xp190 / checkpoint-342 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
29e17ce verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 342,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008771929824561403,
"grad_norm": 39.56407165527344,
"learning_rate": 5.0000000000000004e-08,
"loss": 5.1375,
"step": 1
},
{
"epoch": 0.017543859649122806,
"grad_norm": 40.30452346801758,
"learning_rate": 1.0000000000000001e-07,
"loss": 5.1185,
"step": 2
},
{
"epoch": 0.02631578947368421,
"grad_norm": 40.062313079833984,
"learning_rate": 1.5000000000000002e-07,
"loss": 5.0762,
"step": 3
},
{
"epoch": 0.03508771929824561,
"grad_norm": 39.17148208618164,
"learning_rate": 2.0000000000000002e-07,
"loss": 5.016,
"step": 4
},
{
"epoch": 0.043859649122807015,
"grad_norm": 40.67367172241211,
"learning_rate": 2.5000000000000004e-07,
"loss": 5.0428,
"step": 5
},
{
"epoch": 0.05263157894736842,
"grad_norm": 38.18095016479492,
"learning_rate": 3.0000000000000004e-07,
"loss": 5.2025,
"step": 6
},
{
"epoch": 0.06140350877192982,
"grad_norm": 39.12940979003906,
"learning_rate": 3.5000000000000004e-07,
"loss": 4.9896,
"step": 7
},
{
"epoch": 0.07017543859649122,
"grad_norm": 38.84568405151367,
"learning_rate": 4.0000000000000003e-07,
"loss": 5.1078,
"step": 8
},
{
"epoch": 0.07894736842105263,
"grad_norm": 39.38333511352539,
"learning_rate": 4.5000000000000003e-07,
"loss": 5.0808,
"step": 9
},
{
"epoch": 0.08771929824561403,
"grad_norm": 39.427650451660156,
"learning_rate": 5.000000000000001e-07,
"loss": 5.0534,
"step": 10
},
{
"epoch": 0.09649122807017543,
"grad_norm": 39.29513168334961,
"learning_rate": 5.5e-07,
"loss": 5.058,
"step": 11
},
{
"epoch": 0.10526315789473684,
"grad_norm": 39.641231536865234,
"learning_rate": 6.000000000000001e-07,
"loss": 5.0317,
"step": 12
},
{
"epoch": 0.11403508771929824,
"grad_norm": 37.91259765625,
"learning_rate": 6.5e-07,
"loss": 4.912,
"step": 13
},
{
"epoch": 0.12280701754385964,
"grad_norm": 38.203548431396484,
"learning_rate": 7.000000000000001e-07,
"loss": 4.9705,
"step": 14
},
{
"epoch": 0.13157894736842105,
"grad_norm": 39.15998840332031,
"learning_rate": 7.5e-07,
"loss": 4.6962,
"step": 15
},
{
"epoch": 0.14035087719298245,
"grad_norm": 37.754669189453125,
"learning_rate": 8.000000000000001e-07,
"loss": 4.6262,
"step": 16
},
{
"epoch": 0.14912280701754385,
"grad_norm": 35.871490478515625,
"learning_rate": 8.500000000000001e-07,
"loss": 4.5422,
"step": 17
},
{
"epoch": 0.15789473684210525,
"grad_norm": 36.16888427734375,
"learning_rate": 9.000000000000001e-07,
"loss": 4.664,
"step": 18
},
{
"epoch": 0.16666666666666666,
"grad_norm": 33.520118713378906,
"learning_rate": 9.500000000000001e-07,
"loss": 4.4697,
"step": 19
},
{
"epoch": 0.17543859649122806,
"grad_norm": 30.896282196044922,
"learning_rate": 1.0000000000000002e-06,
"loss": 4.3568,
"step": 20
},
{
"epoch": 0.18421052631578946,
"grad_norm": 29.944643020629883,
"learning_rate": 1.0500000000000001e-06,
"loss": 4.2269,
"step": 21
},
{
"epoch": 0.19298245614035087,
"grad_norm": 25.224485397338867,
"learning_rate": 1.1e-06,
"loss": 4.1272,
"step": 22
},
{
"epoch": 0.20175438596491227,
"grad_norm": 24.410480499267578,
"learning_rate": 1.1500000000000002e-06,
"loss": 4.0585,
"step": 23
},
{
"epoch": 0.21052631578947367,
"grad_norm": 21.480648040771484,
"learning_rate": 1.2000000000000002e-06,
"loss": 3.9472,
"step": 24
},
{
"epoch": 0.21929824561403508,
"grad_norm": 20.61946678161621,
"learning_rate": 1.25e-06,
"loss": 3.8879,
"step": 25
},
{
"epoch": 0.22807017543859648,
"grad_norm": 19.578271865844727,
"learning_rate": 1.3e-06,
"loss": 3.6783,
"step": 26
},
{
"epoch": 0.23684210526315788,
"grad_norm": 17.418983459472656,
"learning_rate": 1.3500000000000002e-06,
"loss": 3.6826,
"step": 27
},
{
"epoch": 0.24561403508771928,
"grad_norm": 18.160301208496094,
"learning_rate": 1.4000000000000001e-06,
"loss": 3.478,
"step": 28
},
{
"epoch": 0.2543859649122807,
"grad_norm": 17.573204040527344,
"learning_rate": 1.45e-06,
"loss": 3.459,
"step": 29
},
{
"epoch": 0.2631578947368421,
"grad_norm": 17.1265869140625,
"learning_rate": 1.5e-06,
"loss": 3.3999,
"step": 30
},
{
"epoch": 0.2719298245614035,
"grad_norm": 15.527145385742188,
"learning_rate": 1.5500000000000002e-06,
"loss": 3.2817,
"step": 31
},
{
"epoch": 0.2807017543859649,
"grad_norm": 14.773847579956055,
"learning_rate": 1.6000000000000001e-06,
"loss": 3.234,
"step": 32
},
{
"epoch": 0.2894736842105263,
"grad_norm": 12.039301872253418,
"learning_rate": 1.6500000000000003e-06,
"loss": 3.132,
"step": 33
},
{
"epoch": 0.2982456140350877,
"grad_norm": 9.217979431152344,
"learning_rate": 1.7000000000000002e-06,
"loss": 3.0548,
"step": 34
},
{
"epoch": 0.30701754385964913,
"grad_norm": 7.575639724731445,
"learning_rate": 1.75e-06,
"loss": 2.9529,
"step": 35
},
{
"epoch": 0.3157894736842105,
"grad_norm": 7.496004104614258,
"learning_rate": 1.8000000000000001e-06,
"loss": 2.8967,
"step": 36
},
{
"epoch": 0.32456140350877194,
"grad_norm": 7.45414924621582,
"learning_rate": 1.85e-06,
"loss": 2.8837,
"step": 37
},
{
"epoch": 0.3333333333333333,
"grad_norm": 8.555658340454102,
"learning_rate": 1.9000000000000002e-06,
"loss": 2.7473,
"step": 38
},
{
"epoch": 0.34210526315789475,
"grad_norm": 10.03805160522461,
"learning_rate": 1.9500000000000004e-06,
"loss": 2.7355,
"step": 39
},
{
"epoch": 0.3508771929824561,
"grad_norm": 9.30649471282959,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.6587,
"step": 40
},
{
"epoch": 0.35964912280701755,
"grad_norm": 8.510339736938477,
"learning_rate": 2.05e-06,
"loss": 2.5977,
"step": 41
},
{
"epoch": 0.3684210526315789,
"grad_norm": 4.709080696105957,
"learning_rate": 2.1000000000000002e-06,
"loss": 2.6286,
"step": 42
},
{
"epoch": 0.37719298245614036,
"grad_norm": 5.128961086273193,
"learning_rate": 2.15e-06,
"loss": 2.4558,
"step": 43
},
{
"epoch": 0.38596491228070173,
"grad_norm": 5.190136432647705,
"learning_rate": 2.2e-06,
"loss": 2.4432,
"step": 44
},
{
"epoch": 0.39473684210526316,
"grad_norm": 4.893551349639893,
"learning_rate": 2.25e-06,
"loss": 2.4939,
"step": 45
},
{
"epoch": 0.40350877192982454,
"grad_norm": 5.2434983253479,
"learning_rate": 2.3000000000000004e-06,
"loss": 2.3381,
"step": 46
},
{
"epoch": 0.41228070175438597,
"grad_norm": 5.122412204742432,
"learning_rate": 2.35e-06,
"loss": 2.313,
"step": 47
},
{
"epoch": 0.42105263157894735,
"grad_norm": 4.577274799346924,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.2236,
"step": 48
},
{
"epoch": 0.4298245614035088,
"grad_norm": 4.722769737243652,
"learning_rate": 2.4500000000000003e-06,
"loss": 2.1987,
"step": 49
},
{
"epoch": 0.43859649122807015,
"grad_norm": 5.059235095977783,
"learning_rate": 2.5e-06,
"loss": 2.1415,
"step": 50
},
{
"epoch": 0.4473684210526316,
"grad_norm": 4.454439640045166,
"learning_rate": 2.55e-06,
"loss": 2.0466,
"step": 51
},
{
"epoch": 0.45614035087719296,
"grad_norm": 4.94586706161499,
"learning_rate": 2.6e-06,
"loss": 1.8762,
"step": 52
},
{
"epoch": 0.4649122807017544,
"grad_norm": 4.704402446746826,
"learning_rate": 2.6500000000000005e-06,
"loss": 1.8012,
"step": 53
},
{
"epoch": 0.47368421052631576,
"grad_norm": 6.125903129577637,
"learning_rate": 2.7000000000000004e-06,
"loss": 1.7669,
"step": 54
},
{
"epoch": 0.4824561403508772,
"grad_norm": 4.5356059074401855,
"learning_rate": 2.7500000000000004e-06,
"loss": 1.6607,
"step": 55
},
{
"epoch": 0.49122807017543857,
"grad_norm": 6.56803035736084,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.6291,
"step": 56
},
{
"epoch": 0.5,
"grad_norm": 4.910050392150879,
"learning_rate": 2.85e-06,
"loss": 1.5545,
"step": 57
},
{
"epoch": 0.5087719298245614,
"grad_norm": 8.733433723449707,
"learning_rate": 2.9e-06,
"loss": 1.4206,
"step": 58
},
{
"epoch": 0.5175438596491229,
"grad_norm": 8.582486152648926,
"learning_rate": 2.95e-06,
"loss": 1.3912,
"step": 59
},
{
"epoch": 0.5263157894736842,
"grad_norm": 13.710689544677734,
"learning_rate": 3e-06,
"loss": 1.3297,
"step": 60
},
{
"epoch": 0.5350877192982456,
"grad_norm": 23.400312423706055,
"learning_rate": 3.05e-06,
"loss": 1.296,
"step": 61
},
{
"epoch": 0.543859649122807,
"grad_norm": 5.678805351257324,
"learning_rate": 3.1000000000000004e-06,
"loss": 1.2259,
"step": 62
},
{
"epoch": 0.5526315789473685,
"grad_norm": 14.700899124145508,
"learning_rate": 3.1500000000000003e-06,
"loss": 1.1087,
"step": 63
},
{
"epoch": 0.5614035087719298,
"grad_norm": 19.38919448852539,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.1805,
"step": 64
},
{
"epoch": 0.5701754385964912,
"grad_norm": 8.460039138793945,
"learning_rate": 3.2500000000000002e-06,
"loss": 1.0963,
"step": 65
},
{
"epoch": 0.5789473684210527,
"grad_norm": 13.371014595031738,
"learning_rate": 3.3000000000000006e-06,
"loss": 1.0627,
"step": 66
},
{
"epoch": 0.5877192982456141,
"grad_norm": 22.380569458007812,
"learning_rate": 3.3500000000000005e-06,
"loss": 1.0869,
"step": 67
},
{
"epoch": 0.5964912280701754,
"grad_norm": 5.780513286590576,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.9991,
"step": 68
},
{
"epoch": 0.6052631578947368,
"grad_norm": 19.850841522216797,
"learning_rate": 3.45e-06,
"loss": 0.9683,
"step": 69
},
{
"epoch": 0.6140350877192983,
"grad_norm": 17.160703659057617,
"learning_rate": 3.5e-06,
"loss": 0.845,
"step": 70
},
{
"epoch": 0.6228070175438597,
"grad_norm": 14.264311790466309,
"learning_rate": 3.5500000000000003e-06,
"loss": 0.8059,
"step": 71
},
{
"epoch": 0.631578947368421,
"grad_norm": 26.39459991455078,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.85,
"step": 72
},
{
"epoch": 0.6403508771929824,
"grad_norm": 51.10348892211914,
"learning_rate": 3.65e-06,
"loss": 0.9755,
"step": 73
},
{
"epoch": 0.6491228070175439,
"grad_norm": 28.795856475830078,
"learning_rate": 3.7e-06,
"loss": 0.8966,
"step": 74
},
{
"epoch": 0.6578947368421053,
"grad_norm": 4.6617937088012695,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.7716,
"step": 75
},
{
"epoch": 0.6666666666666666,
"grad_norm": 15.729666709899902,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.7578,
"step": 76
},
{
"epoch": 0.6754385964912281,
"grad_norm": 7.109970569610596,
"learning_rate": 3.85e-06,
"loss": 0.7055,
"step": 77
},
{
"epoch": 0.6842105263157895,
"grad_norm": 20.84659194946289,
"learning_rate": 3.900000000000001e-06,
"loss": 0.7458,
"step": 78
},
{
"epoch": 0.6929824561403509,
"grad_norm": 21.601303100585938,
"learning_rate": 3.95e-06,
"loss": 0.6879,
"step": 79
},
{
"epoch": 0.7017543859649122,
"grad_norm": 3.6914751529693604,
"learning_rate": 4.000000000000001e-06,
"loss": 0.6179,
"step": 80
},
{
"epoch": 0.7105263157894737,
"grad_norm": 16.539325714111328,
"learning_rate": 4.05e-06,
"loss": 0.5716,
"step": 81
},
{
"epoch": 0.7192982456140351,
"grad_norm": 13.931925773620605,
"learning_rate": 4.1e-06,
"loss": 0.558,
"step": 82
},
{
"epoch": 0.7280701754385965,
"grad_norm": 10.52951717376709,
"learning_rate": 4.15e-06,
"loss": 0.6018,
"step": 83
},
{
"epoch": 0.7368421052631579,
"grad_norm": 17.337060928344727,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.5501,
"step": 84
},
{
"epoch": 0.7456140350877193,
"grad_norm": 13.500468254089355,
"learning_rate": 4.25e-06,
"loss": 0.5214,
"step": 85
},
{
"epoch": 0.7543859649122807,
"grad_norm": 10.290645599365234,
"learning_rate": 4.3e-06,
"loss": 0.4996,
"step": 86
},
{
"epoch": 0.7631578947368421,
"grad_norm": 9.757556915283203,
"learning_rate": 4.350000000000001e-06,
"loss": 0.498,
"step": 87
},
{
"epoch": 0.7719298245614035,
"grad_norm": 9.325140953063965,
"learning_rate": 4.4e-06,
"loss": 0.4721,
"step": 88
},
{
"epoch": 0.7807017543859649,
"grad_norm": 2.9322128295898438,
"learning_rate": 4.450000000000001e-06,
"loss": 0.4528,
"step": 89
},
{
"epoch": 0.7894736842105263,
"grad_norm": 10.484073638916016,
"learning_rate": 4.5e-06,
"loss": 0.445,
"step": 90
},
{
"epoch": 0.7982456140350878,
"grad_norm": 32.7827262878418,
"learning_rate": 4.5500000000000005e-06,
"loss": 0.5105,
"step": 91
},
{
"epoch": 0.8070175438596491,
"grad_norm": 2.8477306365966797,
"learning_rate": 4.600000000000001e-06,
"loss": 0.4117,
"step": 92
},
{
"epoch": 0.8157894736842105,
"grad_norm": 2.7680225372314453,
"learning_rate": 4.65e-06,
"loss": 0.3653,
"step": 93
},
{
"epoch": 0.8245614035087719,
"grad_norm": 2.6512742042541504,
"learning_rate": 4.7e-06,
"loss": 0.3878,
"step": 94
},
{
"epoch": 0.8333333333333334,
"grad_norm": 6.453914165496826,
"learning_rate": 4.75e-06,
"loss": 0.3611,
"step": 95
},
{
"epoch": 0.8421052631578947,
"grad_norm": 3.4594080448150635,
"learning_rate": 4.800000000000001e-06,
"loss": 0.3817,
"step": 96
},
{
"epoch": 0.8508771929824561,
"grad_norm": 3.6144917011260986,
"learning_rate": 4.85e-06,
"loss": 0.3618,
"step": 97
},
{
"epoch": 0.8596491228070176,
"grad_norm": 5.349407196044922,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.3218,
"step": 98
},
{
"epoch": 0.868421052631579,
"grad_norm": 13.671236991882324,
"learning_rate": 4.95e-06,
"loss": 0.3329,
"step": 99
},
{
"epoch": 0.8771929824561403,
"grad_norm": 5.84046745300293,
"learning_rate": 5e-06,
"loss": 0.2967,
"step": 100
},
{
"epoch": 0.8859649122807017,
"grad_norm": 14.005338668823242,
"learning_rate": 4.999963827125897e-06,
"loss": 0.303,
"step": 101
},
{
"epoch": 0.8947368421052632,
"grad_norm": 9.18114185333252,
"learning_rate": 4.999855309550366e-06,
"loss": 0.2762,
"step": 102
},
{
"epoch": 0.9035087719298246,
"grad_norm": 3.0800487995147705,
"learning_rate": 4.999674450413725e-06,
"loss": 0.2628,
"step": 103
},
{
"epoch": 0.9122807017543859,
"grad_norm": 82.03578186035156,
"learning_rate": 4.999421254949728e-06,
"loss": 0.4065,
"step": 104
},
{
"epoch": 0.9210526315789473,
"grad_norm": 77.66315460205078,
"learning_rate": 4.99909573048542e-06,
"loss": 0.4307,
"step": 105
},
{
"epoch": 0.9298245614035088,
"grad_norm": 18.28767967224121,
"learning_rate": 4.998697886440927e-06,
"loss": 0.2571,
"step": 106
},
{
"epoch": 0.9385964912280702,
"grad_norm": 5.960445880889893,
"learning_rate": 4.998227734329177e-06,
"loss": 0.2847,
"step": 107
},
{
"epoch": 0.9473684210526315,
"grad_norm": 5.437699794769287,
"learning_rate": 4.9976852877555755e-06,
"loss": 0.2728,
"step": 108
},
{
"epoch": 0.956140350877193,
"grad_norm": 3.379631280899048,
"learning_rate": 4.997070562417602e-06,
"loss": 0.2467,
"step": 109
},
{
"epoch": 0.9649122807017544,
"grad_norm": 3.1625075340270996,
"learning_rate": 4.996383576104362e-06,
"loss": 0.2273,
"step": 110
},
{
"epoch": 0.9736842105263158,
"grad_norm": 15.588600158691406,
"learning_rate": 4.995624348696071e-06,
"loss": 0.2486,
"step": 111
},
{
"epoch": 0.9824561403508771,
"grad_norm": 2.631044387817383,
"learning_rate": 4.9947929021634815e-06,
"loss": 0.1964,
"step": 112
},
{
"epoch": 0.9912280701754386,
"grad_norm": 4.706504821777344,
"learning_rate": 4.993889260567239e-06,
"loss": 0.1901,
"step": 113
},
{
"epoch": 1.0,
"grad_norm": 10.368465423583984,
"learning_rate": 4.9929134500571954e-06,
"loss": 0.1996,
"step": 114
},
{
"epoch": 1.0087719298245614,
"grad_norm": 30.44986343383789,
"learning_rate": 4.991865498871647e-06,
"loss": 0.2606,
"step": 115
},
{
"epoch": 1.0175438596491229,
"grad_norm": 14.421515464782715,
"learning_rate": 4.99074543733652e-06,
"loss": 0.2394,
"step": 116
},
{
"epoch": 1.0263157894736843,
"grad_norm": 14.072005271911621,
"learning_rate": 4.989553297864489e-06,
"loss": 0.2288,
"step": 117
},
{
"epoch": 1.0350877192982457,
"grad_norm": 4.395325660705566,
"learning_rate": 4.988289114954045e-06,
"loss": 0.2129,
"step": 118
},
{
"epoch": 1.043859649122807,
"grad_norm": 7.286703586578369,
"learning_rate": 4.986952925188489e-06,
"loss": 0.186,
"step": 119
},
{
"epoch": 1.0526315789473684,
"grad_norm": 8.332784652709961,
"learning_rate": 4.98554476723488e-06,
"loss": 0.178,
"step": 120
},
{
"epoch": 1.0614035087719298,
"grad_norm": 1.3646447658538818,
"learning_rate": 4.984064681842917e-06,
"loss": 0.1687,
"step": 121
},
{
"epoch": 1.0701754385964912,
"grad_norm": 4.494940757751465,
"learning_rate": 4.982512711843753e-06,
"loss": 0.1881,
"step": 122
},
{
"epoch": 1.0789473684210527,
"grad_norm": 3.3929836750030518,
"learning_rate": 4.980888902148757e-06,
"loss": 0.1764,
"step": 123
},
{
"epoch": 1.087719298245614,
"grad_norm": 1.8281155824661255,
"learning_rate": 4.979193299748225e-06,
"loss": 0.1602,
"step": 124
},
{
"epoch": 1.0964912280701755,
"grad_norm": 3.494239568710327,
"learning_rate": 4.977425953710005e-06,
"loss": 0.1729,
"step": 125
},
{
"epoch": 1.1052631578947367,
"grad_norm": 1.500410556793213,
"learning_rate": 4.975586915178084e-06,
"loss": 0.1666,
"step": 126
},
{
"epoch": 1.1140350877192982,
"grad_norm": 1.4680222272872925,
"learning_rate": 4.973676237371111e-06,
"loss": 0.159,
"step": 127
},
{
"epoch": 1.1228070175438596,
"grad_norm": 3.0383460521698,
"learning_rate": 4.971693975580851e-06,
"loss": 0.1484,
"step": 128
},
{
"epoch": 1.131578947368421,
"grad_norm": 3.74821138381958,
"learning_rate": 4.969640187170591e-06,
"loss": 0.1586,
"step": 129
},
{
"epoch": 1.1403508771929824,
"grad_norm": 4.682602405548096,
"learning_rate": 4.967514931573473e-06,
"loss": 0.1619,
"step": 130
},
{
"epoch": 1.1491228070175439,
"grad_norm": 3.90673565864563,
"learning_rate": 4.965318270290779e-06,
"loss": 0.164,
"step": 131
},
{
"epoch": 1.1578947368421053,
"grad_norm": 2.2017388343811035,
"learning_rate": 4.963050266890152e-06,
"loss": 0.1499,
"step": 132
},
{
"epoch": 1.1666666666666667,
"grad_norm": 2.4211816787719727,
"learning_rate": 4.960710987003753e-06,
"loss": 0.1387,
"step": 133
},
{
"epoch": 1.1754385964912282,
"grad_norm": 1.7753759622573853,
"learning_rate": 4.958300498326363e-06,
"loss": 0.1441,
"step": 134
},
{
"epoch": 1.1842105263157894,
"grad_norm": 1.5529910326004028,
"learning_rate": 4.955818870613425e-06,
"loss": 0.1304,
"step": 135
},
{
"epoch": 1.1929824561403508,
"grad_norm": 2.090593099594116,
"learning_rate": 4.953266175679023e-06,
"loss": 0.1419,
"step": 136
},
{
"epoch": 1.2017543859649122,
"grad_norm": 2.7141878604888916,
"learning_rate": 4.95064248739381e-06,
"loss": 0.1444,
"step": 137
},
{
"epoch": 1.2105263157894737,
"grad_norm": 2.3690481185913086,
"learning_rate": 4.947947881682861e-06,
"loss": 0.1383,
"step": 138
},
{
"epoch": 1.219298245614035,
"grad_norm": 2.2403147220611572,
"learning_rate": 4.945182436523482e-06,
"loss": 0.1418,
"step": 139
},
{
"epoch": 1.2280701754385965,
"grad_norm": 1.3939160108566284,
"learning_rate": 4.942346231942955e-06,
"loss": 0.1307,
"step": 140
},
{
"epoch": 1.236842105263158,
"grad_norm": 11.276732444763184,
"learning_rate": 4.939439350016214e-06,
"loss": 0.1397,
"step": 141
},
{
"epoch": 1.2456140350877192,
"grad_norm": 8.260516166687012,
"learning_rate": 4.9364618748634794e-06,
"loss": 0.1426,
"step": 142
},
{
"epoch": 1.2543859649122808,
"grad_norm": 2.09720516204834,
"learning_rate": 4.933413892647819e-06,
"loss": 0.1323,
"step": 143
},
{
"epoch": 1.263157894736842,
"grad_norm": 1.802125334739685,
"learning_rate": 4.9302954915726535e-06,
"loss": 0.1304,
"step": 144
},
{
"epoch": 1.2719298245614035,
"grad_norm": 1.7151471376419067,
"learning_rate": 4.927106761879207e-06,
"loss": 0.1264,
"step": 145
},
{
"epoch": 1.280701754385965,
"grad_norm": 1.6970336437225342,
"learning_rate": 4.923847795843894e-06,
"loss": 0.1227,
"step": 146
},
{
"epoch": 1.2894736842105263,
"grad_norm": 16.60441017150879,
"learning_rate": 4.920518687775647e-06,
"loss": 0.1606,
"step": 147
},
{
"epoch": 1.2982456140350878,
"grad_norm": 6.470354080200195,
"learning_rate": 4.917119534013194e-06,
"loss": 0.1447,
"step": 148
},
{
"epoch": 1.3070175438596492,
"grad_norm": 1.4908231496810913,
"learning_rate": 4.913650432922264e-06,
"loss": 0.1343,
"step": 149
},
{
"epoch": 1.3157894736842106,
"grad_norm": 3.19964861869812,
"learning_rate": 4.91011148489274e-06,
"loss": 0.1354,
"step": 150
},
{
"epoch": 1.3245614035087718,
"grad_norm": 2.6052839756011963,
"learning_rate": 4.906502792335761e-06,
"loss": 0.1342,
"step": 151
},
{
"epoch": 1.3333333333333333,
"grad_norm": 2.0719165802001953,
"learning_rate": 4.9028244596807525e-06,
"loss": 0.1359,
"step": 152
},
{
"epoch": 1.3421052631578947,
"grad_norm": 0.8086919784545898,
"learning_rate": 4.899076593372405e-06,
"loss": 0.1279,
"step": 153
},
{
"epoch": 1.3508771929824561,
"grad_norm": 1.0056848526000977,
"learning_rate": 4.8952593018675955e-06,
"loss": 0.1162,
"step": 154
},
{
"epoch": 1.3596491228070176,
"grad_norm": 5.72553014755249,
"learning_rate": 4.891372695632249e-06,
"loss": 0.1315,
"step": 155
},
{
"epoch": 1.368421052631579,
"grad_norm": 1.522894024848938,
"learning_rate": 4.887416887138139e-06,
"loss": 0.1266,
"step": 156
},
{
"epoch": 1.3771929824561404,
"grad_norm": 2.019472122192383,
"learning_rate": 4.883391990859635e-06,
"loss": 0.1262,
"step": 157
},
{
"epoch": 1.3859649122807016,
"grad_norm": 1.8594422340393066,
"learning_rate": 4.879298123270391e-06,
"loss": 0.125,
"step": 158
},
{
"epoch": 1.3947368421052633,
"grad_norm": 1.365377426147461,
"learning_rate": 4.8751354028399725e-06,
"loss": 0.1218,
"step": 159
},
{
"epoch": 1.4035087719298245,
"grad_norm": 3.553309917449951,
"learning_rate": 4.870903950030429e-06,
"loss": 0.1272,
"step": 160
},
{
"epoch": 1.412280701754386,
"grad_norm": 2.1770920753479004,
"learning_rate": 4.866603887292809e-06,
"loss": 0.1213,
"step": 161
},
{
"epoch": 1.4210526315789473,
"grad_norm": 1.6058955192565918,
"learning_rate": 4.862235339063613e-06,
"loss": 0.1173,
"step": 162
},
{
"epoch": 1.4298245614035088,
"grad_norm": 1.3208314180374146,
"learning_rate": 4.857798431761199e-06,
"loss": 0.1183,
"step": 163
},
{
"epoch": 1.4385964912280702,
"grad_norm": 1.282729983329773,
"learning_rate": 4.853293293782118e-06,
"loss": 0.1209,
"step": 164
},
{
"epoch": 1.4473684210526316,
"grad_norm": 1.3838152885437012,
"learning_rate": 4.848720055497401e-06,
"loss": 0.1198,
"step": 165
},
{
"epoch": 1.456140350877193,
"grad_norm": 1.2930737733840942,
"learning_rate": 4.844078849248785e-06,
"loss": 0.1268,
"step": 166
},
{
"epoch": 1.4649122807017543,
"grad_norm": 1.7022266387939453,
"learning_rate": 4.839369809344888e-06,
"loss": 0.1198,
"step": 167
},
{
"epoch": 1.4736842105263157,
"grad_norm": 1.0927815437316895,
"learning_rate": 4.834593072057313e-06,
"loss": 0.1132,
"step": 168
},
{
"epoch": 1.4824561403508771,
"grad_norm": 0.9326333999633789,
"learning_rate": 4.829748775616716e-06,
"loss": 0.1193,
"step": 169
},
{
"epoch": 1.4912280701754386,
"grad_norm": 1.3564742803573608,
"learning_rate": 4.8248370602087954e-06,
"loss": 0.118,
"step": 170
},
{
"epoch": 1.5,
"grad_norm": 1.19778573513031,
"learning_rate": 4.819858067970243e-06,
"loss": 0.1122,
"step": 171
},
{
"epoch": 1.5087719298245614,
"grad_norm": 2.8438351154327393,
"learning_rate": 4.814811942984625e-06,
"loss": 0.1217,
"step": 172
},
{
"epoch": 1.5175438596491229,
"grad_norm": 1.0701063871383667,
"learning_rate": 4.809698831278217e-06,
"loss": 0.1114,
"step": 173
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.9053553938865662,
"learning_rate": 4.804518880815776e-06,
"loss": 0.1178,
"step": 174
},
{
"epoch": 1.5350877192982457,
"grad_norm": 0.42274603247642517,
"learning_rate": 4.799272241496259e-06,
"loss": 0.1091,
"step": 175
},
{
"epoch": 1.543859649122807,
"grad_norm": 0.8576470017433167,
"learning_rate": 4.793959065148484e-06,
"loss": 0.1134,
"step": 176
},
{
"epoch": 1.5526315789473686,
"grad_norm": 0.5910662412643433,
"learning_rate": 4.78857950552674e-06,
"loss": 0.1148,
"step": 177
},
{
"epoch": 1.5614035087719298,
"grad_norm": 0.8761632442474365,
"learning_rate": 4.783133718306331e-06,
"loss": 0.1125,
"step": 178
},
{
"epoch": 1.5701754385964912,
"grad_norm": 1.9190795421600342,
"learning_rate": 4.777621861079079e-06,
"loss": 0.1148,
"step": 179
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.6199957728385925,
"learning_rate": 4.772044093348757e-06,
"loss": 0.1097,
"step": 180
},
{
"epoch": 1.587719298245614,
"grad_norm": 1.562089443206787,
"learning_rate": 4.766400576526479e-06,
"loss": 0.1097,
"step": 181
},
{
"epoch": 1.5964912280701755,
"grad_norm": 1.4957091808319092,
"learning_rate": 4.760691473926021e-06,
"loss": 0.1216,
"step": 182
},
{
"epoch": 1.6052631578947367,
"grad_norm": 0.9863570332527161,
"learning_rate": 4.754916950759105e-06,
"loss": 0.1122,
"step": 183
},
{
"epoch": 1.6140350877192984,
"grad_norm": 0.5803346633911133,
"learning_rate": 4.749077174130609e-06,
"loss": 0.1103,
"step": 184
},
{
"epoch": 1.6228070175438596,
"grad_norm": 1.8789891004562378,
"learning_rate": 4.743172313033738e-06,
"loss": 0.1191,
"step": 185
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.8731380105018616,
"learning_rate": 4.7372025383451285e-06,
"loss": 0.1154,
"step": 186
},
{
"epoch": 1.6403508771929824,
"grad_norm": 1.3535627126693726,
"learning_rate": 4.7311680228199075e-06,
"loss": 0.1123,
"step": 187
},
{
"epoch": 1.6491228070175439,
"grad_norm": 0.7211089134216309,
"learning_rate": 4.725068941086693e-06,
"loss": 0.1134,
"step": 188
},
{
"epoch": 1.6578947368421053,
"grad_norm": 1.4752328395843506,
"learning_rate": 4.718905469642534e-06,
"loss": 0.1185,
"step": 189
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.9822680354118347,
"learning_rate": 4.712677786847814e-06,
"loss": 0.1146,
"step": 190
},
{
"epoch": 1.6754385964912282,
"grad_norm": 1.1308330297470093,
"learning_rate": 4.706386072921083e-06,
"loss": 0.1061,
"step": 191
},
{
"epoch": 1.6842105263157894,
"grad_norm": 5.331939697265625,
"learning_rate": 4.70003050993384e-06,
"loss": 0.1153,
"step": 192
},
{
"epoch": 1.692982456140351,
"grad_norm": 0.6911673545837402,
"learning_rate": 4.6936112818052674e-06,
"loss": 0.1098,
"step": 193
},
{
"epoch": 1.7017543859649122,
"grad_norm": 0.5160980224609375,
"learning_rate": 4.687128574296912e-06,
"loss": 0.1073,
"step": 194
},
{
"epoch": 1.7105263157894737,
"grad_norm": 1.5724798440933228,
"learning_rate": 4.680582575007303e-06,
"loss": 0.121,
"step": 195
},
{
"epoch": 1.719298245614035,
"grad_norm": 1.3960011005401611,
"learning_rate": 4.6739734733665275e-06,
"loss": 0.1145,
"step": 196
},
{
"epoch": 1.7280701754385965,
"grad_norm": 1.4949183464050293,
"learning_rate": 4.6673014606307465e-06,
"loss": 0.1166,
"step": 197
},
{
"epoch": 1.736842105263158,
"grad_norm": 1.6873422861099243,
"learning_rate": 4.660566729876661e-06,
"loss": 0.1115,
"step": 198
},
{
"epoch": 1.7456140350877192,
"grad_norm": 1.3443641662597656,
"learning_rate": 4.653769475995926e-06,
"loss": 0.1119,
"step": 199
},
{
"epoch": 1.7543859649122808,
"grad_norm": 0.807525098323822,
"learning_rate": 4.646909895689508e-06,
"loss": 0.1059,
"step": 200
},
{
"epoch": 1.763157894736842,
"grad_norm": 1.589316964149475,
"learning_rate": 4.639988187461995e-06,
"loss": 0.1151,
"step": 201
},
{
"epoch": 1.7719298245614035,
"grad_norm": 2.474756956100464,
"learning_rate": 4.633004551615851e-06,
"loss": 0.116,
"step": 202
},
{
"epoch": 1.780701754385965,
"grad_norm": 0.6210195422172546,
"learning_rate": 4.62595919024562e-06,
"loss": 0.1097,
"step": 203
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.7217905521392822,
"learning_rate": 4.618852307232078e-06,
"loss": 0.1117,
"step": 204
},
{
"epoch": 1.7982456140350878,
"grad_norm": 1.551251769065857,
"learning_rate": 4.611684108236334e-06,
"loss": 0.113,
"step": 205
},
{
"epoch": 1.807017543859649,
"grad_norm": 0.6619828939437866,
"learning_rate": 4.604454800693874e-06,
"loss": 0.113,
"step": 206
},
{
"epoch": 1.8157894736842106,
"grad_norm": 0.9461805820465088,
"learning_rate": 4.597164593808564e-06,
"loss": 0.1093,
"step": 207
},
{
"epoch": 1.8245614035087718,
"grad_norm": 1.2926547527313232,
"learning_rate": 4.589813698546592e-06,
"loss": 0.1128,
"step": 208
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.8754212856292725,
"learning_rate": 4.582402327630368e-06,
"loss": 0.1104,
"step": 209
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.846051812171936,
"learning_rate": 4.574930695532357e-06,
"loss": 0.1105,
"step": 210
},
{
"epoch": 1.8508771929824561,
"grad_norm": 1.3332515954971313,
"learning_rate": 4.567399018468889e-06,
"loss": 0.1101,
"step": 211
},
{
"epoch": 1.8596491228070176,
"grad_norm": 0.8729192614555359,
"learning_rate": 4.5598075143938855e-06,
"loss": 0.1081,
"step": 212
},
{
"epoch": 1.868421052631579,
"grad_norm": 0.8618345260620117,
"learning_rate": 4.552156402992567e-06,
"loss": 0.1059,
"step": 213
},
{
"epoch": 1.8771929824561404,
"grad_norm": 1.2135930061340332,
"learning_rate": 4.544445905675082e-06,
"loss": 0.1105,
"step": 214
},
{
"epoch": 1.8859649122807016,
"grad_norm": 0.8405666351318359,
"learning_rate": 4.536676245570111e-06,
"loss": 0.1118,
"step": 215
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.42860639095306396,
"learning_rate": 4.528847647518403e-06,
"loss": 0.1093,
"step": 216
},
{
"epoch": 1.9035087719298245,
"grad_norm": 1.1538206338882446,
"learning_rate": 4.520960338066271e-06,
"loss": 0.1088,
"step": 217
},
{
"epoch": 1.912280701754386,
"grad_norm": 0.5870749354362488,
"learning_rate": 4.513014545459038e-06,
"loss": 0.1061,
"step": 218
},
{
"epoch": 1.9210526315789473,
"grad_norm": 0.7279748916625977,
"learning_rate": 4.505010499634427e-06,
"loss": 0.1032,
"step": 219
},
{
"epoch": 1.9298245614035088,
"grad_norm": 0.6331414580345154,
"learning_rate": 4.4969484322159125e-06,
"loss": 0.1109,
"step": 220
},
{
"epoch": 1.9385964912280702,
"grad_norm": 0.9024543166160583,
"learning_rate": 4.488828576506014e-06,
"loss": 0.1094,
"step": 221
},
{
"epoch": 1.9473684210526314,
"grad_norm": 3.540376901626587,
"learning_rate": 4.480651167479545e-06,
"loss": 0.1154,
"step": 222
},
{
"epoch": 1.956140350877193,
"grad_norm": 0.9506739377975464,
"learning_rate": 4.472416441776817e-06,
"loss": 0.108,
"step": 223
},
{
"epoch": 1.9649122807017543,
"grad_norm": 0.6585081815719604,
"learning_rate": 4.464124637696786e-06,
"loss": 0.1033,
"step": 224
},
{
"epoch": 1.973684210526316,
"grad_norm": 1.143038034439087,
"learning_rate": 4.455775995190161e-06,
"loss": 0.1092,
"step": 225
},
{
"epoch": 1.9824561403508771,
"grad_norm": 1.148261547088623,
"learning_rate": 4.4473707558524555e-06,
"loss": 0.1076,
"step": 226
},
{
"epoch": 1.9912280701754386,
"grad_norm": 0.7375811338424683,
"learning_rate": 4.438909162917003e-06,
"loss": 0.108,
"step": 227
},
{
"epoch": 2.0,
"grad_norm": 0.5254591703414917,
"learning_rate": 4.430391461247911e-06,
"loss": 0.1079,
"step": 228
},
{
"epoch": 2.008771929824561,
"grad_norm": 1.0198495388031006,
"learning_rate": 4.42181789733298e-06,
"loss": 0.1083,
"step": 229
},
{
"epoch": 2.017543859649123,
"grad_norm": 0.9234157800674438,
"learning_rate": 4.413188719276569e-06,
"loss": 0.1084,
"step": 230
},
{
"epoch": 2.026315789473684,
"grad_norm": 0.5215068459510803,
"learning_rate": 4.404504176792414e-06,
"loss": 0.1067,
"step": 231
},
{
"epoch": 2.0350877192982457,
"grad_norm": 0.9296736121177673,
"learning_rate": 4.3957645211964065e-06,
"loss": 0.1066,
"step": 232
},
{
"epoch": 2.043859649122807,
"grad_norm": 0.8660671710968018,
"learning_rate": 4.386970005399314e-06,
"loss": 0.108,
"step": 233
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.6014883518218994,
"learning_rate": 4.378120883899467e-06,
"loss": 0.1068,
"step": 234
},
{
"epoch": 2.06140350877193,
"grad_norm": 0.6370371580123901,
"learning_rate": 4.369217412775393e-06,
"loss": 0.1076,
"step": 235
},
{
"epoch": 2.0701754385964914,
"grad_norm": 0.9806828498840332,
"learning_rate": 4.360259849678402e-06,
"loss": 0.1071,
"step": 236
},
{
"epoch": 2.0789473684210527,
"grad_norm": 0.6093440651893616,
"learning_rate": 4.351248453825137e-06,
"loss": 0.1038,
"step": 237
},
{
"epoch": 2.087719298245614,
"grad_norm": 1.3494842052459717,
"learning_rate": 4.3421834859900695e-06,
"loss": 0.1105,
"step": 238
},
{
"epoch": 2.0964912280701755,
"grad_norm": 0.7621576189994812,
"learning_rate": 4.333065208497949e-06,
"loss": 0.1048,
"step": 239
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.5918282866477966,
"learning_rate": 4.3238938852162195e-06,
"loss": 0.1086,
"step": 240
},
{
"epoch": 2.1140350877192984,
"grad_norm": 0.7048676609992981,
"learning_rate": 4.314669781547379e-06,
"loss": 0.1061,
"step": 241
},
{
"epoch": 2.1228070175438596,
"grad_norm": 1.0750821828842163,
"learning_rate": 4.305393164421301e-06,
"loss": 0.1082,
"step": 242
},
{
"epoch": 2.1315789473684212,
"grad_norm": 0.6171414852142334,
"learning_rate": 4.296064302287507e-06,
"loss": 0.1039,
"step": 243
},
{
"epoch": 2.1403508771929824,
"grad_norm": 0.8080905079841614,
"learning_rate": 4.286683465107403e-06,
"loss": 0.1069,
"step": 244
},
{
"epoch": 2.1491228070175437,
"grad_norm": 0.5281466245651245,
"learning_rate": 4.277250924346461e-06,
"loss": 0.1069,
"step": 245
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.8070254325866699,
"learning_rate": 4.267766952966369e-06,
"loss": 0.1061,
"step": 246
},
{
"epoch": 2.1666666666666665,
"grad_norm": 0.8560577630996704,
"learning_rate": 4.25823182541713e-06,
"loss": 0.1116,
"step": 247
},
{
"epoch": 2.175438596491228,
"grad_norm": 0.7772330045700073,
"learning_rate": 4.2486458176291176e-06,
"loss": 0.1092,
"step": 248
},
{
"epoch": 2.1842105263157894,
"grad_norm": 0.814601719379425,
"learning_rate": 4.239009207005096e-06,
"loss": 0.1093,
"step": 249
},
{
"epoch": 2.192982456140351,
"grad_norm": 0.957789957523346,
"learning_rate": 4.2293222724121855e-06,
"loss": 0.1075,
"step": 250
},
{
"epoch": 2.2017543859649122,
"grad_norm": 0.500062108039856,
"learning_rate": 4.219585294173799e-06,
"loss": 0.1048,
"step": 251
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.3866419792175293,
"learning_rate": 4.209798554061527e-06,
"loss": 0.1074,
"step": 252
},
{
"epoch": 2.219298245614035,
"grad_norm": 1.1853291988372803,
"learning_rate": 4.199962335286985e-06,
"loss": 0.1076,
"step": 253
},
{
"epoch": 2.2280701754385963,
"grad_norm": 0.36602887511253357,
"learning_rate": 4.1900769224936125e-06,
"loss": 0.108,
"step": 254
},
{
"epoch": 2.236842105263158,
"grad_norm": 0.2530711889266968,
"learning_rate": 4.180142601748447e-06,
"loss": 0.1041,
"step": 255
},
{
"epoch": 2.245614035087719,
"grad_norm": 1.3067054748535156,
"learning_rate": 4.170159660533834e-06,
"loss": 0.1087,
"step": 256
},
{
"epoch": 2.254385964912281,
"grad_norm": 0.3442043960094452,
"learning_rate": 4.160128387739114e-06,
"loss": 0.1099,
"step": 257
},
{
"epoch": 2.263157894736842,
"grad_norm": 1.174796462059021,
"learning_rate": 4.150049073652262e-06,
"loss": 0.1063,
"step": 258
},
{
"epoch": 2.2719298245614037,
"grad_norm": 0.5719411969184875,
"learning_rate": 4.1399220099514845e-06,
"loss": 0.1043,
"step": 259
},
{
"epoch": 2.280701754385965,
"grad_norm": 0.7268956303596497,
"learning_rate": 4.129747489696781e-06,
"loss": 0.1038,
"step": 260
},
{
"epoch": 2.2894736842105265,
"grad_norm": 0.7028316259384155,
"learning_rate": 4.119525807321467e-06,
"loss": 0.1052,
"step": 261
},
{
"epoch": 2.2982456140350878,
"grad_norm": 1.015335202217102,
"learning_rate": 4.109257258623644e-06,
"loss": 0.1116,
"step": 262
},
{
"epoch": 2.307017543859649,
"grad_norm": 0.7141755819320679,
"learning_rate": 4.098942140757646e-06,
"loss": 0.108,
"step": 263
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.7656403183937073,
"learning_rate": 4.0885807522254435e-06,
"loss": 0.1043,
"step": 264
},
{
"epoch": 2.324561403508772,
"grad_norm": 0.43293774127960205,
"learning_rate": 4.078173392867998e-06,
"loss": 0.1048,
"step": 265
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.6755763292312622,
"learning_rate": 4.0677203638565895e-06,
"loss": 0.1064,
"step": 266
},
{
"epoch": 2.3421052631578947,
"grad_norm": 0.9648827314376831,
"learning_rate": 4.0572219676841e-06,
"loss": 0.1088,
"step": 267
},
{
"epoch": 2.3508771929824563,
"grad_norm": 0.32724836468696594,
"learning_rate": 4.046678508156259e-06,
"loss": 0.1077,
"step": 268
},
{
"epoch": 2.3596491228070176,
"grad_norm": 0.4696657061576843,
"learning_rate": 4.036090290382855e-06,
"loss": 0.1067,
"step": 269
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.33901306986808777,
"learning_rate": 4.025457620768901e-06,
"loss": 0.105,
"step": 270
},
{
"epoch": 2.3771929824561404,
"grad_norm": 0.5703794360160828,
"learning_rate": 4.014780807005775e-06,
"loss": 0.1033,
"step": 271
},
{
"epoch": 2.3859649122807016,
"grad_norm": 0.9639355540275574,
"learning_rate": 4.004060158062306e-06,
"loss": 0.1041,
"step": 272
},
{
"epoch": 2.3947368421052633,
"grad_norm": 0.8851558566093445,
"learning_rate": 3.993295984175845e-06,
"loss": 0.1064,
"step": 273
},
{
"epoch": 2.4035087719298245,
"grad_norm": 0.5200062990188599,
"learning_rate": 3.982488596843276e-06,
"loss": 0.1056,
"step": 274
},
{
"epoch": 2.412280701754386,
"grad_norm": 1.160823106765747,
"learning_rate": 3.971638308812007e-06,
"loss": 0.1069,
"step": 275
},
{
"epoch": 2.4210526315789473,
"grad_norm": 1.0191210508346558,
"learning_rate": 3.9607454340709215e-06,
"loss": 0.1042,
"step": 276
},
{
"epoch": 2.4298245614035086,
"grad_norm": 0.37181487679481506,
"learning_rate": 3.949810287841289e-06,
"loss": 0.1062,
"step": 277
},
{
"epoch": 2.43859649122807,
"grad_norm": 0.9328593611717224,
"learning_rate": 3.9388331865676436e-06,
"loss": 0.1086,
"step": 278
},
{
"epoch": 2.4473684210526314,
"grad_norm": 0.8024734258651733,
"learning_rate": 3.927814447908625e-06,
"loss": 0.1051,
"step": 279
},
{
"epoch": 2.456140350877193,
"grad_norm": 0.9746696352958679,
"learning_rate": 3.916754390727795e-06,
"loss": 0.1041,
"step": 280
},
{
"epoch": 2.4649122807017543,
"grad_norm": 0.5457844138145447,
"learning_rate": 3.905653335084394e-06,
"loss": 0.1052,
"step": 281
},
{
"epoch": 2.473684210526316,
"grad_norm": 1.0736924409866333,
"learning_rate": 3.8945116022240945e-06,
"loss": 0.1075,
"step": 282
},
{
"epoch": 2.482456140350877,
"grad_norm": 0.6335628032684326,
"learning_rate": 3.8833295145696964e-06,
"loss": 0.1036,
"step": 283
},
{
"epoch": 2.4912280701754383,
"grad_norm": 0.6909618377685547,
"learning_rate": 3.872107395711799e-06,
"loss": 0.1089,
"step": 284
},
{
"epoch": 2.5,
"grad_norm": 2.1871702671051025,
"learning_rate": 3.860845570399435e-06,
"loss": 0.1066,
"step": 285
},
{
"epoch": 2.5087719298245617,
"grad_norm": 0.5831722617149353,
"learning_rate": 3.849544364530678e-06,
"loss": 0.1055,
"step": 286
},
{
"epoch": 2.517543859649123,
"grad_norm": 0.5302637815475464,
"learning_rate": 3.838204105143204e-06,
"loss": 0.1057,
"step": 287
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.6348035931587219,
"learning_rate": 3.8268251204048335e-06,
"loss": 0.1089,
"step": 288
},
{
"epoch": 2.5350877192982457,
"grad_norm": 2.1932008266448975,
"learning_rate": 3.815407739604033e-06,
"loss": 0.1043,
"step": 289
},
{
"epoch": 2.543859649122807,
"grad_norm": 0.4388940930366516,
"learning_rate": 3.803952293140385e-06,
"loss": 0.1055,
"step": 290
},
{
"epoch": 2.5526315789473686,
"grad_norm": 0.6853339076042175,
"learning_rate": 3.7924591125150265e-06,
"loss": 0.1036,
"step": 291
},
{
"epoch": 2.56140350877193,
"grad_norm": 0.34744876623153687,
"learning_rate": 3.78092853032106e-06,
"loss": 0.1025,
"step": 292
},
{
"epoch": 2.5701754385964914,
"grad_norm": 0.9523847699165344,
"learning_rate": 3.769360880233922e-06,
"loss": 0.1067,
"step": 293
},
{
"epoch": 2.5789473684210527,
"grad_norm": 1.303745985031128,
"learning_rate": 3.7577564970017338e-06,
"loss": 0.1082,
"step": 294
},
{
"epoch": 2.587719298245614,
"grad_norm": 0.9468981623649597,
"learning_rate": 3.7461157164356103e-06,
"loss": 0.1055,
"step": 295
},
{
"epoch": 2.5964912280701755,
"grad_norm": 0.7204175591468811,
"learning_rate": 3.7344388753999434e-06,
"loss": 0.1055,
"step": 296
},
{
"epoch": 2.6052631578947367,
"grad_norm": 0.5110165476799011,
"learning_rate": 3.7227263118026537e-06,
"loss": 0.1092,
"step": 297
},
{
"epoch": 2.6140350877192984,
"grad_norm": 0.6483246088027954,
"learning_rate": 3.7109783645854116e-06,
"loss": 0.1078,
"step": 298
},
{
"epoch": 2.6228070175438596,
"grad_norm": 0.5058422684669495,
"learning_rate": 3.699195373713831e-06,
"loss": 0.1073,
"step": 299
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.4123518764972687,
"learning_rate": 3.6873776801676265e-06,
"loss": 0.1053,
"step": 300
},
{
"epoch": 2.6403508771929824,
"grad_norm": 1.0864709615707397,
"learning_rate": 3.675525625930751e-06,
"loss": 0.1048,
"step": 301
},
{
"epoch": 2.6491228070175437,
"grad_norm": 1.0264904499053955,
"learning_rate": 3.6636395539814975e-06,
"loss": 0.1059,
"step": 302
},
{
"epoch": 2.6578947368421053,
"grad_norm": 0.7724822163581848,
"learning_rate": 3.651719808282573e-06,
"loss": 0.1063,
"step": 303
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.7474755644798279,
"learning_rate": 3.6397667337711475e-06,
"loss": 0.1034,
"step": 304
},
{
"epoch": 2.675438596491228,
"grad_norm": 0.5628909468650818,
"learning_rate": 3.6277806763488666e-06,
"loss": 0.1026,
"step": 305
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.9070547819137573,
"learning_rate": 3.6157619828718477e-06,
"loss": 0.1031,
"step": 306
},
{
"epoch": 2.692982456140351,
"grad_norm": 0.6968091130256653,
"learning_rate": 3.603711001140641e-06,
"loss": 0.1068,
"step": 307
},
{
"epoch": 2.7017543859649122,
"grad_norm": 0.3764977753162384,
"learning_rate": 3.5916280798901604e-06,
"loss": 0.1038,
"step": 308
},
{
"epoch": 2.7105263157894735,
"grad_norm": 5.012625694274902,
"learning_rate": 3.5795135687795984e-06,
"loss": 0.1129,
"step": 309
},
{
"epoch": 2.719298245614035,
"grad_norm": 0.6745572686195374,
"learning_rate": 3.567367818382303e-06,
"loss": 0.1071,
"step": 310
},
{
"epoch": 2.7280701754385968,
"grad_norm": 1.0659606456756592,
"learning_rate": 3.555191180175634e-06,
"loss": 0.1067,
"step": 311
},
{
"epoch": 2.736842105263158,
"grad_norm": 1.7312604188919067,
"learning_rate": 3.5429840065307924e-06,
"loss": 0.1101,
"step": 312
},
{
"epoch": 2.745614035087719,
"grad_norm": 1.100364327430725,
"learning_rate": 3.5307466507026223e-06,
"loss": 0.1098,
"step": 313
},
{
"epoch": 2.754385964912281,
"grad_norm": 1.0390428304672241,
"learning_rate": 3.5184794668193893e-06,
"loss": 0.1094,
"step": 314
},
{
"epoch": 2.763157894736842,
"grad_norm": 0.3369971811771393,
"learning_rate": 3.5061828098725327e-06,
"loss": 0.1053,
"step": 315
},
{
"epoch": 2.7719298245614032,
"grad_norm": 0.6130257248878479,
"learning_rate": 3.4938570357063906e-06,
"loss": 0.106,
"step": 316
},
{
"epoch": 2.780701754385965,
"grad_norm": 0.6387595534324646,
"learning_rate": 3.481502501007904e-06,
"loss": 0.1044,
"step": 317
},
{
"epoch": 2.7894736842105265,
"grad_norm": 1.0731587409973145,
"learning_rate": 3.469119563296296e-06,
"loss": 0.1097,
"step": 318
},
{
"epoch": 2.7982456140350878,
"grad_norm": 0.8096229434013367,
"learning_rate": 3.4567085809127247e-06,
"loss": 0.1076,
"step": 319
},
{
"epoch": 2.807017543859649,
"grad_norm": 0.5034844279289246,
"learning_rate": 3.444269913009912e-06,
"loss": 0.1071,
"step": 320
},
{
"epoch": 2.8157894736842106,
"grad_norm": 0.675139307975769,
"learning_rate": 3.4318039195417536e-06,
"loss": 0.1039,
"step": 321
},
{
"epoch": 2.824561403508772,
"grad_norm": 0.7330355644226074,
"learning_rate": 3.4193109612528972e-06,
"loss": 0.1044,
"step": 322
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.6558271646499634,
"learning_rate": 3.4067913996683115e-06,
"loss": 0.1051,
"step": 323
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.8411844372749329,
"learning_rate": 3.3942455970828146e-06,
"loss": 0.1063,
"step": 324
},
{
"epoch": 2.8508771929824563,
"grad_norm": 0.4817325174808502,
"learning_rate": 3.3816739165505964e-06,
"loss": 0.105,
"step": 325
},
{
"epoch": 2.8596491228070176,
"grad_norm": 0.424554705619812,
"learning_rate": 3.3690767218747104e-06,
"loss": 0.1037,
"step": 326
},
{
"epoch": 2.8684210526315788,
"grad_norm": 1.0054417848587036,
"learning_rate": 3.3564543775965475e-06,
"loss": 0.1058,
"step": 327
},
{
"epoch": 2.8771929824561404,
"grad_norm": 0.8984584808349609,
"learning_rate": 3.3438072489852837e-06,
"loss": 0.1079,
"step": 328
},
{
"epoch": 2.8859649122807016,
"grad_norm": 0.6779558062553406,
"learning_rate": 3.331135702027311e-06,
"loss": 0.1046,
"step": 329
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.6931657195091248,
"learning_rate": 3.318440103415649e-06,
"loss": 0.1106,
"step": 330
},
{
"epoch": 2.9035087719298245,
"grad_norm": 0.705264151096344,
"learning_rate": 3.305720820539329e-06,
"loss": 0.104,
"step": 331
},
{
"epoch": 2.912280701754386,
"grad_norm": 0.7799407839775085,
"learning_rate": 3.2929782214727657e-06,
"loss": 0.1019,
"step": 332
},
{
"epoch": 2.9210526315789473,
"grad_norm": 0.7583760619163513,
"learning_rate": 3.2802126749651042e-06,
"loss": 0.1049,
"step": 333
},
{
"epoch": 2.9298245614035086,
"grad_norm": 0.6145837306976318,
"learning_rate": 3.2674245504295505e-06,
"loss": 0.104,
"step": 334
},
{
"epoch": 2.93859649122807,
"grad_norm": 0.5170779228210449,
"learning_rate": 3.254614217932679e-06,
"loss": 0.1024,
"step": 335
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.6850940585136414,
"learning_rate": 3.241782048183726e-06,
"loss": 0.1047,
"step": 336
},
{
"epoch": 2.956140350877193,
"grad_norm": 0.7307694554328918,
"learning_rate": 3.2289284125238597e-06,
"loss": 0.1032,
"step": 337
},
{
"epoch": 2.9649122807017543,
"grad_norm": 0.3386179208755493,
"learning_rate": 3.216053682915436e-06,
"loss": 0.1037,
"step": 338
},
{
"epoch": 2.973684210526316,
"grad_norm": 0.7565059065818787,
"learning_rate": 3.203158231931234e-06,
"loss": 0.1048,
"step": 339
},
{
"epoch": 2.982456140350877,
"grad_norm": 0.7902039289474487,
"learning_rate": 3.190242432743673e-06,
"loss": 0.1068,
"step": 340
},
{
"epoch": 2.9912280701754383,
"grad_norm": 0.42595192790031433,
"learning_rate": 3.177306659114015e-06,
"loss": 0.1039,
"step": 341
},
{
"epoch": 3.0,
"grad_norm": 1.1214542388916016,
"learning_rate": 3.164351285381549e-06,
"loss": 0.1062,
"step": 342
}
],
"logging_steps": 1,
"max_steps": 684,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 114,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.037997022244556e+19,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}