38110756107404326,
+      "learning_rate": 1.83014712087187e-05,
+      "loss": 0.4612,
+      "step": 224
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3019413205986473,
+      "learning_rate": 1.8123180211936417e-05,
+      "loss": 0.4738,
+      "step": 225
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.3493524234310519,
+      "learning_rate": 1.794503956576152e-05,
+      "loss": 0.4712,
+      "step": 226
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.28037950243473425,
+      "learning_rate": 1.776706354090585e-05,
+      "loss": 0.4801,
+      "step": 227
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.3752528476673087,
+      "learning_rate": 1.758926639489354e-05,
+      "loss": 0.4769,
+      "step": 228
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.30040214185976755,
+      "learning_rate": 1.7411662370918893e-05,
+      "loss": 0.4785,
+      "step": 229
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.3394484343902263,
+      "learning_rate": 1.7234265696705344e-05,
+      "loss": 0.4772,
+      "step": 230
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.3116248684218825,
+      "learning_rate": 1.7057090583365678e-05,
+      "loss": 0.49,
+      "step": 231
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.33715778998035495,
+      "learning_rate": 1.6880151224263646e-05,
+      "loss": 0.4656,
+      "step": 232
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.32132598320200595,
+      "learning_rate": 1.6703461793876876e-05,
+      "loss": 0.4783,
+      "step": 233
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.3123433780390215,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4583,
+      "step": 234
+    },
+    {
+      "epoch": 3.008,
+      "grad_norm": 0.40270534460041507,
+      "learning_rate": 1.635088931591775e-05,
+      "loss": 0.4601,
+      "step": 235
+    },
+    {
+      "epoch": 3.0208,
+      "grad_norm": 0.3832052503211483,
+      "learning_rate": 1.6175034512658753e-05,
+      "loss": 0.3934,
+      "step": 236
+    },
+    {
+      "epoch": 3.0336,
+      "grad_norm": 0.8324835512594142,
+      "learning_rate": 1.5999486124479115e-05,
+      "loss": 0.4322,
+      "step": 237
+    },
+    {
+      "epoch": 3.0464,
+      "grad_norm": 0.45099281151716625,
+      "learning_rate": 1.5824258214426833e-05,
+      "loss": 0.4099,
+      "step": 238
+    },
+    {
+      "epoch": 3.0592,
+      "grad_norm": 0.4947330786718981,
+      "learning_rate": 1.5649364819876655e-05,
+      "loss": 0.4227,
+      "step": 239
+    },
+    {
+      "epoch": 3.072,
+      "grad_norm": 0.41140040631194025,
+      "learning_rate": 1.547481995140556e-05,
+      "loss": 0.4161,
+      "step": 240
+    },
+    {
+      "epoch": 3.0848,
+      "grad_norm": 0.39409332415502946,
+      "learning_rate": 1.5300637591670357e-05,
+      "loss": 0.3971,
+      "step": 241
+    },
+    {
+      "epoch": 3.0976,
+      "grad_norm": 0.5010026713099779,
+      "learning_rate": 1.5126831694287564e-05,
+      "loss": 0.4059,
+      "step": 242
+    },
+    {
+      "epoch": 3.1104,
+      "grad_norm": 0.38675629193171374,
+      "learning_rate": 1.4953416182715566e-05,
+      "loss": 0.3961,
+      "step": 243
+    },
+    {
+      "epoch": 3.1232,
+      "grad_norm": 0.4328614957549023,
+      "learning_rate": 1.478040494913926e-05,
+      "loss": 0.4089,
+      "step": 244
+    },
+    {
+      "epoch": 3.136,
+      "grad_norm": 0.39946159823496,
+      "learning_rate": 1.460781185335713e-05,
+      "loss": 0.4111,
+      "step": 245
+    },
+    {
+      "epoch": 3.1488,
+      "grad_norm": 0.4261990447434158,
+      "learning_rate": 1.443565072167095e-05,
+      "loss": 0.4184,
+      "step": 246
+    },
+    {
+      "epoch": 3.1616,
+      "grad_norm": 0.37975984472504815,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.4052,
+      "step": 247
+    },
+    {
+      "epoch": 3.1744,
+      "grad_norm": 0.437694596145041,
+      "learning_rate": 1.409267948166718e-05,
+      "loss": 0.4011,
+      "step": 248
+    },
+    {
+      "epoch": 3.1872,
+      "grad_norm": 0.35901948722786137,
+      "learning_rate": 1.3921896848515064e-05,
+      "loss": 0.401,
+      "step": 249
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3650305665036746,
+      "learning_rate": 1.3751601127588849e-05,
+      "loss": 0.4094,
+      "step": 250
+    },
+    {
+      "epoch": 3.2128,
+      "grad_norm": 0.38456740974263476,
+      "learning_rate": 1.3581805961149371e-05,
+      "loss": 0.4081,
+      "step": 251
+    },
+    {
+      "epoch": 3.2256,
+      "grad_norm": 0.35571023954259984,
+      "learning_rate": 1.341252495135841e-05,
+      "loss": 0.4071,
+      "step": 252
+    },
+    {
+      "epoch": 3.2384,
+      "grad_norm": 0.32402895653700925,
+      "learning_rate": 1.324377165918906e-05,
+      "loss": 0.428,
+      "step": 253
+    },
+    {
+      "epoch": 3.2512,
+      "grad_norm": 0.3439002016673635,
+      "learning_rate": 1.3075559603339354e-05,
+      "loss": 0.4055,
+      "step": 254
+    },
+    {
+      "epoch": 3.2640000000000002,
+      "grad_norm": 0.3662462529815356,
+      "learning_rate": 1.2907902259149287e-05,
+      "loss": 0.4157,
+      "step": 255
+    },
+    {
+      "epoch": 3.2768,
+      "grad_norm": 0.33452003926097973,
+      "learning_rate": 1.274081305752135e-05,
+      "loss": 0.3953,
+      "step": 256
+    },
+    {
+      "epoch": 3.2896,
+      "grad_norm": 0.35019843113943855,
+      "learning_rate": 1.2574305383844528e-05,
+      "loss": 0.4133,
+      "step": 257
+    },
+    {
+      "epoch": 3.3024,
+      "grad_norm": 0.33591721458605356,
+      "learning_rate": 1.2408392576922075e-05,
+      "loss": 0.4179,
+      "step": 258
+    },
+    {
+      "epoch": 3.3152,
+      "grad_norm": 0.3667925625479648,
+      "learning_rate": 1.2243087927902905e-05,
+      "loss": 0.397,
+      "step": 259
+    },
+    {
+      "epoch": 3.328,
+      "grad_norm": 0.3157839864337689,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3612,
+      "step": 260
+    },
+    {
+      "epoch": 3.3407999999999998,
+      "grad_norm": 0.3213885312238281,
+      "learning_rate": 1.1914356023513904e-05,
+      "loss": 0.4029,
+      "step": 261
+    },
+    {
+      "epoch": 3.3536,
+      "grad_norm": 0.3302359587917098,
+      "learning_rate": 1.1750955102607193e-05,
+      "loss": 0.4073,
+      "step": 262
+    },
+    {
+      "epoch": 3.3664,
+      "grad_norm": 0.29441598571428257,
+      "learning_rate": 1.1588215006420374e-05,
+      "loss": 0.4004,
+      "step": 263
+    },
+    {
+      "epoch": 3.3792,
+      "grad_norm": 0.30978932262090814,
+      "learning_rate": 1.1426148771938915e-05,
+      "loss": 0.3983,
+      "step": 264
+    },
+    {
+      "epoch": 3.392,
+      "grad_norm": 0.30007203422440043,
+      "learning_rate": 1.1264769382165748e-05,
+      "loss": 0.4367,
+      "step": 265
+    },
+    {
+      "epoch": 3.4048,
+      "grad_norm": 0.29754442735073233,
+      "learning_rate": 1.110408976508118e-05,
+      "loss": 0.3954,
+      "step": 266
+    },
+    {
+      "epoch": 3.4176,
+      "grad_norm": 0.2810672036171426,
+      "learning_rate": 1.094412279260726e-05,
+      "loss": 0.4133,
+      "step": 267
+    },
+    {
+      "epoch": 3.4304,
+      "grad_norm": 0.27872442410858883,
+      "learning_rate": 1.0784881279576635e-05,
+      "loss": 0.4158,
+      "step": 268
+    },
+    {
+      "epoch": 3.4432,
+      "grad_norm": 0.2752414564773996,
+      "learning_rate": 1.0626377982705929e-05,
+      "loss": 0.4003,
+      "step": 269
+    },
+    {
+      "epoch": 3.456,
+      "grad_norm": 0.2825684542341568,
+      "learning_rate": 1.0468625599573842e-05,
+      "loss": 0.3875,
+      "step": 270
+    },
+    {
+      "epoch": 3.4688,
+      "grad_norm": 0.28187758820442765,
+      "learning_rate": 1.0311636767603952e-05,
+      "loss": 0.4092,
+      "step": 271
+    },
+    {
+      "epoch": 3.4816,
+      "grad_norm": 0.25823425566847735,
+      "learning_rate": 1.0155424063052306e-05,
+      "loss": 0.4083,
+      "step": 272
+    },
+    {
+      "epoch": 3.4944,
+      "grad_norm": 0.27031841166677434,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.4134,
+      "step": 273
+    },
+    {
+      "epoch": 3.5072,
+      "grad_norm": 0.2776823550225193,
+      "learning_rate": 9.84537702935065e-06,
+      "loss": 0.3988,
+      "step": 274
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.30650246734285985,
+      "learning_rate": 9.691567537832964e-06,
+      "loss": 0.3755,
+      "step": 275
+    },
+    {
+      "epoch": 3.5328,
+      "grad_norm": 0.32477357150757086,
+      "learning_rate": 9.538583847008452e-06,
+      "loss": 0.416,
+      "step": 276
+    },
+    {
+      "epoch": 3.5456,
+      "grad_norm": 0.2788014705196268,
+      "learning_rate": 9.386438212284372e-06,
+      "loss": 0.4173,
+      "step": 277
+    },
+    {
+      "epoch": 3.5584,
+      "grad_norm": 0.30861774136922526,
+      "learning_rate": 9.235142821931928e-06,
+      "loss": 0.4141,
+      "step": 278
+    },
+    {
+      "epoch": 3.5712,
+      "grad_norm": 0.3306405964299962,
+      "learning_rate": 9.084709796109907e-06,
+      "loss": 0.3963,
+      "step": 279
+    },
+    {
+      "epoch": 3.584,
+      "grad_norm": 0.2655132439233068,
+      "learning_rate": 8.93515118589373e-06,
+      "loss": 0.4107,
+      "step": 280
+    },
+    {
+      "epoch": 3.5968,
+      "grad_norm": 0.2874632947514457,
+      "learning_rate": 8.786478972310023e-06,
+      "loss": 0.3967,
+      "step": 281
+    },
+    {
+      "epoch": 3.6096,
+      "grad_norm": 0.315468335088531,
+      "learning_rate": 8.638705065376887e-06,
+      "loss": 0.4146,
+      "step": 282
+    },
+    {
+      "epoch": 3.6224,
+      "grad_norm": 0.28204155791769175,
+      "learning_rate": 8.491841303149728e-06,
+      "loss": 0.4096,
+      "step": 283
+    },
+    {
+      "epoch": 3.6352,
+      "grad_norm": 0.26847189337376914,
+      "learning_rate": 8.345899450772975e-06,
+      "loss": 0.4051,
+      "step": 284
+    },
+    {
+      "epoch": 3.648,
+      "grad_norm": 0.2792543365674268,
+      "learning_rate": 8.200891199537549e-06,
+      "loss": 0.3995,
+      "step": 285
+    },
+    {
+      "epoch": 3.6608,
+      "grad_norm": 0.29342950668842016,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.4097,
+      "step": 286
+    },
+    {
+      "epoch": 3.6736,
+      "grad_norm": 0.2784342000075087,
+      "learning_rate": 7.913721890773354e-06,
+      "loss": 0.4132,
+      "step": 287
+    },
+    {
+      "epoch": 3.6864,
+      "grad_norm": 0.2997277490020625,
+      "learning_rate": 7.771583838159756e-06,
+      "loss": 0.3957,
+      "step": 288
+    },
+    {
+      "epoch": 3.6992000000000003,
+      "grad_norm": 0.2501811859736929,
+      "learning_rate": 7.630425394674903e-06,
+      "loss": 0.4073,
+      "step": 289
+    },
+    {
+      "epoch": 3.7119999999999997,
+      "grad_norm": 0.2631978829633404,
+      "learning_rate": 7.49025786841445e-06,
+      "loss": 0.4049,
+      "step": 290
+    },
+    {
+      "epoch": 3.7248,
+      "grad_norm": 0.27384311014579543,
+      "learning_rate": 7.3510924880924575e-06,
+      "loss": 0.402,
+      "step": 291
+    },
+    {
+      "epoch": 3.7376,
+      "grad_norm": 0.28437773635271907,
+      "learning_rate": 7.212940402141808e-06,
+      "loss": 0.3984,
+      "step": 292
+    },
+    {
+      "epoch": 3.7504,
+      "grad_norm": 0.25658095866069414,
+      "learning_rate": 7.075812677821145e-06,
+      "loss": 0.4101,
+      "step": 293
+    },
+    {
+      "epoch": 3.7632,
+      "grad_norm": 0.3018970187904178,
+      "learning_rate": 6.939720300328303e-06,
+      "loss": 0.3971,
+      "step": 294
+    },
+    {
+      "epoch": 3.776,
+      "grad_norm": 0.2808611351344886,
+      "learning_rate": 6.8046741719202385e-06,
+      "loss": 0.405,
+      "step": 295
+    },
+    {
+      "epoch": 3.7888,
+      "grad_norm": 0.24692824734018473,
+      "learning_rate": 6.67068511103971e-06,
+      "loss": 0.4179,
+      "step": 296
+    },
+    {
+      "epoch": 3.8016,
+      "grad_norm": 0.2516970002949449,
+      "learning_rate": 6.537763851448593e-06,
+      "loss": 0.4128,
+      "step": 297
+    },
+    {
+      "epoch": 3.8144,
+      "grad_norm": 0.2855497669115216,
+      "learning_rate": 6.4059210413680175e-06,
+      "loss": 0.4124,
+      "step": 298
+    },
+    {
+      "epoch": 3.8272,
+      "grad_norm": 0.2627089532398548,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3971,
+      "step": 299
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.25278969727962275,
+      "learning_rate": 6.145512929808013e-06,
+      "loss": 0.3957,
+      "step": 300
+    },
+    {
+      "epoch": 3.8528000000000002,
+      "grad_norm": 0.2690667867338895,
+      "learning_rate": 6.016968489424572e-06,
+      "loss": 0.3991,
+      "step": 301
+    },
+    {
+      "epoch": 3.8656,
+      "grad_norm": 0.27107813354791993,
+      "learning_rate": 5.889544219072465e-06,
+      "loss": 0.3826,
+      "step": 302
+    },
+    {
+      "epoch": 3.8784,
+      "grad_norm": 0.24557623802864986,
+      "learning_rate": 5.7632503266131925e-06,
+      "loss": 0.4158,
+      "step": 303
+    },
+    {
+      "epoch": 3.8912,
+      "grad_norm": 0.26451015741555556,
+      "learning_rate": 5.638096929354522e-06,
+      "loss": 0.3973,
+      "step": 304
+    },
+    {
+      "epoch": 3.904,
+      "grad_norm": 0.25580789997428777,
+      "learning_rate": 5.514094053240035e-06,
+      "loss": 0.4109,
+      "step": 305
+    },
+    {
+      "epoch": 3.9168,
+      "grad_norm": 0.23648647171218837,
+      "learning_rate": 5.39125163204594e-06,
+      "loss": 0.3874,
+      "step": 306
+    },
+    {
+      "epoch": 3.9295999999999998,
+      "grad_norm": 0.2557805537687245,
+      "learning_rate": 5.269579506585259e-06,
+      "loss": 0.4011,
+      "step": 307
+    },
+    {
+      "epoch": 3.9424,
+      "grad_norm": 0.26075051352920386,
+      "learning_rate": 5.149087423919541e-06,
+      "loss": 0.4092,
+      "step": 308
+    },
+    {
+      "epoch": 3.9552,
+      "grad_norm": 0.24517119714384644,
+      "learning_rate": 5.029785036577976e-06,
+      "loss": 0.3826,
+      "step": 309
+    },
+    {
+      "epoch": 3.968,
+      "grad_norm": 0.24214584145520887,
+      "learning_rate": 4.911681901784198e-06,
+      "loss": 0.4155,
+      "step": 310
+    },
+    {
+      "epoch": 3.9808,
+      "grad_norm": 0.26415663488367225,
+      "learning_rate": 4.794787480690597e-06,
+      "loss": 0.3933,
+      "step": 311
+    },
+    {
+      "epoch": 3.9936,
+      "grad_norm": 0.2578942066880794,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.4027,
+      "step": 312
+    },
+    {
+      "epoch": 4.0064,
+      "grad_norm": 0.307625626165692,
+      "learning_rate": 4.5646621393177e-06,
+      "loss": 0.3813,
+      "step": 313
+    },
+    {
+      "epoch": 4.0192,
+      "grad_norm": 0.359895059636786,
+      "learning_rate": 4.451449654204685e-06,
+      "loss": 0.3595,
+      "step": 314
+    },
+    {
+      "epoch": 4.032,
+      "grad_norm": 0.24909865193186964,
+      "learning_rate": 4.339482751647557e-06,
+      "loss": 0.3616,
+      "step": 315
+    },
+    {
+      "epoch": 4.0448,
+      "grad_norm": 0.3648806485319156,
+      "learning_rate": 4.228770401229824e-06,
+      "loss": 0.369,
+      "step": 316
+    },
+    {
+      "epoch": 4.0576,
+      "grad_norm": 0.4508768732268639,
+      "learning_rate": 4.119321472033779e-06,
+      "loss": 0.3598,
+      "step": 317
+    },
+    {
+      "epoch": 4.0704,
+      "grad_norm": 0.33319265146607197,
+      "learning_rate": 4.011144731929981e-06,
+      "loss": 0.3708,
+      "step": 318
+    },
+    {
+      "epoch": 4.0832,
+      "grad_norm": 0.2772075012948269,
+      "learning_rate": 3.904248846874894e-06,
+      "loss": 0.3732,
+      "step": 319
+    },
+    {
+      "epoch": 4.096,
+      "grad_norm": 0.32215945026006015,
+      "learning_rate": 3.7986423802166705e-06,
+      "loss": 0.3593,
+      "step": 320
+    },
+    {
+      "epoch": 4.1088,
+      "grad_norm": 0.3279649668658508,
+      "learning_rate": 3.694333792009115e-06,
+      "loss": 0.3692,
+      "step": 321
+    },
+    {
+      "epoch": 4.1216,
+      "grad_norm": 0.31012637762523926,
+      "learning_rate": 3.5913314383339937e-06,
+      "loss": 0.3523,
+      "step": 322
+    },
+    {
+      "epoch": 4.1344,
+      "grad_norm": 0.26246355036669916,
+      "learning_rate": 3.4896435706316e-06,
+      "loss": 0.3583,
+      "step": 323
+    },
+    {
+      "epoch": 4.1472,
+      "grad_norm": 0.28165937028456406,
+      "learning_rate": 3.3892783350397675e-06,
+      "loss": 0.3554,
+      "step": 324
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.29778682042645416,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.3614,
+      "step": 325
+    },
+    {
+      "epoch": 4.1728,
+      "grad_norm": 0.29591789039807403,
+      "learning_rate": 3.1925478143197418e-06,
+      "loss": 0.3522,
+      "step": 326
+    },
+    {
+      "epoch": 4.1856,
+      "grad_norm": 0.24016911221949988,
+      "learning_rate": 3.0961982891241083e-06,
+      "loss": 0.359,
+      "step": 327
+    },
+    {
+      "epoch": 4.1984,
+      "grad_norm": 0.23986536636292846,
+      "learning_rate": 3.001202914641628e-06,
+      "loss": 0.357,
+      "step": 328
+    },
+    {
+      "epoch": 4.2112,
+      "grad_norm": 0.26334209884708454,
+      "learning_rate": 2.907569300879596e-06,
+      "loss": 0.3486,
+      "step": 329
+    },
+    {
+      "epoch": 4.224,
+      "grad_norm": 0.2347298521857969,
+      "learning_rate": 2.815304948755664e-06,
+      "loss": 0.3809,
+      "step": 330
+    },
+    {
+      "epoch": 4.2368,
+      "grad_norm": 0.2477382718090426,
+      "learning_rate": 2.7244172494969978e-06,
+      "loss": 0.3667,
+      "step": 331
+    },
+    {
+      "epoch": 4.2496,
+      "grad_norm": 0.24640888582955534,
+      "learning_rate": 2.6349134840481294e-06,
+      "loss": 0.361,
+      "step": 332
+    },
+    {
+      "epoch": 4.2624,
+      "grad_norm": 0.23681448161959096,
+      "learning_rate": 2.546800822487714e-06,
+      "loss": 0.3578,
+      "step": 333
+    },
+    {
+      "epoch": 4.2752,
+      "grad_norm": 0.24377848059734736,
+      "learning_rate": 2.4600863234541338e-06,
+      "loss": 0.3769,
+      "step": 334
+    },
+    {
+      "epoch": 4.288,
+      "grad_norm": 0.24124734082669758,
+      "learning_rate": 2.374776933580025e-06,
+      "loss": 0.3652,
+      "step": 335
+    },
+    {
+      "epoch": 4.3008,
+      "grad_norm": 0.23567196328347514,
+      "learning_rate": 2.2908794869358044e-06,
+      "loss": 0.3581,
+      "step": 336
+    },
+    {
+      "epoch": 4.3136,
+      "grad_norm": 0.2471817925509731,
+      "learning_rate": 2.2084007044821764e-06,
+      "loss": 0.3491,
+      "step": 337
+    },
+    {
+      "epoch": 4.3264,
+      "grad_norm": 0.22505345794144224,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.3806,
+      "step": 338
+    },
+    {
+      "epoch": 4.3392,
+      "grad_norm": 0.23309771793506243,
+      "learning_rate": 2.0477254472197237e-06,
+      "loss": 0.3592,
+      "step": 339
+    },
+    {
+      "epoch": 4.352,
+      "grad_norm": 0.2307999217043977,
+      "learning_rate": 1.96954184398368e-06,
+      "loss": 0.3648,
+      "step": 340
+    },
+    {
+      "epoch": 4.3648,
+      "grad_norm": 0.226707972968475,
+      "learning_rate": 1.8928026470526917e-06,
+      "loss": 0.3676,
+      "step": 341
+    },
+    {
+      "epoch": 4.3776,
+      "grad_norm": 0.22081392972723213,
+      "learning_rate": 1.817514003945524e-06,
+      "loss": 0.345,
+      "step": 342
+    },
+    {
+      "epoch": 4.3904,
+      "grad_norm": 0.2245861482305544,
+      "learning_rate": 1.743681945978184e-06,
+      "loss": 0.3617,
+      "step": 343
+    },
+    {
+      "epoch": 4.4032,
+      "grad_norm": 0.23412311570974062,
+      "learning_rate": 1.6713123877807413e-06,
+      "loss": 0.384,
+      "step": 344
+    },
+    {
+      "epoch": 4.416,
+      "grad_norm": 0.22298020322013068,
+      "learning_rate": 1.6004111268235156e-06,
+      "loss": 0.3442,
+      "step": 345
+    },
+    {
+      "epoch": 4.4288,
+      "grad_norm": 0.22093630814967327,
+      "learning_rate": 1.5309838429526714e-06,
+      "loss": 0.3619,
+      "step": 346
+    },
+    {
+      "epoch": 4.4416,
+      "grad_norm": 0.22555487152430145,
+      "learning_rate": 1.4630360979351644e-06,
+      "loss": 0.3556,
+      "step": 347
+    },
+    {
+      "epoch": 4.4544,
+      "grad_norm": 0.2230100622437554,
+      "learning_rate": 1.396573335013236e-06,
+      "loss": 0.375,
+      "step": 348
+    },
+    {
+      "epoch": 4.4672,
+      "grad_norm": 0.21836815951522665,
+      "learning_rate": 1.3316008784683265e-06,
+      "loss": 0.3517,
+      "step": 349
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.2218280639803106,
+      "learning_rate": 1.2681239331945695e-06,
+      "loss": 0.3576,
+      "step": 350
+    },
+    {
+      "epoch": 4.4928,
+      "grad_norm": 0.20939855098756494,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.3387,
+      "step": 351
+    },
+    {
+      "epoch": 4.5056,
+      "grad_norm": 0.21161810150051819,
+      "learning_rate": 1.1456767966083393e-06,
+      "loss": 0.3592,
+      "step": 352
+    },
+    {
+      "epoch": 4.5184,
+      "grad_norm": 0.21086232167583766,
+      "learning_rate": 1.086716414442952e-06,
+      "loss": 0.3676,
+      "step": 353
+    },
+    {
+      "epoch": 4.5312,
+      "grad_norm": 0.22193056132525696,
+      "learning_rate": 1.0292711610570904e-06,
+      "loss": 0.3395,
+      "step": 354
+    },
+    {
+      "epoch": 4.5440000000000005,
+      "grad_norm": 0.2139593798230955,
+      "learning_rate": 9.733456383463658e-07,
+      "loss": 0.3497,
+      "step": 355
+    },
+    {
+      "epoch": 4.5568,
+      "grad_norm": 0.22173246216531514,
+      "learning_rate": 9.189443264619102e-07,
+      "loss": 0.345,
+      "step": 356
+    },
+    {
+      "epoch": 4.5696,
+      "grad_norm": 0.2223505979608682,
+      "learning_rate": 8.660715834514977e-07,
+      "loss": 0.3311,
+      "step": 357
+    },
+    {
+      "epoch": 4.5824,
+      "grad_norm": 0.22064500116435354,
+      "learning_rate": 8.147316449103959e-07,
+      "loss": 0.3468,
+      "step": 358
+    },
+    {
+      "epoch": 4.5952,
+      "grad_norm": 0.2653736320358872,
+      "learning_rate": 7.649286236420806e-07,
+      "loss": 0.3599,
+      "step": 359
+    },
+    {
+      "epoch": 4.608,
+      "grad_norm": 0.2058638745445594,
+      "learning_rate": 7.166665093287539e-07,
+      "loss": 0.347,
+      "step": 360
+    },
+    {
+      "epoch": 4.6208,
+      "grad_norm": 0.20543274515074986,
+      "learning_rate": 6.69949168211721e-07,
+      "loss": 0.3703,
+      "step": 361
+    },
+    {
+      "epoch": 4.6336,
+      "grad_norm": 0.2110127031713333,
+      "learning_rate": 6.247803427816945e-07,
+      "loss": 0.3455,
+      "step": 362
+    },
+    {
+      "epoch": 4.6464,
+      "grad_norm": 0.21596202266760922,
+      "learning_rate": 5.811636514789598e-07,
+      "loss": 0.3513,
+      "step": 363
+    },
+    {
+      "epoch": 4.6592,
+      "grad_norm": 0.22522996290186065,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.3306,
+      "step": 364
+    },
+    {
+      "epoch": 4.672,
+      "grad_norm": 0.211823793351152,
+      "learning_rate": 4.986005230351954e-07,
+      "loss": 0.35,
+      "step": 365
+    },
+    {
+      "epoch": 4.6848,
+      "grad_norm": 0.2274284269740909,
+      "learning_rate": 4.5966069996365993e-07,
+      "loss": 0.3468,
+      "step": 366
+    },
+    {
+      "epoch": 4.6975999999999996,
+      "grad_norm": 0.2264157008952427,
+      "learning_rate": 4.22286238628562e-07,
+      "loss": 0.3721,
+      "step": 367
+    },
+    {
+      "epoch": 4.7104,
+      "grad_norm": 0.2087623130384067,
+      "learning_rate": 3.8648013306960664e-07,
+      "loss": 0.3556,
+      "step": 368
+    },
+    {
+      "epoch": 4.7232,
+      "grad_norm": 0.2184256543912645,
+      "learning_rate": 3.522452516867048e-07,
+      "loss": 0.3552,
+      "step": 369
+    },
+    {
+      "epoch": 4.736,
+      "grad_norm": 0.2036290746246842,
+      "learning_rate": 3.1958433701019697e-07,
+      "loss": 0.3655,
+      "step": 370
+    },
+    {
+      "epoch": 4.7488,
+      "grad_norm": 0.2157222040338971,
+      "learning_rate": 2.8850000548115155e-07,
+      "loss": 0.3519,
+      "step": 371
+    },
+    {
+      "epoch": 4.7616,
+      "grad_norm": 0.20350721467845614,
+      "learning_rate": 2.5899474724174313e-07,
+      "loss": 0.3604,
+      "step": 372
+    },
+    {
+      "epoch": 4.7744,
+      "grad_norm": 0.2099619313758119,
+      "learning_rate": 2.3107092593579905e-07,
+      "loss": 0.3544,
+      "step": 373
+    },
+    {
+      "epoch": 4.7872,
+      "grad_norm": 0.2032473270845786,
+      "learning_rate": 2.0473077851942858e-07,
+      "loss": 0.3726,
+      "step": 374
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.20459497279966066,
+      "learning_rate": 1.799764150818306e-07,
+      "loss": 0.3418,
+      "step": 375
+    },
+    {
+      "epoch": 4.8128,
+      "grad_norm": 0.20198939605952645,
+      "learning_rate": 1.5680981867625566e-07,
+      "loss": 0.3799,
+      "step": 376
+    },
+    {
+      "epoch": 4.8256,
+      "grad_norm": 0.21873468599263174,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.3684,
+      "step": 377
+    },
+    {
+      "epoch": 4.8384,
+      "grad_norm": 0.20799442470283128,
+      "learning_rate": 1.1524722305144231e-07,
+      "loss": 0.3705,
+      "step": 378
+    },
+    {
+      "epoch": 4.8512,
+      "grad_norm": 0.21174809799681588,
+      "learning_rate": 9.685455338016347e-08,
+      "loss": 0.3683,
+      "step": 379
+    },
+    {
+      "epoch": 4.864,
+      "grad_norm": 0.20958435877026446,
+      "learning_rate": 8.005630957010014e-08,
+      "loss": 0.3633,
+      "step": 380
+    },
+    {
+      "epoch": 4.8768,
+      "grad_norm": 0.2176512792959373,
+      "learning_rate": 6.485383731580142e-08,
+      "loss": 0.3888,
+      "step": 381
+    },
+    {
+      "epoch": 4.8896,
+      "grad_norm": 0.2073198282002002,
+      "learning_rate": 5.1248354475768034e-08,
+      "loss": 0.3529,
+      "step": 382
+    },
+    {
+      "epoch": 4.9024,
+      "grad_norm": 0.2154539076366996,
+      "learning_rate": 3.924095097489922e-08,
+      "loss": 0.3405,
+      "step": 383
+    },
+    {
+      "epoch": 4.9152000000000005,
+      "grad_norm": 0.20959975179624796,
+      "learning_rate": 2.8832588717164766e-08,
+      "loss": 0.3743,
+      "step": 384
+    },
+    {
+      "epoch": 4.928,
+      "grad_norm": 0.20556679449289894,
+      "learning_rate": 2.0024101508555604e-08,
+      "loss": 0.3555,
+      "step": 385
+    },
+    {
+      "epoch": 4.9408,
+      "grad_norm": 0.21217068333574515,
+      "learning_rate": 1.281619499029274e-08,
+      "loss": 0.3458,
+      "step": 386
+    },
+    {
+      "epoch": 4.9536,
+      "grad_norm": 0.20702970651106367,
+      "learning_rate": 7.209446582292501e-09,
+      "loss": 0.3545,
+      "step": 387
+    },
+    {
+      "epoch": 4.9664,
+      "grad_norm": 0.2123866845066802,
+      "learning_rate": 3.2043054369057523e-09,
+      "loss": 0.3672,
+      "step": 388
+    },
+    {
+      "epoch": 4.9792,
+      "grad_norm": 0.2007902799746407,
+      "learning_rate": 8.010924029533406e-10,
+      "loss": 0.3521,
+      "step": 389
+    },
+    {
+      "epoch": 4.992,
+      "grad_norm": 0.21132775167360818,
+      "learning_rate": 0.0,
+      "loss": 0.37,
+      "step": 390
+    },
+    {
+      "epoch": 4.992,
+      "step": 390,
+      "total_flos": 1.9434542678430188e+18,
+      "train_loss": 0.5044988004060892,
+      "train_runtime": 62031.5499,
+      "train_samples_per_second": 0.806,
+      "train_steps_per_second": 0.006
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 390,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.9434542678430188e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}