Instructions to use ulas742/lora-mid with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use ulas742/lora-mid with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.2-1B") model = PeftModel.from_pretrained(base_model, "ulas742/lora-mid") - Notebooks
- Google Colab
- Kaggle
Download checkpoint-5000/trainer_state.json from ulas742/lora-mid: direct link, hf CLI and curl.
- Browser
- Download file 162 kB
-
https://huggingface.co/ulas742/lora-mid/resolve/main/checkpoint-5000/trainer_state.json
- Command line
-
hf download hf://ulas742/lora-mid/checkpoint-5000/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/ulas742/lora-mid/resolve/main/checkpoint-5000/trainer_state.json
162 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001,
"grad_norm": 133.26564025878906,
"learning_rate": 1.6000000000000001e-06,
"loss": 319.5714,
"step": 5
},
{
"epoch": 0.002,
"grad_norm": 116.60221099853516,
"learning_rate": 3.6e-06,
"loss": 321.6409,
"step": 10
},
{
"epoch": 0.003,
"grad_norm": 99.86441040039062,
"learning_rate": 5.600000000000001e-06,
"loss": 321.4713,
"step": 15
},
{
"epoch": 0.004,
"grad_norm": 120.4101333618164,
"learning_rate": 7.6e-06,
"loss": 326.6742,
"step": 20
},
{
"epoch": 0.005,
"grad_norm": 131.2949676513672,
"learning_rate": 9.600000000000001e-06,
"loss": 325.8993,
"step": 25
},
{
"epoch": 0.006,
"grad_norm": 109.72811889648438,
"learning_rate": 1.16e-05,
"loss": 313.0747,
"step": 30
},
{
"epoch": 0.007,
"grad_norm": 93.34612274169922,
"learning_rate": 1.3600000000000002e-05,
"loss": 315.1154,
"step": 35
},
{
"epoch": 0.008,
"grad_norm": 89.763427734375,
"learning_rate": 1.56e-05,
"loss": 307.5251,
"step": 40
},
{
"epoch": 0.009,
"grad_norm": 79.69110870361328,
"learning_rate": 1.76e-05,
"loss": 310.1657,
"step": 45
},
{
"epoch": 0.01,
"grad_norm": 83.45806121826172,
"learning_rate": 1.9600000000000002e-05,
"loss": 322.3456,
"step": 50
},
{
"epoch": 0.011,
"grad_norm": 83.1032943725586,
"learning_rate": 2.16e-05,
"loss": 307.5482,
"step": 55
},
{
"epoch": 0.012,
"grad_norm": 100.54994201660156,
"learning_rate": 2.36e-05,
"loss": 308.4983,
"step": 60
},
{
"epoch": 0.013,
"grad_norm": 74.47264099121094,
"learning_rate": 2.5600000000000002e-05,
"loss": 313.1681,
"step": 65
},
{
"epoch": 0.014,
"grad_norm": 85.1391372680664,
"learning_rate": 2.7600000000000003e-05,
"loss": 302.5862,
"step": 70
},
{
"epoch": 0.015,
"grad_norm": 86.149169921875,
"learning_rate": 2.96e-05,
"loss": 296.1115,
"step": 75
},
{
"epoch": 0.016,
"grad_norm": 82.54764556884766,
"learning_rate": 3.16e-05,
"loss": 299.7468,
"step": 80
},
{
"epoch": 0.017,
"grad_norm": 76.23515319824219,
"learning_rate": 3.3600000000000004e-05,
"loss": 314.8733,
"step": 85
},
{
"epoch": 0.018,
"grad_norm": 86.42366027832031,
"learning_rate": 3.56e-05,
"loss": 309.0613,
"step": 90
},
{
"epoch": 0.019,
"grad_norm": 79.41493225097656,
"learning_rate": 3.76e-05,
"loss": 311.0322,
"step": 95
},
{
"epoch": 0.02,
"grad_norm": 97.31949615478516,
"learning_rate": 3.960000000000001e-05,
"loss": 303.7566,
"step": 100
},
{
"epoch": 0.021,
"grad_norm": 91.24207305908203,
"learning_rate": 4.16e-05,
"loss": 307.9841,
"step": 105
},
{
"epoch": 0.022,
"grad_norm": 82.55218505859375,
"learning_rate": 4.36e-05,
"loss": 303.4123,
"step": 110
},
{
"epoch": 0.023,
"grad_norm": 98.32549285888672,
"learning_rate": 4.5600000000000004e-05,
"loss": 302.9958,
"step": 115
},
{
"epoch": 0.024,
"grad_norm": 90.6205062866211,
"learning_rate": 4.76e-05,
"loss": 306.3299,
"step": 120
},
{
"epoch": 0.025,
"grad_norm": 88.87057495117188,
"learning_rate": 4.96e-05,
"loss": 299.5162,
"step": 125
},
{
"epoch": 0.026,
"grad_norm": 84.20787048339844,
"learning_rate": 5.16e-05,
"loss": 302.7059,
"step": 130
},
{
"epoch": 0.027,
"grad_norm": 95.52533721923828,
"learning_rate": 5.360000000000001e-05,
"loss": 302.2274,
"step": 135
},
{
"epoch": 0.028,
"grad_norm": 117.1716537475586,
"learning_rate": 5.560000000000001e-05,
"loss": 299.0743,
"step": 140
},
{
"epoch": 0.029,
"grad_norm": 77.68347930908203,
"learning_rate": 5.76e-05,
"loss": 295.1787,
"step": 145
},
{
"epoch": 0.03,
"grad_norm": 76.69217681884766,
"learning_rate": 5.96e-05,
"loss": 303.3868,
"step": 150
},
{
"epoch": 0.031,
"grad_norm": 91.29154205322266,
"learning_rate": 6.16e-05,
"loss": 285.78,
"step": 155
},
{
"epoch": 0.032,
"grad_norm": 80.40809631347656,
"learning_rate": 6.36e-05,
"loss": 298.3949,
"step": 160
},
{
"epoch": 0.033,
"grad_norm": 91.36680603027344,
"learning_rate": 6.560000000000001e-05,
"loss": 304.5564,
"step": 165
},
{
"epoch": 0.034,
"grad_norm": 112.79888916015625,
"learning_rate": 6.76e-05,
"loss": 242.7879,
"step": 170
},
{
"epoch": 0.035,
"grad_norm": 98.3003158569336,
"learning_rate": 6.96e-05,
"loss": 239.308,
"step": 175
},
{
"epoch": 0.036,
"grad_norm": 88.80140686035156,
"learning_rate": 7.16e-05,
"loss": 280.2297,
"step": 180
},
{
"epoch": 0.037,
"grad_norm": 75.21417236328125,
"learning_rate": 7.36e-05,
"loss": 285.2648,
"step": 185
},
{
"epoch": 0.038,
"grad_norm": 76.6654052734375,
"learning_rate": 7.560000000000001e-05,
"loss": 297.0549,
"step": 190
},
{
"epoch": 0.039,
"grad_norm": 101.13793182373047,
"learning_rate": 7.76e-05,
"loss": 287.9887,
"step": 195
},
{
"epoch": 0.04,
"grad_norm": 79.47496032714844,
"learning_rate": 7.960000000000001e-05,
"loss": 267.6669,
"step": 200
},
{
"epoch": 0.041,
"grad_norm": 86.7893295288086,
"learning_rate": 8.16e-05,
"loss": 301.4663,
"step": 205
},
{
"epoch": 0.042,
"grad_norm": 98.68603515625,
"learning_rate": 8.36e-05,
"loss": 297.1799,
"step": 210
},
{
"epoch": 0.043,
"grad_norm": 69.82585144042969,
"learning_rate": 8.560000000000001e-05,
"loss": 284.9405,
"step": 215
},
{
"epoch": 0.044,
"grad_norm": 77.55110931396484,
"learning_rate": 8.76e-05,
"loss": 289.2016,
"step": 220
},
{
"epoch": 0.045,
"grad_norm": 67.33943176269531,
"learning_rate": 8.960000000000001e-05,
"loss": 289.2506,
"step": 225
},
{
"epoch": 0.046,
"grad_norm": 66.90230560302734,
"learning_rate": 9.16e-05,
"loss": 297.8193,
"step": 230
},
{
"epoch": 0.047,
"grad_norm": 280.9183044433594,
"learning_rate": 9.360000000000001e-05,
"loss": 299.363,
"step": 235
},
{
"epoch": 0.048,
"grad_norm": 87.29924011230469,
"learning_rate": 9.56e-05,
"loss": 285.1276,
"step": 240
},
{
"epoch": 0.049,
"grad_norm": 69.45281982421875,
"learning_rate": 9.76e-05,
"loss": 291.0465,
"step": 245
},
{
"epoch": 0.05,
"grad_norm": 95.85892486572266,
"learning_rate": 9.960000000000001e-05,
"loss": 253.5561,
"step": 250
},
{
"epoch": 0.051,
"grad_norm": 87.18708801269531,
"learning_rate": 0.0001016,
"loss": 278.3737,
"step": 255
},
{
"epoch": 0.052,
"grad_norm": 59.54637145996094,
"learning_rate": 0.00010360000000000001,
"loss": 284.0256,
"step": 260
},
{
"epoch": 0.053,
"grad_norm": 75.85100555419922,
"learning_rate": 0.0001056,
"loss": 288.8852,
"step": 265
},
{
"epoch": 0.054,
"grad_norm": 79.74699401855469,
"learning_rate": 0.00010760000000000001,
"loss": 235.3202,
"step": 270
},
{
"epoch": 0.055,
"grad_norm": 93.92021942138672,
"learning_rate": 0.00010960000000000001,
"loss": 291.399,
"step": 275
},
{
"epoch": 0.056,
"grad_norm": 73.23931121826172,
"learning_rate": 0.00011160000000000002,
"loss": 297.0533,
"step": 280
},
{
"epoch": 0.057,
"grad_norm": 74.23943328857422,
"learning_rate": 0.0001136,
"loss": 301.6619,
"step": 285
},
{
"epoch": 0.058,
"grad_norm": 76.8428726196289,
"learning_rate": 0.00011559999999999999,
"loss": 297.2145,
"step": 290
},
{
"epoch": 0.059,
"grad_norm": 63.19135284423828,
"learning_rate": 0.0001176,
"loss": 259.2088,
"step": 295
},
{
"epoch": 0.06,
"grad_norm": 74.82901000976562,
"learning_rate": 0.00011960000000000001,
"loss": 269.6802,
"step": 300
},
{
"epoch": 0.061,
"grad_norm": 118.87765502929688,
"learning_rate": 0.0001216,
"loss": 254.0412,
"step": 305
},
{
"epoch": 0.062,
"grad_norm": 74.98046875,
"learning_rate": 0.0001236,
"loss": 294.7132,
"step": 310
},
{
"epoch": 0.063,
"grad_norm": 70.42806243896484,
"learning_rate": 0.00012560000000000002,
"loss": 256.933,
"step": 315
},
{
"epoch": 0.064,
"grad_norm": 86.4557113647461,
"learning_rate": 0.0001276,
"loss": 235.0536,
"step": 320
},
{
"epoch": 0.065,
"grad_norm": 62.878108978271484,
"learning_rate": 0.0001296,
"loss": 255.1298,
"step": 325
},
{
"epoch": 0.066,
"grad_norm": 62.64030838012695,
"learning_rate": 0.0001316,
"loss": 286.0793,
"step": 330
},
{
"epoch": 0.067,
"grad_norm": 82.25812530517578,
"learning_rate": 0.00013360000000000002,
"loss": 255.7905,
"step": 335
},
{
"epoch": 0.068,
"grad_norm": 67.82426452636719,
"learning_rate": 0.00013560000000000002,
"loss": 286.4674,
"step": 340
},
{
"epoch": 0.069,
"grad_norm": 69.97148132324219,
"learning_rate": 0.00013759999999999998,
"loss": 287.4435,
"step": 345
},
{
"epoch": 0.07,
"grad_norm": 127.98161315917969,
"learning_rate": 0.0001396,
"loss": 273.2139,
"step": 350
},
{
"epoch": 0.071,
"grad_norm": 92.13734436035156,
"learning_rate": 0.0001416,
"loss": 283.4005,
"step": 355
},
{
"epoch": 0.072,
"grad_norm": 142.50759887695312,
"learning_rate": 0.0001436,
"loss": 267.1752,
"step": 360
},
{
"epoch": 0.073,
"grad_norm": 110.20304107666016,
"learning_rate": 0.00014560000000000002,
"loss": 248.1763,
"step": 365
},
{
"epoch": 0.074,
"grad_norm": 76.13386535644531,
"learning_rate": 0.0001476,
"loss": 286.8581,
"step": 370
},
{
"epoch": 0.075,
"grad_norm": 169.9783172607422,
"learning_rate": 0.0001496,
"loss": 273.6802,
"step": 375
},
{
"epoch": 0.076,
"grad_norm": 88.17061614990234,
"learning_rate": 0.0001516,
"loss": 279.6191,
"step": 380
},
{
"epoch": 0.077,
"grad_norm": 85.39425659179688,
"learning_rate": 0.00015360000000000002,
"loss": 247.7376,
"step": 385
},
{
"epoch": 0.078,
"grad_norm": 71.32750701904297,
"learning_rate": 0.00015560000000000001,
"loss": 271.884,
"step": 390
},
{
"epoch": 0.079,
"grad_norm": 65.19810485839844,
"learning_rate": 0.0001576,
"loss": 277.6798,
"step": 395
},
{
"epoch": 0.08,
"grad_norm": 94.19916534423828,
"learning_rate": 0.0001596,
"loss": 276.7407,
"step": 400
},
{
"epoch": 0.081,
"grad_norm": 133.42376708984375,
"learning_rate": 0.00016160000000000002,
"loss": 230.0227,
"step": 405
},
{
"epoch": 0.082,
"grad_norm": 74.1983642578125,
"learning_rate": 0.0001636,
"loss": 254.8563,
"step": 410
},
{
"epoch": 0.083,
"grad_norm": 81.3104248046875,
"learning_rate": 0.0001656,
"loss": 266.7406,
"step": 415
},
{
"epoch": 0.084,
"grad_norm": 79.57012176513672,
"learning_rate": 0.0001676,
"loss": 250.5589,
"step": 420
},
{
"epoch": 0.085,
"grad_norm": 77.29656219482422,
"learning_rate": 0.0001696,
"loss": 282.2854,
"step": 425
},
{
"epoch": 0.086,
"grad_norm": 79.8593978881836,
"learning_rate": 0.0001716,
"loss": 252.3088,
"step": 430
},
{
"epoch": 0.087,
"grad_norm": 60.186832427978516,
"learning_rate": 0.00017360000000000002,
"loss": 286.7576,
"step": 435
},
{
"epoch": 0.088,
"grad_norm": 65.36947631835938,
"learning_rate": 0.0001756,
"loss": 281.5767,
"step": 440
},
{
"epoch": 0.089,
"grad_norm": 75.44194030761719,
"learning_rate": 0.0001776,
"loss": 278.4216,
"step": 445
},
{
"epoch": 0.09,
"grad_norm": 66.71389770507812,
"learning_rate": 0.0001796,
"loss": 270.1713,
"step": 450
},
{
"epoch": 0.091,
"grad_norm": 68.38360595703125,
"learning_rate": 0.00018160000000000002,
"loss": 275.1372,
"step": 455
},
{
"epoch": 0.092,
"grad_norm": 72.85200500488281,
"learning_rate": 0.00018360000000000002,
"loss": 281.6097,
"step": 460
},
{
"epoch": 0.093,
"grad_norm": 65.46351623535156,
"learning_rate": 0.0001856,
"loss": 285.4045,
"step": 465
},
{
"epoch": 0.094,
"grad_norm": 77.94449615478516,
"learning_rate": 0.0001876,
"loss": 275.6515,
"step": 470
},
{
"epoch": 0.095,
"grad_norm": 104.31855773925781,
"learning_rate": 0.0001896,
"loss": 252.4393,
"step": 475
},
{
"epoch": 0.096,
"grad_norm": 85.45411682128906,
"learning_rate": 0.0001916,
"loss": 260.0067,
"step": 480
},
{
"epoch": 0.097,
"grad_norm": 71.21846008300781,
"learning_rate": 0.00019360000000000002,
"loss": 282.9384,
"step": 485
},
{
"epoch": 0.098,
"grad_norm": 66.82077026367188,
"learning_rate": 0.0001956,
"loss": 268.286,
"step": 490
},
{
"epoch": 0.099,
"grad_norm": 66.0440673828125,
"learning_rate": 0.0001976,
"loss": 280.2809,
"step": 495
},
{
"epoch": 0.1,
"grad_norm": 65.30158996582031,
"learning_rate": 0.0001996,
"loss": 277.9199,
"step": 500
},
{
"epoch": 0.101,
"grad_norm": 69.67945098876953,
"learning_rate": 0.00019999961008995606,
"loss": 289.9011,
"step": 505
},
{
"epoch": 0.102,
"grad_norm": 67.6677474975586,
"learning_rate": 0.0001999980260856137,
"loss": 277.9196,
"step": 510
},
{
"epoch": 0.103,
"grad_norm": 74.7071762084961,
"learning_rate": 0.000199995223636881,
"loss": 262.7933,
"step": 515
},
{
"epoch": 0.104,
"grad_norm": 69.05998992919922,
"learning_rate": 0.00019999120277790477,
"loss": 287.7239,
"step": 520
},
{
"epoch": 0.105,
"grad_norm": 69.93698120117188,
"learning_rate": 0.00019998596355767805,
"loss": 275.5523,
"step": 525
},
{
"epoch": 0.106,
"grad_norm": 61.50294876098633,
"learning_rate": 0.000199979506040039,
"loss": 280.7647,
"step": 530
},
{
"epoch": 0.107,
"grad_norm": 68.64286041259766,
"learning_rate": 0.00019997183030367048,
"loss": 271.8449,
"step": 535
},
{
"epoch": 0.108,
"grad_norm": 63.09992980957031,
"learning_rate": 0.00019996293644209887,
"loss": 278.2004,
"step": 540
},
{
"epoch": 0.109,
"grad_norm": 71.78915405273438,
"learning_rate": 0.0001999528245636931,
"loss": 278.1076,
"step": 545
},
{
"epoch": 0.11,
"grad_norm": 119.38697052001953,
"learning_rate": 0.00019994149479166324,
"loss": 250.7288,
"step": 550
},
{
"epoch": 0.111,
"grad_norm": 68.61959075927734,
"learning_rate": 0.00019992894726405893,
"loss": 210.8162,
"step": 555
},
{
"epoch": 0.112,
"grad_norm": 76.88066864013672,
"learning_rate": 0.00019991518213376787,
"loss": 280.71,
"step": 560
},
{
"epoch": 0.113,
"grad_norm": 145.10508728027344,
"learning_rate": 0.00019990019956851382,
"loss": 267.4712,
"step": 565
},
{
"epoch": 0.114,
"grad_norm": 60.84817123413086,
"learning_rate": 0.0001998839997508546,
"loss": 271.589,
"step": 570
},
{
"epoch": 0.115,
"grad_norm": 68.52861785888672,
"learning_rate": 0.00019986658287817987,
"loss": 264.3261,
"step": 575
},
{
"epoch": 0.116,
"grad_norm": 74.1113052368164,
"learning_rate": 0.00019984794916270876,
"loss": 246.1759,
"step": 580
},
{
"epoch": 0.117,
"grad_norm": 72.84656524658203,
"learning_rate": 0.00019982809883148722,
"loss": 260.5045,
"step": 585
},
{
"epoch": 0.118,
"grad_norm": 70.28099822998047,
"learning_rate": 0.00019980703212638522,
"loss": 240.8895,
"step": 590
},
{
"epoch": 0.119,
"grad_norm": 72.57321166992188,
"learning_rate": 0.00019978474930409395,
"loss": 267.5912,
"step": 595
},
{
"epoch": 0.12,
"grad_norm": 82.38754272460938,
"learning_rate": 0.00019976125063612252,
"loss": 260.8692,
"step": 600
},
{
"epoch": 0.121,
"grad_norm": 62.69374084472656,
"learning_rate": 0.00019973653640879485,
"loss": 267.0439,
"step": 605
},
{
"epoch": 0.122,
"grad_norm": 183.09005737304688,
"learning_rate": 0.00019971060692324598,
"loss": 261.9547,
"step": 610
},
{
"epoch": 0.123,
"grad_norm": 77.94440460205078,
"learning_rate": 0.00019968346249541846,
"loss": 262.4955,
"step": 615
},
{
"epoch": 0.124,
"grad_norm": 303.3373107910156,
"learning_rate": 0.00019965510345605866,
"loss": 259.7711,
"step": 620
},
{
"epoch": 0.125,
"grad_norm": 74.0142822265625,
"learning_rate": 0.0001996255301507125,
"loss": 270.3866,
"step": 625
},
{
"epoch": 0.126,
"grad_norm": 71.14913177490234,
"learning_rate": 0.00019959474293972129,
"loss": 272.4547,
"step": 630
},
{
"epoch": 0.127,
"grad_norm": 135.9524688720703,
"learning_rate": 0.00019956274219821757,
"loss": 253.1282,
"step": 635
},
{
"epoch": 0.128,
"grad_norm": 66.36004638671875,
"learning_rate": 0.00019952952831612026,
"loss": 273.6564,
"step": 640
},
{
"epoch": 0.129,
"grad_norm": 70.32064056396484,
"learning_rate": 0.00019949510169813003,
"loss": 280.9516,
"step": 645
},
{
"epoch": 0.13,
"grad_norm": 59.25802993774414,
"learning_rate": 0.00019945946276372434,
"loss": 254.8434,
"step": 650
},
{
"epoch": 0.131,
"grad_norm": 70.53588104248047,
"learning_rate": 0.00019942261194715236,
"loss": 245.8175,
"step": 655
},
{
"epoch": 0.132,
"grad_norm": 69.76761627197266,
"learning_rate": 0.00019938454969742968,
"loss": 275.741,
"step": 660
},
{
"epoch": 0.133,
"grad_norm": 64.74825286865234,
"learning_rate": 0.00019934527647833276,
"loss": 266.5821,
"step": 665
},
{
"epoch": 0.134,
"grad_norm": 65.31316375732422,
"learning_rate": 0.00019930479276839344,
"loss": 275.6624,
"step": 670
},
{
"epoch": 0.135,
"grad_norm": 62.38485336303711,
"learning_rate": 0.0001992630990608929,
"loss": 261.4533,
"step": 675
},
{
"epoch": 0.136,
"grad_norm": 71.19464111328125,
"learning_rate": 0.00019922019586385585,
"loss": 262.5512,
"step": 680
},
{
"epoch": 0.137,
"grad_norm": 168.85684204101562,
"learning_rate": 0.00019917608370004417,
"loss": 234.3873,
"step": 685
},
{
"epoch": 0.138,
"grad_norm": 75.73576354980469,
"learning_rate": 0.00019913076310695068,
"loss": 231.9306,
"step": 690
},
{
"epoch": 0.139,
"grad_norm": 143.76568603515625,
"learning_rate": 0.00019908423463679248,
"loss": 264.1436,
"step": 695
},
{
"epoch": 0.14,
"grad_norm": 83.88737487792969,
"learning_rate": 0.0001990364988565043,
"loss": 281.3076,
"step": 700
},
{
"epoch": 0.141,
"grad_norm": 70.77721405029297,
"learning_rate": 0.00019898755634773158,
"loss": 268.7908,
"step": 705
},
{
"epoch": 0.142,
"grad_norm": 84.32762145996094,
"learning_rate": 0.00019893740770682335,
"loss": 269.243,
"step": 710
},
{
"epoch": 0.143,
"grad_norm": 60.8754768371582,
"learning_rate": 0.0001988860535448249,
"loss": 247.6113,
"step": 715
},
{
"epoch": 0.144,
"grad_norm": 72.7818603515625,
"learning_rate": 0.00019883349448747062,
"loss": 268.8609,
"step": 720
},
{
"epoch": 0.145,
"grad_norm": 88.45670318603516,
"learning_rate": 0.0001987797311751759,
"loss": 268.0048,
"step": 725
},
{
"epoch": 0.146,
"grad_norm": 72.0194320678711,
"learning_rate": 0.00019872476426302982,
"loss": 275.5222,
"step": 730
},
{
"epoch": 0.147,
"grad_norm": 62.675167083740234,
"learning_rate": 0.0001986685944207868,
"loss": 267.5268,
"step": 735
},
{
"epoch": 0.148,
"grad_norm": 82.43748474121094,
"learning_rate": 0.0001986112223328587,
"loss": 274.9757,
"step": 740
},
{
"epoch": 0.149,
"grad_norm": 71.96373748779297,
"learning_rate": 0.00019855264869830629,
"loss": 264.831,
"step": 745
},
{
"epoch": 0.15,
"grad_norm": 66.41895294189453,
"learning_rate": 0.00019849287423083078,
"loss": 263.1631,
"step": 750
},
{
"epoch": 0.151,
"grad_norm": 87.85332489013672,
"learning_rate": 0.00019843189965876526,
"loss": 260.5555,
"step": 755
},
{
"epoch": 0.152,
"grad_norm": 74.49442291259766,
"learning_rate": 0.00019836972572506557,
"loss": 275.7229,
"step": 760
},
{
"epoch": 0.153,
"grad_norm": 77.08104705810547,
"learning_rate": 0.00019830635318730154,
"loss": 256.9746,
"step": 765
},
{
"epoch": 0.154,
"grad_norm": 73.9883041381836,
"learning_rate": 0.00019824178281764753,
"loss": 276.2443,
"step": 770
},
{
"epoch": 0.155,
"grad_norm": 73.93590545654297,
"learning_rate": 0.00019817601540287306,
"loss": 254.5362,
"step": 775
},
{
"epoch": 0.156,
"grad_norm": 68.59113311767578,
"learning_rate": 0.0001981090517443334,
"loss": 262.6415,
"step": 780
},
{
"epoch": 0.157,
"grad_norm": 61.02271270751953,
"learning_rate": 0.0001980408926579596,
"loss": 252.363,
"step": 785
},
{
"epoch": 0.158,
"grad_norm": 62.54066848754883,
"learning_rate": 0.00019797153897424852,
"loss": 268.4424,
"step": 790
},
{
"epoch": 0.159,
"grad_norm": 63.734649658203125,
"learning_rate": 0.000197900991538253,
"loss": 255.699,
"step": 795
},
{
"epoch": 0.16,
"grad_norm": 487.8463134765625,
"learning_rate": 0.00019782925120957124,
"loss": 268.1905,
"step": 800
},
{
"epoch": 0.161,
"grad_norm": 70.30332946777344,
"learning_rate": 0.00019775631886233654,
"loss": 260.2901,
"step": 805
},
{
"epoch": 0.162,
"grad_norm": 81.92864990234375,
"learning_rate": 0.0001976821953852065,
"loss": 257.2851,
"step": 810
},
{
"epoch": 0.163,
"grad_norm": 72.4359130859375,
"learning_rate": 0.00019760688168135232,
"loss": 258.2725,
"step": 815
},
{
"epoch": 0.164,
"grad_norm": 72.5333023071289,
"learning_rate": 0.00019753037866844771,
"loss": 245.3567,
"step": 820
},
{
"epoch": 0.165,
"grad_norm": 70.82955932617188,
"learning_rate": 0.00019745268727865774,
"loss": 247.9339,
"step": 825
},
{
"epoch": 0.166,
"grad_norm": 66.89386749267578,
"learning_rate": 0.00019737380845862745,
"loss": 267.3392,
"step": 830
},
{
"epoch": 0.167,
"grad_norm": 62.468936920166016,
"learning_rate": 0.0001972937431694704,
"loss": 249.5989,
"step": 835
},
{
"epoch": 0.168,
"grad_norm": 66.95137786865234,
"learning_rate": 0.00019721249238675688,
"loss": 256.6585,
"step": 840
},
{
"epoch": 0.169,
"grad_norm": 69.7963638305664,
"learning_rate": 0.000197130057100502,
"loss": 266.6019,
"step": 845
},
{
"epoch": 0.17,
"grad_norm": 64.68950653076172,
"learning_rate": 0.00019704643831515374,
"loss": 268.029,
"step": 850
},
{
"epoch": 0.171,
"grad_norm": 71.6061019897461,
"learning_rate": 0.0001969616370495806,
"loss": 261.1283,
"step": 855
},
{
"epoch": 0.172,
"grad_norm": 62.72983932495117,
"learning_rate": 0.00019687565433705926,
"loss": 267.8803,
"step": 860
},
{
"epoch": 0.173,
"grad_norm": 64.09201049804688,
"learning_rate": 0.00019678849122526187,
"loss": 243.9025,
"step": 865
},
{
"epoch": 0.174,
"grad_norm": 61.98489761352539,
"learning_rate": 0.00019670014877624353,
"loss": 241.4365,
"step": 870
},
{
"epoch": 0.175,
"grad_norm": 59.48838424682617,
"learning_rate": 0.00019661062806642903,
"loss": 250.3907,
"step": 875
},
{
"epoch": 0.176,
"grad_norm": 70.55399322509766,
"learning_rate": 0.0001965199301866,
"loss": 242.1836,
"step": 880
},
{
"epoch": 0.177,
"grad_norm": 63.19829177856445,
"learning_rate": 0.00019642805624188147,
"loss": 246.7926,
"step": 885
},
{
"epoch": 0.178,
"grad_norm": 70.59349060058594,
"learning_rate": 0.0001963350073517285,
"loss": 235.0772,
"step": 890
},
{
"epoch": 0.179,
"grad_norm": 70.43497467041016,
"learning_rate": 0.0001962407846499124,
"loss": 246.707,
"step": 895
},
{
"epoch": 0.18,
"grad_norm": 85.28739166259766,
"learning_rate": 0.0001961453892845071,
"loss": 240.7082,
"step": 900
},
{
"epoch": 0.181,
"grad_norm": 71.00152587890625,
"learning_rate": 0.00019604882241787498,
"loss": 241.3172,
"step": 905
},
{
"epoch": 0.182,
"grad_norm": 70.0228042602539,
"learning_rate": 0.0001959510852266529,
"loss": 259.7299,
"step": 910
},
{
"epoch": 0.183,
"grad_norm": 74.89047241210938,
"learning_rate": 0.0001958521789017376,
"loss": 258.1866,
"step": 915
},
{
"epoch": 0.184,
"grad_norm": 69.18814086914062,
"learning_rate": 0.00019575210464827149,
"loss": 270.8136,
"step": 920
},
{
"epoch": 0.185,
"grad_norm": 69.73190307617188,
"learning_rate": 0.0001956508636856278,
"loss": 258.3369,
"step": 925
},
{
"epoch": 0.186,
"grad_norm": 66.07145690917969,
"learning_rate": 0.00019554845724739566,
"loss": 251.0459,
"step": 930
},
{
"epoch": 0.187,
"grad_norm": 85.49530029296875,
"learning_rate": 0.00019544488658136523,
"loss": 266.0341,
"step": 935
},
{
"epoch": 0.188,
"grad_norm": 68.31378173828125,
"learning_rate": 0.00019534015294951233,
"loss": 264.6955,
"step": 940
},
{
"epoch": 0.189,
"grad_norm": 82.6740951538086,
"learning_rate": 0.00019523425762798329,
"loss": 266.1459,
"step": 945
},
{
"epoch": 0.19,
"grad_norm": 83.74392700195312,
"learning_rate": 0.00019512720190707913,
"loss": 247.1594,
"step": 950
},
{
"epoch": 0.191,
"grad_norm": 83.73538208007812,
"learning_rate": 0.00019501898709124008,
"loss": 260.7229,
"step": 955
},
{
"epoch": 0.192,
"grad_norm": 66.07109069824219,
"learning_rate": 0.00019490961449902946,
"loss": 263.6804,
"step": 960
},
{
"epoch": 0.193,
"grad_norm": 63.7003059387207,
"learning_rate": 0.00019479908546311781,
"loss": 264.5129,
"step": 965
},
{
"epoch": 0.194,
"grad_norm": 65.40355682373047,
"learning_rate": 0.0001946874013302666,
"loss": 260.6219,
"step": 970
},
{
"epoch": 0.195,
"grad_norm": 65.47498321533203,
"learning_rate": 0.0001945745634613117,
"loss": 263.1286,
"step": 975
},
{
"epoch": 0.196,
"grad_norm": 68.73960876464844,
"learning_rate": 0.0001944605732311469,
"loss": 265.6815,
"step": 980
},
{
"epoch": 0.197,
"grad_norm": 84.80562591552734,
"learning_rate": 0.00019434543202870725,
"loss": 253.9608,
"step": 985
},
{
"epoch": 0.198,
"grad_norm": 66.91242218017578,
"learning_rate": 0.0001942291412569519,
"loss": 262.0103,
"step": 990
},
{
"epoch": 0.199,
"grad_norm": 61.25504684448242,
"learning_rate": 0.00019411170233284727,
"loss": 269.1613,
"step": 995
},
{
"epoch": 0.2,
"grad_norm": 56.45645523071289,
"learning_rate": 0.00019399311668734956,
"loss": 264.6178,
"step": 1000
},
{
"epoch": 0.201,
"grad_norm": 63.47613525390625,
"learning_rate": 0.00019387338576538744,
"loss": 270.6375,
"step": 1005
},
{
"epoch": 0.202,
"grad_norm": 71.85877227783203,
"learning_rate": 0.0001937525110258444,
"loss": 259.4103,
"step": 1010
},
{
"epoch": 0.203,
"grad_norm": 70.76119995117188,
"learning_rate": 0.00019363049394154094,
"loss": 267.0952,
"step": 1015
},
{
"epoch": 0.204,
"grad_norm": 62.38040542602539,
"learning_rate": 0.00019350733599921683,
"loss": 251.4135,
"step": 1020
},
{
"epoch": 0.205,
"grad_norm": 72.8625259399414,
"learning_rate": 0.00019338303869951269,
"loss": 256.7074,
"step": 1025
},
{
"epoch": 0.206,
"grad_norm": 74.69538879394531,
"learning_rate": 0.00019325760355695188,
"loss": 267.8689,
"step": 1030
},
{
"epoch": 0.207,
"grad_norm": 71.6656723022461,
"learning_rate": 0.00019313103209992204,
"loss": 252.3839,
"step": 1035
},
{
"epoch": 0.208,
"grad_norm": 57.08998107910156,
"learning_rate": 0.0001930033258706564,
"loss": 257.3785,
"step": 1040
},
{
"epoch": 0.209,
"grad_norm": 75.07803344726562,
"learning_rate": 0.00019287448642521513,
"loss": 267.4957,
"step": 1045
},
{
"epoch": 0.21,
"grad_norm": 79.76412963867188,
"learning_rate": 0.00019274451533346615,
"loss": 261.6118,
"step": 1050
},
{
"epoch": 0.211,
"grad_norm": 81.56097412109375,
"learning_rate": 0.00019261341417906621,
"loss": 265.8841,
"step": 1055
},
{
"epoch": 0.212,
"grad_norm": 70.84222412109375,
"learning_rate": 0.0001924811845594415,
"loss": 250.3418,
"step": 1060
},
{
"epoch": 0.213,
"grad_norm": 67.82208251953125,
"learning_rate": 0.00019234782808576824,
"loss": 263.8429,
"step": 1065
},
{
"epoch": 0.214,
"grad_norm": 72.11970520019531,
"learning_rate": 0.00019221334638295294,
"loss": 250.9662,
"step": 1070
},
{
"epoch": 0.215,
"grad_norm": 74.35679626464844,
"learning_rate": 0.00019207774108961272,
"loss": 259.2442,
"step": 1075
},
{
"epoch": 0.216,
"grad_norm": 69.58717346191406,
"learning_rate": 0.0001919410138580553,
"loss": 258.0087,
"step": 1080
},
{
"epoch": 0.217,
"grad_norm": 71.62444305419922,
"learning_rate": 0.0001918031663542588,
"loss": 253.524,
"step": 1085
},
{
"epoch": 0.218,
"grad_norm": 64.11453247070312,
"learning_rate": 0.00019166420025785164,
"loss": 263.3739,
"step": 1090
},
{
"epoch": 0.219,
"grad_norm": 67.0963134765625,
"learning_rate": 0.00019152411726209176,
"loss": 264.8893,
"step": 1095
},
{
"epoch": 0.22,
"grad_norm": 70.35098266601562,
"learning_rate": 0.0001913829190738463,
"loss": 237.8675,
"step": 1100
},
{
"epoch": 0.221,
"grad_norm": 70.07389831542969,
"learning_rate": 0.00019124060741357063,
"loss": 265.9235,
"step": 1105
},
{
"epoch": 0.222,
"grad_norm": 70.8550796508789,
"learning_rate": 0.0001910971840152874,
"loss": 260.5005,
"step": 1110
},
{
"epoch": 0.223,
"grad_norm": 93.7171630859375,
"learning_rate": 0.00019095265062656544,
"loss": 254.5492,
"step": 1115
},
{
"epoch": 0.224,
"grad_norm": 151.48504638671875,
"learning_rate": 0.00019080700900849851,
"loss": 253.4584,
"step": 1120
},
{
"epoch": 0.225,
"grad_norm": 72.2385482788086,
"learning_rate": 0.00019066026093568378,
"loss": 259.241,
"step": 1125
},
{
"epoch": 0.226,
"grad_norm": 60.79322814941406,
"learning_rate": 0.00019051240819620014,
"loss": 259.5886,
"step": 1130
},
{
"epoch": 0.227,
"grad_norm": 65.25270080566406,
"learning_rate": 0.00019036345259158667,
"loss": 260.1056,
"step": 1135
},
{
"epoch": 0.228,
"grad_norm": 60.94587707519531,
"learning_rate": 0.00019021339593682028,
"loss": 257.8644,
"step": 1140
},
{
"epoch": 0.229,
"grad_norm": 84.38846588134766,
"learning_rate": 0.00019006224006029406,
"loss": 244.2875,
"step": 1145
},
{
"epoch": 0.23,
"grad_norm": 69.91790771484375,
"learning_rate": 0.00018990998680379456,
"loss": 258.3608,
"step": 1150
},
{
"epoch": 0.231,
"grad_norm": 81.54023742675781,
"learning_rate": 0.00018975663802247976,
"loss": 272.5109,
"step": 1155
},
{
"epoch": 0.232,
"grad_norm": 69.9112319946289,
"learning_rate": 0.0001896021955848561,
"loss": 255.1114,
"step": 1160
},
{
"epoch": 0.233,
"grad_norm": 78.4234619140625,
"learning_rate": 0.000189446661372756,
"loss": 259.7902,
"step": 1165
},
{
"epoch": 0.234,
"grad_norm": 59.534034729003906,
"learning_rate": 0.0001892900372813147,
"loss": 265.1972,
"step": 1170
},
{
"epoch": 0.235,
"grad_norm": 67.2402114868164,
"learning_rate": 0.00018913232521894732,
"loss": 257.1966,
"step": 1175
},
{
"epoch": 0.236,
"grad_norm": 64.15892028808594,
"learning_rate": 0.00018897352710732564,
"loss": 256.5153,
"step": 1180
},
{
"epoch": 0.237,
"grad_norm": 106.17436981201172,
"learning_rate": 0.00018881364488135448,
"loss": 267.3872,
"step": 1185
},
{
"epoch": 0.238,
"grad_norm": 68.61872100830078,
"learning_rate": 0.00018865268048914828,
"loss": 253.5069,
"step": 1190
},
{
"epoch": 0.239,
"grad_norm": 68.29266357421875,
"learning_rate": 0.00018849063589200743,
"loss": 260.4851,
"step": 1195
},
{
"epoch": 0.24,
"grad_norm": 69.44786834716797,
"learning_rate": 0.00018832751306439418,
"loss": 252.4921,
"step": 1200
},
{
"epoch": 0.241,
"grad_norm": 70.99056243896484,
"learning_rate": 0.0001881633139939087,
"loss": 262.0189,
"step": 1205
},
{
"epoch": 0.242,
"grad_norm": 60.21652603149414,
"learning_rate": 0.00018799804068126485,
"loss": 253.4937,
"step": 1210
},
{
"epoch": 0.243,
"grad_norm": 66.3786392211914,
"learning_rate": 0.00018783169514026578,
"loss": 255.9939,
"step": 1215
},
{
"epoch": 0.244,
"grad_norm": 68.00920104980469,
"learning_rate": 0.00018766427939777945,
"loss": 257.3594,
"step": 1220
},
{
"epoch": 0.245,
"grad_norm": 65.09504699707031,
"learning_rate": 0.0001874957954937138,
"loss": 260.6237,
"step": 1225
},
{
"epoch": 0.246,
"grad_norm": 60.35883331298828,
"learning_rate": 0.00018732624548099204,
"loss": 257.7881,
"step": 1230
},
{
"epoch": 0.247,
"grad_norm": 117.36902618408203,
"learning_rate": 0.00018715563142552758,
"loss": 256.1025,
"step": 1235
},
{
"epoch": 0.248,
"grad_norm": 61.20811080932617,
"learning_rate": 0.0001869839554061988,
"loss": 255.0857,
"step": 1240
},
{
"epoch": 0.249,
"grad_norm": 78.92390441894531,
"learning_rate": 0.00018681121951482393,
"loss": 259.8825,
"step": 1245
},
{
"epoch": 0.25,
"grad_norm": 62.366363525390625,
"learning_rate": 0.00018663742585613518,
"loss": 258.528,
"step": 1250
},
{
"epoch": 0.251,
"grad_norm": 61.07777786254883,
"learning_rate": 0.0001864625765477535,
"loss": 259.0611,
"step": 1255
},
{
"epoch": 0.252,
"grad_norm": 62.74348449707031,
"learning_rate": 0.0001862866737201625,
"loss": 256.3734,
"step": 1260
},
{
"epoch": 0.253,
"grad_norm": 63.42617416381836,
"learning_rate": 0.00018610971951668265,
"loss": 269.0647,
"step": 1265
},
{
"epoch": 0.254,
"grad_norm": 67.448974609375,
"learning_rate": 0.00018593171609344503,
"loss": 250.4642,
"step": 1270
},
{
"epoch": 0.255,
"grad_norm": 61.498104095458984,
"learning_rate": 0.00018575266561936523,
"loss": 259.9946,
"step": 1275
},
{
"epoch": 0.256,
"grad_norm": 61.68122100830078,
"learning_rate": 0.00018557257027611675,
"loss": 257.0583,
"step": 1280
},
{
"epoch": 0.257,
"grad_norm": 62.55679702758789,
"learning_rate": 0.0001853914322581045,
"loss": 256.0573,
"step": 1285
},
{
"epoch": 0.258,
"grad_norm": 112.78436279296875,
"learning_rate": 0.0001852092537724381,
"loss": 244.0424,
"step": 1290
},
{
"epoch": 0.259,
"grad_norm": 79.3133544921875,
"learning_rate": 0.00018502603703890488,
"loss": 250.8328,
"step": 1295
},
{
"epoch": 0.26,
"grad_norm": 70.6105728149414,
"learning_rate": 0.0001848417842899429,
"loss": 258.7902,
"step": 1300
},
{
"epoch": 0.261,
"grad_norm": 68.15289306640625,
"learning_rate": 0.0001846564977706138,
"loss": 250.6485,
"step": 1305
},
{
"epoch": 0.262,
"grad_norm": 65.69000244140625,
"learning_rate": 0.00018447017973857532,
"loss": 248.6376,
"step": 1310
},
{
"epoch": 0.263,
"grad_norm": 58.19040298461914,
"learning_rate": 0.0001842828324640539,
"loss": 263.3102,
"step": 1315
},
{
"epoch": 0.264,
"grad_norm": 64.97654724121094,
"learning_rate": 0.00018409445822981693,
"loss": 256.339,
"step": 1320
},
{
"epoch": 0.265,
"grad_norm": 450.18511962890625,
"learning_rate": 0.000183905059331145,
"loss": 254.6749,
"step": 1325
},
{
"epoch": 0.266,
"grad_norm": 82.3821029663086,
"learning_rate": 0.000183714638075804,
"loss": 254.4249,
"step": 1330
},
{
"epoch": 0.267,
"grad_norm": 75.35969543457031,
"learning_rate": 0.00018352319678401676,
"loss": 254.6711,
"step": 1335
},
{
"epoch": 0.268,
"grad_norm": 64.18514251708984,
"learning_rate": 0.000183330737788435,
"loss": 259.8325,
"step": 1340
},
{
"epoch": 0.269,
"grad_norm": 66.32669830322266,
"learning_rate": 0.00018313726343411086,
"loss": 253.2079,
"step": 1345
},
{
"epoch": 0.27,
"grad_norm": 74.89440155029297,
"learning_rate": 0.00018294277607846832,
"loss": 246.6619,
"step": 1350
},
{
"epoch": 0.271,
"grad_norm": 62.38987731933594,
"learning_rate": 0.00018274727809127438,
"loss": 247.1976,
"step": 1355
},
{
"epoch": 0.272,
"grad_norm": 69.08052062988281,
"learning_rate": 0.00018255077185461038,
"loss": 244.9568,
"step": 1360
},
{
"epoch": 0.273,
"grad_norm": 62.61387634277344,
"learning_rate": 0.00018235325976284275,
"loss": 243.8006,
"step": 1365
},
{
"epoch": 0.274,
"grad_norm": 60.06017303466797,
"learning_rate": 0.00018215474422259402,
"loss": 241.652,
"step": 1370
},
{
"epoch": 0.275,
"grad_norm": 74.84953308105469,
"learning_rate": 0.0001819552276527134,
"loss": 243.0105,
"step": 1375
},
{
"epoch": 0.276,
"grad_norm": 61.174224853515625,
"learning_rate": 0.0001817547124842473,
"loss": 252.2928,
"step": 1380
},
{
"epoch": 0.277,
"grad_norm": 73.44465637207031,
"learning_rate": 0.00018155320116040982,
"loss": 259.0343,
"step": 1385
},
{
"epoch": 0.278,
"grad_norm": 72.30426788330078,
"learning_rate": 0.0001813506961365528,
"loss": 249.5622,
"step": 1390
},
{
"epoch": 0.279,
"grad_norm": 61.13508987426758,
"learning_rate": 0.00018114719988013612,
"loss": 254.7038,
"step": 1395
},
{
"epoch": 0.28,
"grad_norm": 60.89707946777344,
"learning_rate": 0.00018094271487069735,
"loss": 261.7343,
"step": 1400
},
{
"epoch": 0.281,
"grad_norm": 60.29357147216797,
"learning_rate": 0.00018073724359982186,
"loss": 263.7504,
"step": 1405
},
{
"epoch": 0.282,
"grad_norm": 64.27754974365234,
"learning_rate": 0.0001805307885711122,
"loss": 259.5201,
"step": 1410
},
{
"epoch": 0.283,
"grad_norm": 66.30209350585938,
"learning_rate": 0.0001803233523001578,
"loss": 239.3289,
"step": 1415
},
{
"epoch": 0.284,
"grad_norm": 59.92820358276367,
"learning_rate": 0.00018011493731450413,
"loss": 252.4354,
"step": 1420
},
{
"epoch": 0.285,
"grad_norm": 61.41872024536133,
"learning_rate": 0.00017990554615362198,
"loss": 270.7561,
"step": 1425
},
{
"epoch": 0.286,
"grad_norm": 58.28062438964844,
"learning_rate": 0.00017969518136887663,
"loss": 254.5627,
"step": 1430
},
{
"epoch": 0.287,
"grad_norm": 69.97006225585938,
"learning_rate": 0.00017948384552349657,
"loss": 250.7488,
"step": 1435
},
{
"epoch": 0.288,
"grad_norm": 64.22581481933594,
"learning_rate": 0.00017927154119254236,
"loss": 236.5071,
"step": 1440
},
{
"epoch": 0.289,
"grad_norm": 62.77339172363281,
"learning_rate": 0.0001790582709628753,
"loss": 243.0065,
"step": 1445
},
{
"epoch": 0.29,
"grad_norm": 72.30258178710938,
"learning_rate": 0.00017884403743312582,
"loss": 252.5576,
"step": 1450
},
{
"epoch": 0.291,
"grad_norm": 68.7700424194336,
"learning_rate": 0.00017862884321366188,
"loss": 246.0682,
"step": 1455
},
{
"epoch": 0.292,
"grad_norm": 101.85023498535156,
"learning_rate": 0.00017841269092655715,
"loss": 241.2078,
"step": 1460
},
{
"epoch": 0.293,
"grad_norm": 70.41748046875,
"learning_rate": 0.000178195583205559,
"loss": 232.2595,
"step": 1465
},
{
"epoch": 0.294,
"grad_norm": 56.64640426635742,
"learning_rate": 0.00017797752269605653,
"loss": 234.7589,
"step": 1470
},
{
"epoch": 0.295,
"grad_norm": 73.50944519042969,
"learning_rate": 0.0001777585120550482,
"loss": 250.9359,
"step": 1475
},
{
"epoch": 0.296,
"grad_norm": 82.7099609375,
"learning_rate": 0.0001775385539511096,
"loss": 227.0061,
"step": 1480
},
{
"epoch": 0.297,
"grad_norm": 96.73185729980469,
"learning_rate": 0.00017731765106436073,
"loss": 233.9576,
"step": 1485
},
{
"epoch": 0.298,
"grad_norm": 65.25926971435547,
"learning_rate": 0.00017709580608643363,
"loss": 235.6743,
"step": 1490
},
{
"epoch": 0.299,
"grad_norm": 61.761226654052734,
"learning_rate": 0.00017687302172043933,
"loss": 237.7806,
"step": 1495
},
{
"epoch": 0.3,
"grad_norm": 59.79208755493164,
"learning_rate": 0.00017664930068093498,
"loss": 244.9097,
"step": 1500
},
{
"epoch": 0.301,
"grad_norm": 75.16264343261719,
"learning_rate": 0.0001764246456938909,
"loss": 258.5652,
"step": 1505
},
{
"epoch": 0.302,
"grad_norm": 68.61026763916016,
"learning_rate": 0.0001761990594966572,
"loss": 264.604,
"step": 1510
},
{
"epoch": 0.303,
"grad_norm": 64.94346618652344,
"learning_rate": 0.00017597254483793048,
"loss": 251.489,
"step": 1515
},
{
"epoch": 0.304,
"grad_norm": 66.63980102539062,
"learning_rate": 0.00017574510447772039,
"loss": 255.6003,
"step": 1520
},
{
"epoch": 0.305,
"grad_norm": 72.99297332763672,
"learning_rate": 0.00017551674118731591,
"loss": 252.4185,
"step": 1525
},
{
"epoch": 0.306,
"grad_norm": 59.50336837768555,
"learning_rate": 0.00017528745774925172,
"loss": 247.8548,
"step": 1530
},
{
"epoch": 0.307,
"grad_norm": 66.31355285644531,
"learning_rate": 0.00017505725695727412,
"loss": 250.6218,
"step": 1535
},
{
"epoch": 0.308,
"grad_norm": 58.865623474121094,
"learning_rate": 0.00017482614161630714,
"loss": 260.7944,
"step": 1540
},
{
"epoch": 0.309,
"grad_norm": 63.617034912109375,
"learning_rate": 0.00017459411454241822,
"loss": 267.0334,
"step": 1545
},
{
"epoch": 0.31,
"grad_norm": 62.7519645690918,
"learning_rate": 0.0001743611785627841,
"loss": 245.0342,
"step": 1550
},
{
"epoch": 0.311,
"grad_norm": 85.38328552246094,
"learning_rate": 0.0001741273365156561,
"loss": 254.2524,
"step": 1555
},
{
"epoch": 0.312,
"grad_norm": 67.7838363647461,
"learning_rate": 0.0001738925912503259,
"loss": 256.7063,
"step": 1560
},
{
"epoch": 0.313,
"grad_norm": 110.21266174316406,
"learning_rate": 0.00017365694562709034,
"loss": 243.5006,
"step": 1565
},
{
"epoch": 0.314,
"grad_norm": 69.58531188964844,
"learning_rate": 0.00017342040251721702,
"loss": 254.3385,
"step": 1570
},
{
"epoch": 0.315,
"grad_norm": 65.52213287353516,
"learning_rate": 0.0001731829648029091,
"loss": 260.2927,
"step": 1575
},
{
"epoch": 0.316,
"grad_norm": 62.20018768310547,
"learning_rate": 0.00017294463537727024,
"loss": 253.9355,
"step": 1580
},
{
"epoch": 0.317,
"grad_norm": 72.21735382080078,
"learning_rate": 0.0001727054171442692,
"loss": 248.7483,
"step": 1585
},
{
"epoch": 0.318,
"grad_norm": 63.43247604370117,
"learning_rate": 0.0001724653130187047,
"loss": 229.2525,
"step": 1590
},
{
"epoch": 0.319,
"grad_norm": 58.46535873413086,
"learning_rate": 0.0001722243259261697,
"loss": 241.5975,
"step": 1595
},
{
"epoch": 0.32,
"grad_norm": 71.36355590820312,
"learning_rate": 0.0001719824588030159,
"loss": 246.6689,
"step": 1600
},
{
"epoch": 0.321,
"grad_norm": 77.30906677246094,
"learning_rate": 0.00017173971459631787,
"loss": 247.9697,
"step": 1605
},
{
"epoch": 0.322,
"grad_norm": 63.93739318847656,
"learning_rate": 0.00017149609626383717,
"loss": 260.0547,
"step": 1610
},
{
"epoch": 0.323,
"grad_norm": 63.72947311401367,
"learning_rate": 0.00017125160677398626,
"loss": 255.5164,
"step": 1615
},
{
"epoch": 0.324,
"grad_norm": 64.98088073730469,
"learning_rate": 0.0001710062491057925,
"loss": 251.8339,
"step": 1620
},
{
"epoch": 0.325,
"grad_norm": 78.54566192626953,
"learning_rate": 0.00017076002624886154,
"loss": 248.8351,
"step": 1625
},
{
"epoch": 0.326,
"grad_norm": 76.1285629272461,
"learning_rate": 0.00017051294120334125,
"loss": 242.5996,
"step": 1630
},
{
"epoch": 0.327,
"grad_norm": 59.719017028808594,
"learning_rate": 0.00017026499697988493,
"loss": 249.7222,
"step": 1635
},
{
"epoch": 0.328,
"grad_norm": 68.43321990966797,
"learning_rate": 0.00017001619659961467,
"loss": 240.5313,
"step": 1640
},
{
"epoch": 0.329,
"grad_norm": 69.43534088134766,
"learning_rate": 0.00016976654309408464,
"loss": 227.6783,
"step": 1645
},
{
"epoch": 0.33,
"grad_norm": 91.16776275634766,
"learning_rate": 0.000169516039505244,
"loss": 246.3205,
"step": 1650
},
{
"epoch": 0.331,
"grad_norm": 76.1478500366211,
"learning_rate": 0.0001692646888853999,
"loss": 252.1556,
"step": 1655
},
{
"epoch": 0.332,
"grad_norm": 136.00778198242188,
"learning_rate": 0.00016901249429718032,
"loss": 239.3608,
"step": 1660
},
{
"epoch": 0.333,
"grad_norm": 71.8016128540039,
"learning_rate": 0.00016875945881349676,
"loss": 240.7788,
"step": 1665
},
{
"epoch": 0.334,
"grad_norm": 66.53947448730469,
"learning_rate": 0.0001685055855175067,
"loss": 228.5486,
"step": 1670
},
{
"epoch": 0.335,
"grad_norm": 66.0619888305664,
"learning_rate": 0.0001682508775025762,
"loss": 259.8584,
"step": 1675
},
{
"epoch": 0.336,
"grad_norm": 66.97184753417969,
"learning_rate": 0.00016799533787224192,
"loss": 250.809,
"step": 1680
},
{
"epoch": 0.337,
"grad_norm": 64.1236801147461,
"learning_rate": 0.00016773896974017373,
"loss": 247.3749,
"step": 1685
},
{
"epoch": 0.338,
"grad_norm": 65.39366149902344,
"learning_rate": 0.00016748177623013638,
"loss": 237.2529,
"step": 1690
},
{
"epoch": 0.339,
"grad_norm": 65.25891876220703,
"learning_rate": 0.00016722376047595164,
"loss": 242.7048,
"step": 1695
},
{
"epoch": 0.34,
"grad_norm": 62.32688522338867,
"learning_rate": 0.00016696492562145996,
"loss": 246.5424,
"step": 1700
},
{
"epoch": 0.341,
"grad_norm": 62.54330825805664,
"learning_rate": 0.00016670527482048246,
"loss": 244.564,
"step": 1705
},
{
"epoch": 0.342,
"grad_norm": 69.17015838623047,
"learning_rate": 0.00016644481123678217,
"loss": 236.5604,
"step": 1710
},
{
"epoch": 0.343,
"grad_norm": 94.22887420654297,
"learning_rate": 0.00016618353804402568,
"loss": 231.0441,
"step": 1715
},
{
"epoch": 0.344,
"grad_norm": 69.00772857666016,
"learning_rate": 0.00016592145842574433,
"loss": 241.5581,
"step": 1720
},
{
"epoch": 0.345,
"grad_norm": 56.885257720947266,
"learning_rate": 0.00016565857557529566,
"loss": 259.9843,
"step": 1725
},
{
"epoch": 0.346,
"grad_norm": 55.5030632019043,
"learning_rate": 0.00016539489269582416,
"loss": 252.2137,
"step": 1730
},
{
"epoch": 0.347,
"grad_norm": 68.91424560546875,
"learning_rate": 0.00016513041300022255,
"loss": 245.8943,
"step": 1735
},
{
"epoch": 0.348,
"grad_norm": 70.65258026123047,
"learning_rate": 0.00016486513971109243,
"loss": 255.4661,
"step": 1740
},
{
"epoch": 0.349,
"grad_norm": 58.84693908691406,
"learning_rate": 0.0001645990760607051,
"loss": 237.8968,
"step": 1745
},
{
"epoch": 0.35,
"grad_norm": 64.97807312011719,
"learning_rate": 0.0001643322252909622,
"loss": 251.2854,
"step": 1750
},
{
"epoch": 0.351,
"grad_norm": 62.77329635620117,
"learning_rate": 0.00016406459065335615,
"loss": 230.4382,
"step": 1755
},
{
"epoch": 0.352,
"grad_norm": 64.9175033569336,
"learning_rate": 0.00016379617540893056,
"loss": 233.1863,
"step": 1760
},
{
"epoch": 0.353,
"grad_norm": 61.643211364746094,
"learning_rate": 0.00016352698282824044,
"loss": 238.2037,
"step": 1765
},
{
"epoch": 0.354,
"grad_norm": 63.59113311767578,
"learning_rate": 0.00016325701619131246,
"loss": 244.5164,
"step": 1770
},
{
"epoch": 0.355,
"grad_norm": 58.9718017578125,
"learning_rate": 0.00016298627878760487,
"loss": 234.3176,
"step": 1775
},
{
"epoch": 0.356,
"grad_norm": 61.498809814453125,
"learning_rate": 0.00016271477391596753,
"loss": 251.9608,
"step": 1780
},
{
"epoch": 0.357,
"grad_norm": 61.9296989440918,
"learning_rate": 0.00016244250488460158,
"loss": 249.1438,
"step": 1785
},
{
"epoch": 0.358,
"grad_norm": 73.51992797851562,
"learning_rate": 0.0001621694750110193,
"loss": 246.7759,
"step": 1790
},
{
"epoch": 0.359,
"grad_norm": 59.58364486694336,
"learning_rate": 0.00016189568762200348,
"loss": 237.6455,
"step": 1795
},
{
"epoch": 0.36,
"grad_norm": 139.42498779296875,
"learning_rate": 0.00016162114605356703,
"loss": 240.8839,
"step": 1800
},
{
"epoch": 0.361,
"grad_norm": 61.32443618774414,
"learning_rate": 0.00016134585365091243,
"loss": 240.3019,
"step": 1805
},
{
"epoch": 0.362,
"grad_norm": 66.80059051513672,
"learning_rate": 0.00016106981376839066,
"loss": 235.7607,
"step": 1810
},
{
"epoch": 0.363,
"grad_norm": 96.10183715820312,
"learning_rate": 0.00016079302976946055,
"loss": 241.8637,
"step": 1815
},
{
"epoch": 0.364,
"grad_norm": 174.83294677734375,
"learning_rate": 0.0001605155050266478,
"loss": 235.3204,
"step": 1820
},
{
"epoch": 0.365,
"grad_norm": 83.66526794433594,
"learning_rate": 0.00016023724292150385,
"loss": 239.0025,
"step": 1825
},
{
"epoch": 0.366,
"grad_norm": 62.8675651550293,
"learning_rate": 0.00015995824684456465,
"loss": 241.9004,
"step": 1830
},
{
"epoch": 0.367,
"grad_norm": 57.88111877441406,
"learning_rate": 0.00015967852019530926,
"loss": 250.5243,
"step": 1835
},
{
"epoch": 0.368,
"grad_norm": 60.821659088134766,
"learning_rate": 0.0001593980663821187,
"loss": 238.8469,
"step": 1840
},
{
"epoch": 0.369,
"grad_norm": 60.44630813598633,
"learning_rate": 0.0001591168888222342,
"loss": 244.9853,
"step": 1845
},
{
"epoch": 0.37,
"grad_norm": 68.4105224609375,
"learning_rate": 0.00015883499094171554,
"loss": 260.3014,
"step": 1850
},
{
"epoch": 0.371,
"grad_norm": 64.6832275390625,
"learning_rate": 0.0001585523761753994,
"loss": 239.2754,
"step": 1855
},
{
"epoch": 0.372,
"grad_norm": 58.168155670166016,
"learning_rate": 0.00015826904796685762,
"loss": 248.5865,
"step": 1860
},
{
"epoch": 0.373,
"grad_norm": 83.49576568603516,
"learning_rate": 0.00015798500976835493,
"loss": 255.9171,
"step": 1865
},
{
"epoch": 0.374,
"grad_norm": 101.22307586669922,
"learning_rate": 0.00015770026504080718,
"loss": 248.0609,
"step": 1870
},
{
"epoch": 0.375,
"grad_norm": 68.59806060791016,
"learning_rate": 0.000157414817253739,
"loss": 236.3886,
"step": 1875
},
{
"epoch": 0.376,
"grad_norm": 71.27215576171875,
"learning_rate": 0.00015712866988524155,
"loss": 247.2733,
"step": 1880
},
{
"epoch": 0.377,
"grad_norm": 65.266357421875,
"learning_rate": 0.0001568418264219303,
"loss": 236.9641,
"step": 1885
},
{
"epoch": 0.378,
"grad_norm": 59.90776824951172,
"learning_rate": 0.0001565542903589023,
"loss": 215.6263,
"step": 1890
},
{
"epoch": 0.379,
"grad_norm": 61.711795806884766,
"learning_rate": 0.0001562660651996937,
"loss": 249.6925,
"step": 1895
},
{
"epoch": 0.38,
"grad_norm": 57.719783782958984,
"learning_rate": 0.00015597715445623712,
"loss": 241.6331,
"step": 1900
},
{
"epoch": 0.381,
"grad_norm": 60.528297424316406,
"learning_rate": 0.00015568756164881882,
"loss": 237.7176,
"step": 1905
},
{
"epoch": 0.382,
"grad_norm": 69.33570098876953,
"learning_rate": 0.00015539729030603574,
"loss": 235.1876,
"step": 1910
},
{
"epoch": 0.383,
"grad_norm": 64.02123260498047,
"learning_rate": 0.0001551063439647526,
"loss": 234.8362,
"step": 1915
},
{
"epoch": 0.384,
"grad_norm": 57.75348663330078,
"learning_rate": 0.00015481472617005876,
"loss": 236.5208,
"step": 1920
},
{
"epoch": 0.385,
"grad_norm": 53.08407974243164,
"learning_rate": 0.00015452244047522502,
"loss": 233.4317,
"step": 1925
},
{
"epoch": 0.386,
"grad_norm": 73.68172454833984,
"learning_rate": 0.0001542294904416603,
"loss": 238.0298,
"step": 1930
},
{
"epoch": 0.387,
"grad_norm": 57.79017639160156,
"learning_rate": 0.00015393587963886835,
"loss": 230.5418,
"step": 1935
},
{
"epoch": 0.388,
"grad_norm": 65.8370590209961,
"learning_rate": 0.0001536416116444041,
"loss": 247.8104,
"step": 1940
},
{
"epoch": 0.389,
"grad_norm": 67.43762969970703,
"learning_rate": 0.0001533466900438303,
"loss": 258.8561,
"step": 1945
},
{
"epoch": 0.39,
"grad_norm": 57.806129455566406,
"learning_rate": 0.0001530511184306734,
"loss": 254.9089,
"step": 1950
},
{
"epoch": 0.391,
"grad_norm": 69.66842651367188,
"learning_rate": 0.00015275490040638038,
"loss": 248.4691,
"step": 1955
},
{
"epoch": 0.392,
"grad_norm": 84.50879669189453,
"learning_rate": 0.00015245803958027434,
"loss": 237.1378,
"step": 1960
},
{
"epoch": 0.393,
"grad_norm": 65.10584259033203,
"learning_rate": 0.0001521605395695108,
"loss": 242.4159,
"step": 1965
},
{
"epoch": 0.394,
"grad_norm": 66.4300537109375,
"learning_rate": 0.00015186240399903342,
"loss": 245.825,
"step": 1970
},
{
"epoch": 0.395,
"grad_norm": 109.91886901855469,
"learning_rate": 0.00015156363650153012,
"loss": 250.4606,
"step": 1975
},
{
"epoch": 0.396,
"grad_norm": 72.5931396484375,
"learning_rate": 0.0001512642407173885,
"loss": 248.6551,
"step": 1980
},
{
"epoch": 0.397,
"grad_norm": 73.90507507324219,
"learning_rate": 0.00015096422029465178,
"loss": 244.256,
"step": 1985
},
{
"epoch": 0.398,
"grad_norm": 65.20423126220703,
"learning_rate": 0.0001506635788889741,
"loss": 220.3735,
"step": 1990
},
{
"epoch": 0.399,
"grad_norm": 71.23815155029297,
"learning_rate": 0.0001503623201635761,
"loss": 247.7928,
"step": 1995
},
{
"epoch": 0.4,
"grad_norm": 76.0530014038086,
"learning_rate": 0.0001500604477892003,
"loss": 245.9678,
"step": 2000
},
{
"epoch": 0.401,
"grad_norm": 67.73207092285156,
"learning_rate": 0.00014975796544406625,
"loss": 229.0977,
"step": 2005
},
{
"epoch": 0.402,
"grad_norm": 68.48651123046875,
"learning_rate": 0.00014945487681382598,
"loss": 236.5397,
"step": 2010
},
{
"epoch": 0.403,
"grad_norm": 61.38922119140625,
"learning_rate": 0.0001491511855915187,
"loss": 238.7702,
"step": 2015
},
{
"epoch": 0.404,
"grad_norm": 68.66989135742188,
"learning_rate": 0.0001488468954775262,
"loss": 242.4966,
"step": 2020
},
{
"epoch": 0.405,
"grad_norm": 70.91650390625,
"learning_rate": 0.0001485420101795274,
"loss": 241.4282,
"step": 2025
},
{
"epoch": 0.406,
"grad_norm": 61.95619583129883,
"learning_rate": 0.00014823653341245353,
"loss": 255.1551,
"step": 2030
},
{
"epoch": 0.407,
"grad_norm": 82.40103149414062,
"learning_rate": 0.0001479304688984425,
"loss": 255.0193,
"step": 2035
},
{
"epoch": 0.408,
"grad_norm": 65.26828002929688,
"learning_rate": 0.0001476238203667939,
"loss": 256.2934,
"step": 2040
},
{
"epoch": 0.409,
"grad_norm": 58.06156539916992,
"learning_rate": 0.00014731659155392332,
"loss": 244.8165,
"step": 2045
},
{
"epoch": 0.41,
"grad_norm": 69.94371032714844,
"learning_rate": 0.00014700878620331684,
"loss": 236.5389,
"step": 2050
},
{
"epoch": 0.411,
"grad_norm": 57.47764205932617,
"learning_rate": 0.00014670040806548555,
"loss": 232.9615,
"step": 2055
},
{
"epoch": 0.412,
"grad_norm": 61.665435791015625,
"learning_rate": 0.0001463914608979197,
"loss": 244.6553,
"step": 2060
},
{
"epoch": 0.413,
"grad_norm": 63.331031799316406,
"learning_rate": 0.0001460819484650431,
"loss": 240.4683,
"step": 2065
},
{
"epoch": 0.414,
"grad_norm": 61.5399284362793,
"learning_rate": 0.000145771874538167,
"loss": 233.3042,
"step": 2070
},
{
"epoch": 0.415,
"grad_norm": 192.06161499023438,
"learning_rate": 0.0001454612428954444,
"loss": 251.3163,
"step": 2075
},
{
"epoch": 0.416,
"grad_norm": 66.12826538085938,
"learning_rate": 0.00014515005732182383,
"loss": 233.3447,
"step": 2080
},
{
"epoch": 0.417,
"grad_norm": 70.17752075195312,
"learning_rate": 0.00014483832160900326,
"loss": 239.6131,
"step": 2085
},
{
"epoch": 0.418,
"grad_norm": 64.34825134277344,
"learning_rate": 0.00014452603955538397,
"loss": 250.7617,
"step": 2090
},
{
"epoch": 0.419,
"grad_norm": 68.52357482910156,
"learning_rate": 0.00014421321496602428,
"loss": 244.4657,
"step": 2095
},
{
"epoch": 0.42,
"grad_norm": 82.43126678466797,
"learning_rate": 0.00014389985165259308,
"loss": 245.4597,
"step": 2100
},
{
"epoch": 0.421,
"grad_norm": 65.874267578125,
"learning_rate": 0.00014358595343332342,
"loss": 239.2493,
"step": 2105
},
{
"epoch": 0.422,
"grad_norm": 65.05198669433594,
"learning_rate": 0.00014327152413296608,
"loss": 239.5945,
"step": 2110
},
{
"epoch": 0.423,
"grad_norm": 60.770965576171875,
"learning_rate": 0.00014295656758274284,
"loss": 249.9315,
"step": 2115
},
{
"epoch": 0.424,
"grad_norm": 61.743896484375,
"learning_rate": 0.0001426410876202999,
"loss": 230.7119,
"step": 2120
},
{
"epoch": 0.425,
"grad_norm": 55.88385772705078,
"learning_rate": 0.00014232508808966098,
"loss": 239.291,
"step": 2125
},
{
"epoch": 0.426,
"grad_norm": 61.7985954284668,
"learning_rate": 0.00014200857284118066,
"loss": 233.6485,
"step": 2130
},
{
"epoch": 0.427,
"grad_norm": 80.9444351196289,
"learning_rate": 0.00014169154573149737,
"loss": 224.9626,
"step": 2135
},
{
"epoch": 0.428,
"grad_norm": 56.159523010253906,
"learning_rate": 0.00014137401062348638,
"loss": 254.8871,
"step": 2140
},
{
"epoch": 0.429,
"grad_norm": 71.07969665527344,
"learning_rate": 0.0001410559713862128,
"loss": 235.4285,
"step": 2145
},
{
"epoch": 0.43,
"grad_norm": 60.81352996826172,
"learning_rate": 0.00014073743189488435,
"loss": 246.4281,
"step": 2150
},
{
"epoch": 0.431,
"grad_norm": 59.1260986328125,
"learning_rate": 0.00014041839603080422,
"loss": 245.2993,
"step": 2155
},
{
"epoch": 0.432,
"grad_norm": 63.77505111694336,
"learning_rate": 0.00014009886768132375,
"loss": 222.7855,
"step": 2160
},
{
"epoch": 0.433,
"grad_norm": 96.66449737548828,
"learning_rate": 0.000139778850739795,
"loss": 245.9477,
"step": 2165
},
{
"epoch": 0.434,
"grad_norm": 56.731544494628906,
"learning_rate": 0.0001394583491055234,
"loss": 243.3353,
"step": 2170
},
{
"epoch": 0.435,
"grad_norm": 57.319236755371094,
"learning_rate": 0.00013913736668372026,
"loss": 249.849,
"step": 2175
},
{
"epoch": 0.436,
"grad_norm": 64.18098449707031,
"learning_rate": 0.00013881590738545508,
"loss": 253.3436,
"step": 2180
},
{
"epoch": 0.437,
"grad_norm": 61.97779083251953,
"learning_rate": 0.00013849397512760795,
"loss": 244.4615,
"step": 2185
},
{
"epoch": 0.438,
"grad_norm": 69.94448852539062,
"learning_rate": 0.00013817157383282184,
"loss": 225.9712,
"step": 2190
},
{
"epoch": 0.439,
"grad_norm": 61.12146759033203,
"learning_rate": 0.00013784870742945482,
"loss": 239.8827,
"step": 2195
},
{
"epoch": 0.44,
"grad_norm": 63.54712677001953,
"learning_rate": 0.0001375253798515321,
"loss": 234.7807,
"step": 2200
},
{
"epoch": 0.441,
"grad_norm": 64.55406188964844,
"learning_rate": 0.00013720159503869815,
"loss": 234.2377,
"step": 2205
},
{
"epoch": 0.442,
"grad_norm": 67.22713470458984,
"learning_rate": 0.00013687735693616876,
"loss": 241.5794,
"step": 2210
},
{
"epoch": 0.443,
"grad_norm": 61.47258377075195,
"learning_rate": 0.0001365526694946829,
"loss": 239.5455,
"step": 2215
},
{
"epoch": 0.444,
"grad_norm": 63.785343170166016,
"learning_rate": 0.00013622753667045457,
"loss": 241.8949,
"step": 2220
},
{
"epoch": 0.445,
"grad_norm": 59.00518035888672,
"learning_rate": 0.00013590196242512463,
"loss": 234.9362,
"step": 2225
},
{
"epoch": 0.446,
"grad_norm": 81.03519439697266,
"learning_rate": 0.0001355759507257125,
"loss": 237.0903,
"step": 2230
},
{
"epoch": 0.447,
"grad_norm": 60.88785171508789,
"learning_rate": 0.00013524950554456784,
"loss": 235.0176,
"step": 2235
},
{
"epoch": 0.448,
"grad_norm": 62.00077438354492,
"learning_rate": 0.00013492263085932224,
"loss": 239.6677,
"step": 2240
},
{
"epoch": 0.449,
"grad_norm": 54.97482681274414,
"learning_rate": 0.0001345953306528405,
"loss": 240.5235,
"step": 2245
},
{
"epoch": 0.45,
"grad_norm": 62.751216888427734,
"learning_rate": 0.00013426760891317236,
"loss": 237.0468,
"step": 2250
},
{
"epoch": 0.451,
"grad_norm": 69.67955017089844,
"learning_rate": 0.00013393946963350382,
"loss": 239.1589,
"step": 2255
},
{
"epoch": 0.452,
"grad_norm": 64.26301574707031,
"learning_rate": 0.00013361091681210845,
"loss": 247.5965,
"step": 2260
},
{
"epoch": 0.453,
"grad_norm": 56.22570037841797,
"learning_rate": 0.00013328195445229868,
"loss": 233.4828,
"step": 2265
},
{
"epoch": 0.454,
"grad_norm": 61.460750579833984,
"learning_rate": 0.00013295258656237702,
"loss": 232.6884,
"step": 2270
},
{
"epoch": 0.455,
"grad_norm": 57.480690002441406,
"learning_rate": 0.00013262281715558736,
"loss": 250.5228,
"step": 2275
},
{
"epoch": 0.456,
"grad_norm": 54.52292251586914,
"learning_rate": 0.00013229265025006576,
"loss": 244.7536,
"step": 2280
},
{
"epoch": 0.457,
"grad_norm": 62.742244720458984,
"learning_rate": 0.0001319620898687918,
"loss": 253.2492,
"step": 2285
},
{
"epoch": 0.458,
"grad_norm": 58.441490173339844,
"learning_rate": 0.0001316311400395394,
"loss": 238.2327,
"step": 2290
},
{
"epoch": 0.459,
"grad_norm": 120.33824157714844,
"learning_rate": 0.00013129980479482782,
"loss": 246.7062,
"step": 2295
},
{
"epoch": 0.46,
"grad_norm": 59.58649826049805,
"learning_rate": 0.00013096808817187242,
"loss": 241.4882,
"step": 2300
},
{
"epoch": 0.461,
"grad_norm": 65.39479064941406,
"learning_rate": 0.00013063599421253558,
"loss": 235.64,
"step": 2305
},
{
"epoch": 0.462,
"grad_norm": 70.0931625366211,
"learning_rate": 0.00013030352696327742,
"loss": 238.6676,
"step": 2310
},
{
"epoch": 0.463,
"grad_norm": 58.46336364746094,
"learning_rate": 0.0001299706904751064,
"loss": 237.8939,
"step": 2315
},
{
"epoch": 0.464,
"grad_norm": 63.86176300048828,
"learning_rate": 0.00012963748880353011,
"loss": 244.4889,
"step": 2320
},
{
"epoch": 0.465,
"grad_norm": 61.63530349731445,
"learning_rate": 0.00012930392600850573,
"loss": 242.4194,
"step": 2325
},
{
"epoch": 0.466,
"grad_norm": 59.6197509765625,
"learning_rate": 0.00012897000615439075,
"loss": 243.6793,
"step": 2330
},
{
"epoch": 0.467,
"grad_norm": 61.077152252197266,
"learning_rate": 0.00012863573330989313,
"loss": 235.2319,
"step": 2335
},
{
"epoch": 0.468,
"grad_norm": 106.57664489746094,
"learning_rate": 0.00012830111154802203,
"loss": 252.657,
"step": 2340
},
{
"epoch": 0.469,
"grad_norm": 63.63618087768555,
"learning_rate": 0.000127966144946038,
"loss": 246.2469,
"step": 2345
},
{
"epoch": 0.47,
"grad_norm": 62.414955139160156,
"learning_rate": 0.00012763083758540337,
"loss": 247.511,
"step": 2350
},
{
"epoch": 0.471,
"grad_norm": 58.36800003051758,
"learning_rate": 0.00012729519355173254,
"loss": 258.8127,
"step": 2355
},
{
"epoch": 0.472,
"grad_norm": 60.2503662109375,
"learning_rate": 0.00012695921693474212,
"loss": 235.5709,
"step": 2360
},
{
"epoch": 0.473,
"grad_norm": 61.302093505859375,
"learning_rate": 0.00012662291182820114,
"loss": 231.9777,
"step": 2365
},
{
"epoch": 0.474,
"grad_norm": 170.6962127685547,
"learning_rate": 0.0001262862823298812,
"loss": 243.1137,
"step": 2370
},
{
"epoch": 0.475,
"grad_norm": 61.58894348144531,
"learning_rate": 0.00012594933254150655,
"loss": 237.433,
"step": 2375
},
{
"epoch": 0.476,
"grad_norm": 68.40026092529297,
"learning_rate": 0.00012561206656870396,
"loss": 246.8944,
"step": 2380
},
{
"epoch": 0.477,
"grad_norm": 63.99790573120117,
"learning_rate": 0.00012527448852095295,
"loss": 240.999,
"step": 2385
},
{
"epoch": 0.478,
"grad_norm": 80.36054992675781,
"learning_rate": 0.0001249366025115354,
"loss": 234.8044,
"step": 2390
},
{
"epoch": 0.479,
"grad_norm": 72.70059967041016,
"learning_rate": 0.0001245984126574858,
"loss": 233.5652,
"step": 2395
},
{
"epoch": 0.48,
"grad_norm": 66.48823547363281,
"learning_rate": 0.00012425992307954075,
"loss": 232.478,
"step": 2400
},
{
"epoch": 0.481,
"grad_norm": 58.06026840209961,
"learning_rate": 0.00012392113790208895,
"loss": 207.8611,
"step": 2405
},
{
"epoch": 0.482,
"grad_norm": 58.839759826660156,
"learning_rate": 0.00012358206125312083,
"loss": 231.4109,
"step": 2410
},
{
"epoch": 0.483,
"grad_norm": 74.88693237304688,
"learning_rate": 0.0001232426972641784,
"loss": 253.2388,
"step": 2415
},
{
"epoch": 0.484,
"grad_norm": 55.45047378540039,
"learning_rate": 0.00012290305007030478,
"loss": 253.2637,
"step": 2420
},
{
"epoch": 0.485,
"grad_norm": 80.15945434570312,
"learning_rate": 0.00012256312380999376,
"loss": 239.2875,
"step": 2425
},
{
"epoch": 0.486,
"grad_norm": 60.26356506347656,
"learning_rate": 0.00012222292262513965,
"loss": 227.7976,
"step": 2430
},
{
"epoch": 0.487,
"grad_norm": 60.471561431884766,
"learning_rate": 0.00012188245066098647,
"loss": 235.5319,
"step": 2435
},
{
"epoch": 0.488,
"grad_norm": 61.756874084472656,
"learning_rate": 0.00012154171206607764,
"loss": 229.1146,
"step": 2440
},
{
"epoch": 0.489,
"grad_norm": 53.89438247680664,
"learning_rate": 0.00012120071099220549,
"loss": 240.3418,
"step": 2445
},
{
"epoch": 0.49,
"grad_norm": 58.24182891845703,
"learning_rate": 0.00012085945159436038,
"loss": 236.7947,
"step": 2450
},
{
"epoch": 0.491,
"grad_norm": 64.80339813232422,
"learning_rate": 0.00012051793803068046,
"loss": 234.0121,
"step": 2455
},
{
"epoch": 0.492,
"grad_norm": 56.22035217285156,
"learning_rate": 0.0001201761744624007,
"loss": 226.0305,
"step": 2460
},
{
"epoch": 0.493,
"grad_norm": 60.1837272644043,
"learning_rate": 0.00011983416505380234,
"loss": 244.352,
"step": 2465
},
{
"epoch": 0.494,
"grad_norm": 58.84907531738281,
"learning_rate": 0.00011949191397216206,
"loss": 246.7269,
"step": 2470
},
{
"epoch": 0.495,
"grad_norm": 58.51205825805664,
"learning_rate": 0.00011914942538770131,
"loss": 243.1532,
"step": 2475
},
{
"epoch": 0.496,
"grad_norm": 79.28509521484375,
"learning_rate": 0.00011880670347353539,
"loss": 242.6627,
"step": 2480
},
{
"epoch": 0.497,
"grad_norm": 187.86900329589844,
"learning_rate": 0.0001184637524056227,
"loss": 255.2222,
"step": 2485
},
{
"epoch": 0.498,
"grad_norm": 63.020999908447266,
"learning_rate": 0.00011812057636271374,
"loss": 227.5514,
"step": 2490
},
{
"epoch": 0.499,
"grad_norm": 66.07195281982422,
"learning_rate": 0.00011777717952630031,
"loss": 252.7046,
"step": 2495
},
{
"epoch": 0.5,
"grad_norm": 59.861446380615234,
"learning_rate": 0.00011743356608056449,
"loss": 238.4142,
"step": 2500
},
{
"epoch": 0.501,
"grad_norm": 91.85547637939453,
"learning_rate": 0.00011708974021232769,
"loss": 249.0073,
"step": 2505
},
{
"epoch": 0.502,
"grad_norm": 66.29524230957031,
"learning_rate": 0.00011674570611099955,
"loss": 232.3617,
"step": 2510
},
{
"epoch": 0.503,
"grad_norm": 60.96319580078125,
"learning_rate": 0.00011640146796852711,
"loss": 230.5008,
"step": 2515
},
{
"epoch": 0.504,
"grad_norm": 66.24161529541016,
"learning_rate": 0.00011605702997934345,
"loss": 232.6192,
"step": 2520
},
{
"epoch": 0.505,
"grad_norm": 62.35693359375,
"learning_rate": 0.00011571239634031679,
"loss": 239.766,
"step": 2525
},
{
"epoch": 0.506,
"grad_norm": 61.14425277709961,
"learning_rate": 0.00011536757125069923,
"loss": 249.4623,
"step": 2530
},
{
"epoch": 0.507,
"grad_norm": 68.11719512939453,
"learning_rate": 0.00011502255891207572,
"loss": 243.9944,
"step": 2535
},
{
"epoch": 0.508,
"grad_norm": 62.079551696777344,
"learning_rate": 0.00011467736352831266,
"loss": 237.1768,
"step": 2540
},
{
"epoch": 0.509,
"grad_norm": 56.01610565185547,
"learning_rate": 0.00011433198930550695,
"loss": 239.001,
"step": 2545
},
{
"epoch": 0.51,
"grad_norm": 58.486572265625,
"learning_rate": 0.00011398644045193444,
"loss": 243.1911,
"step": 2550
},
{
"epoch": 0.511,
"grad_norm": 57.300201416015625,
"learning_rate": 0.00011364072117799885,
"loss": 235.3636,
"step": 2555
},
{
"epoch": 0.512,
"grad_norm": 61.27510070800781,
"learning_rate": 0.00011329483569618045,
"loss": 237.3513,
"step": 2560
},
{
"epoch": 0.513,
"grad_norm": 59.30598068237305,
"learning_rate": 0.00011294878822098469,
"loss": 246.4485,
"step": 2565
},
{
"epoch": 0.514,
"grad_norm": 72.56581115722656,
"learning_rate": 0.00011260258296889086,
"loss": 227.0799,
"step": 2570
},
{
"epoch": 0.515,
"grad_norm": 138.9378662109375,
"learning_rate": 0.00011225622415830068,
"loss": 181.89,
"step": 2575
},
{
"epoch": 0.516,
"grad_norm": 62.93873596191406,
"learning_rate": 0.00011190971600948699,
"loss": 219.6437,
"step": 2580
},
{
"epoch": 0.517,
"grad_norm": 58.3951530456543,
"learning_rate": 0.00011156306274454218,
"loss": 231.366,
"step": 2585
},
{
"epoch": 0.518,
"grad_norm": 76.44886016845703,
"learning_rate": 0.0001112162685873269,
"loss": 234.1057,
"step": 2590
},
{
"epoch": 0.519,
"grad_norm": 60.203487396240234,
"learning_rate": 0.00011086933776341852,
"loss": 239.2624,
"step": 2595
},
{
"epoch": 0.52,
"grad_norm": 68.33686065673828,
"learning_rate": 0.00011052227450005967,
"loss": 224.2315,
"step": 2600
},
{
"epoch": 0.521,
"grad_norm": 66.6567611694336,
"learning_rate": 0.00011017508302610664,
"loss": 228.0241,
"step": 2605
},
{
"epoch": 0.522,
"grad_norm": 64.11446380615234,
"learning_rate": 0.00010982776757197799,
"loss": 260.6547,
"step": 2610
},
{
"epoch": 0.523,
"grad_norm": 87.50645446777344,
"learning_rate": 0.00010948033236960294,
"loss": 243.3,
"step": 2615
},
{
"epoch": 0.524,
"grad_norm": 60.33220672607422,
"learning_rate": 0.00010913278165236978,
"loss": 244.7448,
"step": 2620
},
{
"epoch": 0.525,
"grad_norm": 64.48478698730469,
"learning_rate": 0.00010878511965507434,
"loss": 242.857,
"step": 2625
},
{
"epoch": 0.526,
"grad_norm": 75.51817321777344,
"learning_rate": 0.00010843735061386828,
"loss": 223.696,
"step": 2630
},
{
"epoch": 0.527,
"grad_norm": 61.36201477050781,
"learning_rate": 0.00010808947876620767,
"loss": 228.7271,
"step": 2635
},
{
"epoch": 0.528,
"grad_norm": 60.587257385253906,
"learning_rate": 0.00010774150835080119,
"loss": 223.0957,
"step": 2640
},
{
"epoch": 0.529,
"grad_norm": 60.65777587890625,
"learning_rate": 0.00010739344360755852,
"loss": 253.583,
"step": 2645
},
{
"epoch": 0.53,
"grad_norm": 54.75935745239258,
"learning_rate": 0.0001070452887775387,
"loss": 236.5906,
"step": 2650
},
{
"epoch": 0.531,
"grad_norm": 62.956722259521484,
"learning_rate": 0.00010669704810289851,
"loss": 237.0601,
"step": 2655
},
{
"epoch": 0.532,
"grad_norm": 92.0921401977539,
"learning_rate": 0.00010634872582684061,
"loss": 241.9507,
"step": 2660
},
{
"epoch": 0.533,
"grad_norm": 65.49005889892578,
"learning_rate": 0.00010600032619356209,
"loss": 243.4771,
"step": 2665
},
{
"epoch": 0.534,
"grad_norm": 59.5872802734375,
"learning_rate": 0.00010565185344820247,
"loss": 241.934,
"step": 2670
},
{
"epoch": 0.535,
"grad_norm": 56.114322662353516,
"learning_rate": 0.00010530331183679218,
"loss": 237.9316,
"step": 2675
},
{
"epoch": 0.536,
"grad_norm": 57.16724395751953,
"learning_rate": 0.00010495470560620083,
"loss": 241.5337,
"step": 2680
},
{
"epoch": 0.537,
"grad_norm": 61.376731872558594,
"learning_rate": 0.00010460603900408523,
"loss": 239.176,
"step": 2685
},
{
"epoch": 0.538,
"grad_norm": 63.183937072753906,
"learning_rate": 0.00010425731627883797,
"loss": 236.7282,
"step": 2690
},
{
"epoch": 0.539,
"grad_norm": 64.90032196044922,
"learning_rate": 0.00010390854167953537,
"loss": 236.8318,
"step": 2695
},
{
"epoch": 0.54,
"grad_norm": 59.82840347290039,
"learning_rate": 0.00010355971945588585,
"loss": 243.5185,
"step": 2700
},
{
"epoch": 0.541,
"grad_norm": 61.49772262573242,
"learning_rate": 0.00010321085385817817,
"loss": 225.9077,
"step": 2705
},
{
"epoch": 0.542,
"grad_norm": 58.72164535522461,
"learning_rate": 0.00010286194913722948,
"loss": 257.1229,
"step": 2710
},
{
"epoch": 0.543,
"grad_norm": 65.40806579589844,
"learning_rate": 0.00010251300954433376,
"loss": 223.6326,
"step": 2715
},
{
"epoch": 0.544,
"grad_norm": 59.297889709472656,
"learning_rate": 0.00010216403933120979,
"loss": 235.6933,
"step": 2720
},
{
"epoch": 0.545,
"grad_norm": 61.0966682434082,
"learning_rate": 0.00010181504274994948,
"loss": 246.2851,
"step": 2725
},
{
"epoch": 0.546,
"grad_norm": 59.07041931152344,
"learning_rate": 0.00010146602405296607,
"loss": 239.2856,
"step": 2730
},
{
"epoch": 0.547,
"grad_norm": 75.01293182373047,
"learning_rate": 0.00010111698749294223,
"loss": 240.9919,
"step": 2735
},
{
"epoch": 0.548,
"grad_norm": 63.14535903930664,
"learning_rate": 0.00010076793732277829,
"loss": 236.2562,
"step": 2740
},
{
"epoch": 0.549,
"grad_norm": 60.572410583496094,
"learning_rate": 0.0001004188777955404,
"loss": 227.527,
"step": 2745
},
{
"epoch": 0.55,
"grad_norm": 58.98918533325195,
"learning_rate": 0.00010006981316440875,
"loss": 234.3241,
"step": 2750
},
{
"epoch": 0.551,
"grad_norm": 58.03236389160156,
"learning_rate": 9.972074768262576e-05,
"loss": 239.9354,
"step": 2755
},
{
"epoch": 0.552,
"grad_norm": 54.80331039428711,
"learning_rate": 9.937168560344412e-05,
"loss": 233.2195,
"step": 2760
},
{
"epoch": 0.553,
"grad_norm": 63.448822021484375,
"learning_rate": 9.902263118007513e-05,
"loss": 240.1099,
"step": 2765
},
{
"epoch": 0.554,
"grad_norm": 58.246917724609375,
"learning_rate": 9.867358866563673e-05,
"loss": 227.2533,
"step": 2770
},
{
"epoch": 0.555,
"grad_norm": 64.1023178100586,
"learning_rate": 9.832456231310189e-05,
"loss": 250.562,
"step": 2775
},
{
"epoch": 0.556,
"grad_norm": 76.42041015625,
"learning_rate": 9.797555637524649e-05,
"loss": 249.508,
"step": 2780
},
{
"epoch": 0.557,
"grad_norm": 65.2816162109375,
"learning_rate": 9.762657510459783e-05,
"loss": 239.5315,
"step": 2785
},
{
"epoch": 0.558,
"grad_norm": 60.70484924316406,
"learning_rate": 9.727762275338246e-05,
"loss": 239.3723,
"step": 2790
},
{
"epoch": 0.559,
"grad_norm": 164.0494384765625,
"learning_rate": 9.692870357347473e-05,
"loss": 234.6554,
"step": 2795
},
{
"epoch": 0.56,
"grad_norm": 56.92321014404297,
"learning_rate": 9.657982181634475e-05,
"loss": 242.3019,
"step": 2800
},
{
"epoch": 0.561,
"grad_norm": 52.640716552734375,
"learning_rate": 9.623098173300654e-05,
"loss": 243.621,
"step": 2805
},
{
"epoch": 0.562,
"grad_norm": 55.68557357788086,
"learning_rate": 9.588218757396655e-05,
"loss": 238.2811,
"step": 2810
},
{
"epoch": 0.563,
"grad_norm": 54.3975944519043,
"learning_rate": 9.55334435891714e-05,
"loss": 249.4321,
"step": 2815
},
{
"epoch": 0.564,
"grad_norm": 77.16236877441406,
"learning_rate": 9.518475402795661e-05,
"loss": 222.2912,
"step": 2820
},
{
"epoch": 0.565,
"grad_norm": 63.21393585205078,
"learning_rate": 9.483612313899435e-05,
"loss": 223.6223,
"step": 2825
},
{
"epoch": 0.566,
"grad_norm": 58.208290100097656,
"learning_rate": 9.448755517024206e-05,
"loss": 228.3,
"step": 2830
},
{
"epoch": 0.567,
"grad_norm": 56.82502365112305,
"learning_rate": 9.413905436889035e-05,
"loss": 245.1693,
"step": 2835
},
{
"epoch": 0.568,
"grad_norm": 77.26526641845703,
"learning_rate": 9.37906249813116e-05,
"loss": 228.4223,
"step": 2840
},
{
"epoch": 0.569,
"grad_norm": 64.27784729003906,
"learning_rate": 9.344227125300788e-05,
"loss": 236.0278,
"step": 2845
},
{
"epoch": 0.57,
"grad_norm": 60.73809814453125,
"learning_rate": 9.309399742855942e-05,
"loss": 245.0785,
"step": 2850
},
{
"epoch": 0.571,
"grad_norm": 64.869140625,
"learning_rate": 9.274580775157294e-05,
"loss": 248.7473,
"step": 2855
},
{
"epoch": 0.572,
"grad_norm": 58.31768035888672,
"learning_rate": 9.239770646462968e-05,
"loss": 238.6827,
"step": 2860
},
{
"epoch": 0.573,
"grad_norm": 216.96249389648438,
"learning_rate": 9.204969780923403e-05,
"loss": 227.2639,
"step": 2865
},
{
"epoch": 0.574,
"grad_norm": 66.54228973388672,
"learning_rate": 9.17017860257616e-05,
"loss": 229.4464,
"step": 2870
},
{
"epoch": 0.575,
"grad_norm": 59.51702880859375,
"learning_rate": 9.135397535340773e-05,
"loss": 241.1056,
"step": 2875
},
{
"epoch": 0.576,
"grad_norm": 68.89649200439453,
"learning_rate": 9.100627003013562e-05,
"loss": 232.2868,
"step": 2880
},
{
"epoch": 0.577,
"grad_norm": 65.72486877441406,
"learning_rate": 9.065867429262496e-05,
"loss": 238.6274,
"step": 2885
},
{
"epoch": 0.578,
"grad_norm": 60.892086029052734,
"learning_rate": 9.03111923762201e-05,
"loss": 225.4908,
"step": 2890
},
{
"epoch": 0.579,
"grad_norm": 55.909690856933594,
"learning_rate": 8.99638285148785e-05,
"loss": 247.2271,
"step": 2895
},
{
"epoch": 0.58,
"grad_norm": 65.2547378540039,
"learning_rate": 8.961658694111929e-05,
"loss": 242.1955,
"step": 2900
},
{
"epoch": 0.581,
"grad_norm": 62.73838806152344,
"learning_rate": 8.926947188597134e-05,
"loss": 241.343,
"step": 2905
},
{
"epoch": 0.582,
"grad_norm": 57.31295394897461,
"learning_rate": 8.892248757892214e-05,
"loss": 232.0151,
"step": 2910
},
{
"epoch": 0.583,
"grad_norm": 60.45274353027344,
"learning_rate": 8.857563824786596e-05,
"loss": 230.4105,
"step": 2915
},
{
"epoch": 0.584,
"grad_norm": 56.33414840698242,
"learning_rate": 8.822892811905237e-05,
"loss": 227.1643,
"step": 2920
},
{
"epoch": 0.585,
"grad_norm": 60.07317352294922,
"learning_rate": 8.788236141703498e-05,
"loss": 234.3683,
"step": 2925
},
{
"epoch": 0.586,
"grad_norm": 162.65834045410156,
"learning_rate": 8.753594236461957e-05,
"loss": 232.3262,
"step": 2930
},
{
"epoch": 0.587,
"grad_norm": 54.98447799682617,
"learning_rate": 8.718967518281307e-05,
"loss": 244.4766,
"step": 2935
},
{
"epoch": 0.588,
"grad_norm": 58.16379928588867,
"learning_rate": 8.684356409077176e-05,
"loss": 231.27,
"step": 2940
},
{
"epoch": 0.589,
"grad_norm": 61.21174240112305,
"learning_rate": 8.64976133057501e-05,
"loss": 232.6313,
"step": 2945
},
{
"epoch": 0.59,
"grad_norm": 60.70001220703125,
"learning_rate": 8.615182704304918e-05,
"loss": 233.5037,
"step": 2950
},
{
"epoch": 0.591,
"grad_norm": 67.02996826171875,
"learning_rate": 8.580620951596557e-05,
"loss": 244.4164,
"step": 2955
},
{
"epoch": 0.592,
"grad_norm": 59.221439361572266,
"learning_rate": 8.546076493573972e-05,
"loss": 244.167,
"step": 2960
},
{
"epoch": 0.593,
"grad_norm": 59.17390060424805,
"learning_rate": 8.511549751150479e-05,
"loss": 232.5588,
"step": 2965
},
{
"epoch": 0.594,
"grad_norm": 127.00204467773438,
"learning_rate": 8.477041145023546e-05,
"loss": 224.3063,
"step": 2970
},
{
"epoch": 0.595,
"grad_norm": 64.2029037475586,
"learning_rate": 8.442551095669639e-05,
"loss": 233.2887,
"step": 2975
},
{
"epoch": 0.596,
"grad_norm": 60.42882537841797,
"learning_rate": 8.408080023339133e-05,
"loss": 238.2239,
"step": 2980
},
{
"epoch": 0.597,
"grad_norm": 72.35857391357422,
"learning_rate": 8.373628348051165e-05,
"loss": 229.4668,
"step": 2985
},
{
"epoch": 0.598,
"grad_norm": 54.91132354736328,
"learning_rate": 8.339196489588523e-05,
"loss": 236.2862,
"step": 2990
},
{
"epoch": 0.599,
"grad_norm": 60.4171142578125,
"learning_rate": 8.30478486749254e-05,
"loss": 237.8247,
"step": 2995
},
{
"epoch": 0.6,
"grad_norm": 136.08200073242188,
"learning_rate": 8.270393901057964e-05,
"loss": 242.6742,
"step": 3000
},
{
"epoch": 0.601,
"grad_norm": 70.57664489746094,
"learning_rate": 8.236024009327879e-05,
"loss": 246.3756,
"step": 3005
},
{
"epoch": 0.602,
"grad_norm": 58.46903610229492,
"learning_rate": 8.201675611088558e-05,
"loss": 241.5537,
"step": 3010
},
{
"epoch": 0.603,
"grad_norm": 57.734466552734375,
"learning_rate": 8.167349124864405e-05,
"loss": 240.1889,
"step": 3015
},
{
"epoch": 0.604,
"grad_norm": 64.36956024169922,
"learning_rate": 8.133044968912811e-05,
"loss": 230.3479,
"step": 3020
},
{
"epoch": 0.605,
"grad_norm": 68.69734954833984,
"learning_rate": 8.0987635612191e-05,
"loss": 245.7289,
"step": 3025
},
{
"epoch": 0.606,
"grad_norm": 55.32661437988281,
"learning_rate": 8.064505319491398e-05,
"loss": 229.8225,
"step": 3030
},
{
"epoch": 0.607,
"grad_norm": 83.67379760742188,
"learning_rate": 8.030270661155574e-05,
"loss": 224.3079,
"step": 3035
},
{
"epoch": 0.608,
"grad_norm": 57.175315856933594,
"learning_rate": 7.996060003350139e-05,
"loss": 246.3686,
"step": 3040
},
{
"epoch": 0.609,
"grad_norm": 60.30739974975586,
"learning_rate": 7.961873762921153e-05,
"loss": 230.407,
"step": 3045
},
{
"epoch": 0.61,
"grad_norm": 59.62334442138672,
"learning_rate": 7.927712356417176e-05,
"loss": 235.6676,
"step": 3050
},
{
"epoch": 0.611,
"grad_norm": 57.50238037109375,
"learning_rate": 7.893576200084159e-05,
"loss": 236.9519,
"step": 3055
},
{
"epoch": 0.612,
"grad_norm": 59.17082595825195,
"learning_rate": 7.8594657098604e-05,
"loss": 229.6581,
"step": 3060
},
{
"epoch": 0.613,
"grad_norm": 57.422332763671875,
"learning_rate": 7.825381301371452e-05,
"loss": 227.6457,
"step": 3065
},
{
"epoch": 0.614,
"grad_norm": 69.13914489746094,
"learning_rate": 7.791323389925084e-05,
"loss": 244.3604,
"step": 3070
},
{
"epoch": 0.615,
"grad_norm": 55.45825958251953,
"learning_rate": 7.75729239050619e-05,
"loss": 239.3342,
"step": 3075
},
{
"epoch": 0.616,
"grad_norm": 106.14602661132812,
"learning_rate": 7.723288717771761e-05,
"loss": 234.9215,
"step": 3080
},
{
"epoch": 0.617,
"grad_norm": 53.32939147949219,
"learning_rate": 7.689312786045823e-05,
"loss": 248.6481,
"step": 3085
},
{
"epoch": 0.618,
"grad_norm": 57.58500289916992,
"learning_rate": 7.655365009314374e-05,
"loss": 237.943,
"step": 3090
},
{
"epoch": 0.619,
"grad_norm": 60.41383743286133,
"learning_rate": 7.621445801220371e-05,
"loss": 241.1509,
"step": 3095
},
{
"epoch": 0.62,
"grad_norm": 62.4730224609375,
"learning_rate": 7.587555575058649e-05,
"loss": 226.4815,
"step": 3100
},
{
"epoch": 0.621,
"grad_norm": 60.28714370727539,
"learning_rate": 7.553694743770928e-05,
"loss": 238.3285,
"step": 3105
},
{
"epoch": 0.622,
"grad_norm": 55.270790100097656,
"learning_rate": 7.519863719940748e-05,
"loss": 231.9199,
"step": 3110
},
{
"epoch": 0.623,
"grad_norm": 58.7408332824707,
"learning_rate": 7.486062915788452e-05,
"loss": 233.2264,
"step": 3115
},
{
"epoch": 0.624,
"grad_norm": 56.40263748168945,
"learning_rate": 7.45229274316618e-05,
"loss": 228.0927,
"step": 3120
},
{
"epoch": 0.625,
"grad_norm": 55.0942497253418,
"learning_rate": 7.418553613552825e-05,
"loss": 239.51,
"step": 3125
},
{
"epoch": 0.626,
"grad_norm": 99.64218139648438,
"learning_rate": 7.384845938049031e-05,
"loss": 216.8472,
"step": 3130
},
{
"epoch": 0.627,
"grad_norm": 105.22872924804688,
"learning_rate": 7.351170127372191e-05,
"loss": 157.4052,
"step": 3135
},
{
"epoch": 0.628,
"grad_norm": 82.74718475341797,
"learning_rate": 7.317526591851433e-05,
"loss": 133.1311,
"step": 3140
},
{
"epoch": 0.629,
"grad_norm": 87.29386901855469,
"learning_rate": 7.283915741422612e-05,
"loss": 133.5219,
"step": 3145
},
{
"epoch": 0.63,
"grad_norm": 62.17641067504883,
"learning_rate": 7.250337985623342e-05,
"loss": 130.8998,
"step": 3150
},
{
"epoch": 0.631,
"grad_norm": 62.76249694824219,
"learning_rate": 7.216793733587976e-05,
"loss": 179.7419,
"step": 3155
},
{
"epoch": 0.632,
"grad_norm": 64.08108520507812,
"learning_rate": 7.183283394042634e-05,
"loss": 217.4482,
"step": 3160
},
{
"epoch": 0.633,
"grad_norm": 60.380897521972656,
"learning_rate": 7.149807375300239e-05,
"loss": 237.668,
"step": 3165
},
{
"epoch": 0.634,
"grad_norm": 67.55339813232422,
"learning_rate": 7.11636608525551e-05,
"loss": 234.2804,
"step": 3170
},
{
"epoch": 0.635,
"grad_norm": 61.59585189819336,
"learning_rate": 7.082959931380011e-05,
"loss": 239.367,
"step": 3175
},
{
"epoch": 0.636,
"grad_norm": 52.35560607910156,
"learning_rate": 7.049589320717186e-05,
"loss": 227.8226,
"step": 3180
},
{
"epoch": 0.637,
"grad_norm": 77.83943939208984,
"learning_rate": 7.016254659877398e-05,
"loss": 232.8947,
"step": 3185
},
{
"epoch": 0.638,
"grad_norm": 65.90837860107422,
"learning_rate": 6.982956355032968e-05,
"loss": 227.8474,
"step": 3190
},
{
"epoch": 0.639,
"grad_norm": 61.64802932739258,
"learning_rate": 6.949694811913225e-05,
"loss": 228.967,
"step": 3195
},
{
"epoch": 0.64,
"grad_norm": 55.6593017578125,
"learning_rate": 6.916470435799587e-05,
"loss": 220.3826,
"step": 3200
},
{
"epoch": 0.641,
"grad_norm": 61.60590744018555,
"learning_rate": 6.883283631520582e-05,
"loss": 220.2009,
"step": 3205
},
{
"epoch": 0.642,
"grad_norm": 63.62286376953125,
"learning_rate": 6.850134803446954e-05,
"loss": 231.0152,
"step": 3210
},
{
"epoch": 0.643,
"grad_norm": 67.09355163574219,
"learning_rate": 6.817024355486706e-05,
"loss": 221.5273,
"step": 3215
},
{
"epoch": 0.644,
"grad_norm": 72.17182922363281,
"learning_rate": 6.783952691080203e-05,
"loss": 230.9787,
"step": 3220
},
{
"epoch": 0.645,
"grad_norm": 53.96472930908203,
"learning_rate": 6.750920213195238e-05,
"loss": 239.7588,
"step": 3225
},
{
"epoch": 0.646,
"grad_norm": 59.04903030395508,
"learning_rate": 6.717927324322124e-05,
"loss": 220.5329,
"step": 3230
},
{
"epoch": 0.647,
"grad_norm": 65.50021362304688,
"learning_rate": 6.684974426468808e-05,
"loss": 249.2473,
"step": 3235
},
{
"epoch": 0.648,
"grad_norm": 59.610469818115234,
"learning_rate": 6.652061921155943e-05,
"loss": 236.1873,
"step": 3240
},
{
"epoch": 0.649,
"grad_norm": 64.26549530029297,
"learning_rate": 6.61919020941203e-05,
"loss": 237.7919,
"step": 3245
},
{
"epoch": 0.65,
"grad_norm": 74.63209533691406,
"learning_rate": 6.58635969176849e-05,
"loss": 245.7763,
"step": 3250
},
{
"epoch": 0.651,
"grad_norm": 60.74374771118164,
"learning_rate": 6.55357076825483e-05,
"loss": 228.4005,
"step": 3255
},
{
"epoch": 0.652,
"grad_norm": 62.64879608154297,
"learning_rate": 6.520823838393731e-05,
"loss": 229.0408,
"step": 3260
},
{
"epoch": 0.653,
"grad_norm": 58.77146911621094,
"learning_rate": 6.488119301196201e-05,
"loss": 235.8309,
"step": 3265
},
{
"epoch": 0.654,
"grad_norm": 67.26822662353516,
"learning_rate": 6.455457555156705e-05,
"loss": 223.6326,
"step": 3270
},
{
"epoch": 0.655,
"grad_norm": 64.14057922363281,
"learning_rate": 6.422838998248307e-05,
"loss": 222.6345,
"step": 3275
},
{
"epoch": 0.656,
"grad_norm": 58.894004821777344,
"learning_rate": 6.390264027917836e-05,
"loss": 229.5242,
"step": 3280
},
{
"epoch": 0.657,
"grad_norm": 158.8521270751953,
"learning_rate": 6.357733041081018e-05,
"loss": 245.897,
"step": 3285
},
{
"epoch": 0.658,
"grad_norm": 61.3464241027832,
"learning_rate": 6.325246434117668e-05,
"loss": 225.3231,
"step": 3290
},
{
"epoch": 0.659,
"grad_norm": 60.92818069458008,
"learning_rate": 6.292804602866834e-05,
"loss": 246.2732,
"step": 3295
},
{
"epoch": 0.66,
"grad_norm": 56.73927688598633,
"learning_rate": 6.260407942621998e-05,
"loss": 235.801,
"step": 3300
},
{
"epoch": 0.661,
"grad_norm": 75.34114074707031,
"learning_rate": 6.228056848126236e-05,
"loss": 235.6205,
"step": 3305
},
{
"epoch": 0.662,
"grad_norm": 58.10975646972656,
"learning_rate": 6.195751713567427e-05,
"loss": 227.5968,
"step": 3310
},
{
"epoch": 0.663,
"grad_norm": 60.49566650390625,
"learning_rate": 6.163492932573438e-05,
"loss": 233.5354,
"step": 3315
},
{
"epoch": 0.664,
"grad_norm": 125.74470520019531,
"learning_rate": 6.131280898207339e-05,
"loss": 235.8063,
"step": 3320
},
{
"epoch": 0.665,
"grad_norm": 77.7129898071289,
"learning_rate": 6.099116002962604e-05,
"loss": 241.5039,
"step": 3325
},
{
"epoch": 0.666,
"grad_norm": 60.046573638916016,
"learning_rate": 6.066998638758326e-05,
"loss": 229.5891,
"step": 3330
},
{
"epoch": 0.667,
"grad_norm": 57.028751373291016,
"learning_rate": 6.034929196934459e-05,
"loss": 232.005,
"step": 3335
},
{
"epoch": 0.668,
"grad_norm": 60.27423858642578,
"learning_rate": 6.002908068247024e-05,
"loss": 237.8356,
"step": 3340
},
{
"epoch": 0.669,
"grad_norm": 56.60969161987305,
"learning_rate": 5.9709356428633746e-05,
"loss": 249.0651,
"step": 3345
},
{
"epoch": 0.67,
"grad_norm": 58.64129638671875,
"learning_rate": 5.9390123103574215e-05,
"loss": 239.3238,
"step": 3350
},
{
"epoch": 0.671,
"grad_norm": 70.50381469726562,
"learning_rate": 5.907138459704895e-05,
"loss": 236.6368,
"step": 3355
},
{
"epoch": 0.672,
"grad_norm": 55.20785903930664,
"learning_rate": 5.8753144792786096e-05,
"loss": 238.1108,
"step": 3360
},
{
"epoch": 0.673,
"grad_norm": 57.345367431640625,
"learning_rate": 5.8435407568437216e-05,
"loss": 227.6418,
"step": 3365
},
{
"epoch": 0.674,
"grad_norm": 77.11515808105469,
"learning_rate": 5.8118176795530176e-05,
"loss": 240.6088,
"step": 3370
},
{
"epoch": 0.675,
"grad_norm": 60.07567596435547,
"learning_rate": 5.780145633942173e-05,
"loss": 230.2157,
"step": 3375
},
{
"epoch": 0.676,
"grad_norm": 70.93324279785156,
"learning_rate": 5.748525005925074e-05,
"loss": 240.1354,
"step": 3380
},
{
"epoch": 0.677,
"grad_norm": 55.7655143737793,
"learning_rate": 5.716956180789098e-05,
"loss": 235.4092,
"step": 3385
},
{
"epoch": 0.678,
"grad_norm": 62.266727447509766,
"learning_rate": 5.6854395431904094e-05,
"loss": 220.1855,
"step": 3390
},
{
"epoch": 0.679,
"grad_norm": 57.26567077636719,
"learning_rate": 5.653975477149298e-05,
"loss": 229.9771,
"step": 3395
},
{
"epoch": 0.68,
"grad_norm": 63.264583587646484,
"learning_rate": 5.622564366045472e-05,
"loss": 227.0424,
"step": 3400
},
{
"epoch": 0.681,
"grad_norm": 58.400047302246094,
"learning_rate": 5.591206592613416e-05,
"loss": 229.6065,
"step": 3405
},
{
"epoch": 0.682,
"grad_norm": 58.05074691772461,
"learning_rate": 5.5599025389376935e-05,
"loss": 240.6718,
"step": 3410
},
{
"epoch": 0.683,
"grad_norm": 57.55448913574219,
"learning_rate": 5.5286525864483285e-05,
"loss": 233.1961,
"step": 3415
},
{
"epoch": 0.684,
"grad_norm": 63.032962799072266,
"learning_rate": 5.497457115916127e-05,
"loss": 231.0708,
"step": 3420
},
{
"epoch": 0.685,
"grad_norm": 58.53230285644531,
"learning_rate": 5.466316507448048e-05,
"loss": 228.709,
"step": 3425
},
{
"epoch": 0.686,
"grad_norm": 58.597530364990234,
"learning_rate": 5.435231140482587e-05,
"loss": 226.7799,
"step": 3430
},
{
"epoch": 0.687,
"grad_norm": 59.05603790283203,
"learning_rate": 5.404201393785122e-05,
"loss": 221.3859,
"step": 3435
},
{
"epoch": 0.688,
"grad_norm": 64.90831756591797,
"learning_rate": 5.373227645443332e-05,
"loss": 228.4797,
"step": 3440
},
{
"epoch": 0.689,
"grad_norm": 65.26719665527344,
"learning_rate": 5.3423102728625574e-05,
"loss": 224.3567,
"step": 3445
},
{
"epoch": 0.69,
"grad_norm": 76.78877258300781,
"learning_rate": 5.311449652761235e-05,
"loss": 241.4243,
"step": 3450
},
{
"epoch": 0.691,
"grad_norm": 63.909393310546875,
"learning_rate": 5.2806461611662735e-05,
"loss": 239.889,
"step": 3455
},
{
"epoch": 0.692,
"grad_norm": 58.965003967285156,
"learning_rate": 5.2499001734085044e-05,
"loss": 231.3724,
"step": 3460
},
{
"epoch": 0.693,
"grad_norm": 57.41484069824219,
"learning_rate": 5.2192120641180786e-05,
"loss": 215.7092,
"step": 3465
},
{
"epoch": 0.694,
"grad_norm": 62.95941162109375,
"learning_rate": 5.188582207219931e-05,
"loss": 226.5267,
"step": 3470
},
{
"epoch": 0.695,
"grad_norm": 61.71487045288086,
"learning_rate": 5.158010975929193e-05,
"loss": 236.4199,
"step": 3475
},
{
"epoch": 0.696,
"grad_norm": 56.94986343383789,
"learning_rate": 5.127498742746675e-05,
"loss": 247.0517,
"step": 3480
},
{
"epoch": 0.697,
"grad_norm": 120.18292236328125,
"learning_rate": 5.097045879454313e-05,
"loss": 232.1091,
"step": 3485
},
{
"epoch": 0.698,
"grad_norm": 60.40031433105469,
"learning_rate": 5.066652757110628e-05,
"loss": 234.0605,
"step": 3490
},
{
"epoch": 0.699,
"grad_norm": 59.25654220581055,
"learning_rate": 5.036319746046232e-05,
"loss": 239.6799,
"step": 3495
},
{
"epoch": 0.7,
"grad_norm": 61.800086975097656,
"learning_rate": 5.006047215859289e-05,
"loss": 235.1895,
"step": 3500
},
{
"epoch": 0.701,
"grad_norm": 76.94608306884766,
"learning_rate": 4.97583553541102e-05,
"loss": 235.3422,
"step": 3505
},
{
"epoch": 0.702,
"grad_norm": 55.355552673339844,
"learning_rate": 4.945685072821227e-05,
"loss": 212.9436,
"step": 3510
},
{
"epoch": 0.703,
"grad_norm": 71.21341705322266,
"learning_rate": 4.915596195463773e-05,
"loss": 218.8237,
"step": 3515
},
{
"epoch": 0.704,
"grad_norm": 83.03817749023438,
"learning_rate": 4.885569269962142e-05,
"loss": 224.1068,
"step": 3520
},
{
"epoch": 0.705,
"grad_norm": 67.02042388916016,
"learning_rate": 4.8556046621849346e-05,
"loss": 228.4471,
"step": 3525
},
{
"epoch": 0.706,
"grad_norm": 56.48438262939453,
"learning_rate": 4.825702737241452e-05,
"loss": 209.1283,
"step": 3530
},
{
"epoch": 0.707,
"grad_norm": 60.945011138916016,
"learning_rate": 4.7958638594772064e-05,
"loss": 227.8682,
"step": 3535
},
{
"epoch": 0.708,
"grad_norm": 84.68537902832031,
"learning_rate": 4.766088392469506e-05,
"loss": 175.3404,
"step": 3540
},
{
"epoch": 0.709,
"grad_norm": 106.09636688232422,
"learning_rate": 4.736376699023023e-05,
"loss": 182.3729,
"step": 3545
},
{
"epoch": 0.71,
"grad_norm": 55.742774963378906,
"learning_rate": 4.706729141165361e-05,
"loss": 246.4256,
"step": 3550
},
{
"epoch": 0.711,
"grad_norm": 56.66408920288086,
"learning_rate": 4.6771460801426635e-05,
"loss": 232.1106,
"step": 3555
},
{
"epoch": 0.712,
"grad_norm": 56.370574951171875,
"learning_rate": 4.647627876415186e-05,
"loss": 245.1462,
"step": 3560
},
{
"epoch": 0.713,
"grad_norm": 74.42236328125,
"learning_rate": 4.6181748896529273e-05,
"loss": 228.35,
"step": 3565
},
{
"epoch": 0.714,
"grad_norm": 57.87800598144531,
"learning_rate": 4.588787478731242e-05,
"loss": 234.2479,
"step": 3570
},
{
"epoch": 0.715,
"grad_norm": 59.219154357910156,
"learning_rate": 4.559466001726451e-05,
"loss": 213.6988,
"step": 3575
},
{
"epoch": 0.716,
"grad_norm": 58.77073669433594,
"learning_rate": 4.530210815911504e-05,
"loss": 218.4181,
"step": 3580
},
{
"epoch": 0.717,
"grad_norm": 61.42266845703125,
"learning_rate": 4.501022277751602e-05,
"loss": 221.6765,
"step": 3585
},
{
"epoch": 0.718,
"grad_norm": 59.34172821044922,
"learning_rate": 4.471900742899876e-05,
"loss": 224.4874,
"step": 3590
},
{
"epoch": 0.719,
"grad_norm": 65.31617736816406,
"learning_rate": 4.4428465661930343e-05,
"loss": 221.9394,
"step": 3595
},
{
"epoch": 0.72,
"grad_norm": 57.800697326660156,
"learning_rate": 4.413860101647055e-05,
"loss": 238.7757,
"step": 3600
},
{
"epoch": 0.721,
"grad_norm": 53.778724670410156,
"learning_rate": 4.3849417024528564e-05,
"loss": 286.4774,
"step": 3605
},
{
"epoch": 0.722,
"grad_norm": 65.23345184326172,
"learning_rate": 4.356091720972011e-05,
"loss": 237.3981,
"step": 3610
},
{
"epoch": 0.723,
"grad_norm": 63.572113037109375,
"learning_rate": 4.327310508732437e-05,
"loss": 229.8879,
"step": 3615
},
{
"epoch": 0.724,
"grad_norm": 71.95780181884766,
"learning_rate": 4.29859841642412e-05,
"loss": 246.5536,
"step": 3620
},
{
"epoch": 0.725,
"grad_norm": 60.0914306640625,
"learning_rate": 4.26995579389485e-05,
"loss": 224.2951,
"step": 3625
},
{
"epoch": 0.726,
"grad_norm": 68.65311431884766,
"learning_rate": 4.2413829901459344e-05,
"loss": 250.7042,
"step": 3630
},
{
"epoch": 0.727,
"grad_norm": 58.95090103149414,
"learning_rate": 4.212880353327979e-05,
"loss": 235.579,
"step": 3635
},
{
"epoch": 0.728,
"grad_norm": 56.65614700317383,
"learning_rate": 4.184448230736613e-05,
"loss": 237.8454,
"step": 3640
},
{
"epoch": 0.729,
"grad_norm": 67.07494354248047,
"learning_rate": 4.15608696880828e-05,
"loss": 233.2988,
"step": 3645
},
{
"epoch": 0.73,
"grad_norm": 57.59606170654297,
"learning_rate": 4.1277969131160045e-05,
"loss": 226.3549,
"step": 3650
},
{
"epoch": 0.731,
"grad_norm": 85.04194641113281,
"learning_rate": 4.099578408365191e-05,
"loss": 226.1134,
"step": 3655
},
{
"epoch": 0.732,
"grad_norm": 58.70685577392578,
"learning_rate": 4.071431798389408e-05,
"loss": 226.9804,
"step": 3660
},
{
"epoch": 0.733,
"grad_norm": 53.8144416809082,
"learning_rate": 4.0433574261462206e-05,
"loss": 221.1446,
"step": 3665
},
{
"epoch": 0.734,
"grad_norm": 55.033111572265625,
"learning_rate": 4.015355633712996e-05,
"loss": 234.1043,
"step": 3670
},
{
"epoch": 0.735,
"grad_norm": 62.225669860839844,
"learning_rate": 3.987426762282733e-05,
"loss": 222.0488,
"step": 3675
},
{
"epoch": 0.736,
"grad_norm": 56.79927444458008,
"learning_rate": 3.9595711521599224e-05,
"loss": 240.1699,
"step": 3680
},
{
"epoch": 0.737,
"grad_norm": 68.60034942626953,
"learning_rate": 3.931789142756377e-05,
"loss": 232.5952,
"step": 3685
},
{
"epoch": 0.738,
"grad_norm": 85.82996368408203,
"learning_rate": 3.904081072587119e-05,
"loss": 240.5267,
"step": 3690
},
{
"epoch": 0.739,
"grad_norm": 58.34128189086914,
"learning_rate": 3.876447279266238e-05,
"loss": 231.7139,
"step": 3695
},
{
"epoch": 0.74,
"grad_norm": 58.81357955932617,
"learning_rate": 3.8488880995027786e-05,
"loss": 218.3256,
"step": 3700
},
{
"epoch": 0.741,
"grad_norm": 76.90341186523438,
"learning_rate": 3.821403869096658e-05,
"loss": 233.1653,
"step": 3705
},
{
"epoch": 0.742,
"grad_norm": 81.11316680908203,
"learning_rate": 3.793994922934544e-05,
"loss": 225.9958,
"step": 3710
},
{
"epoch": 0.743,
"grad_norm": 53.793087005615234,
"learning_rate": 3.766661594985801e-05,
"loss": 249.5571,
"step": 3715
},
{
"epoch": 0.744,
"grad_norm": 85.10948181152344,
"learning_rate": 3.739404218298398e-05,
"loss": 239.715,
"step": 3720
},
{
"epoch": 0.745,
"grad_norm": 55.54294204711914,
"learning_rate": 3.712223124994875e-05,
"loss": 239.5781,
"step": 3725
},
{
"epoch": 0.746,
"grad_norm": 58.83379364013672,
"learning_rate": 3.685118646268272e-05,
"loss": 227.0683,
"step": 3730
},
{
"epoch": 0.747,
"grad_norm": 89.98795318603516,
"learning_rate": 3.6580911123781056e-05,
"loss": 237.3153,
"step": 3735
},
{
"epoch": 0.748,
"grad_norm": 64.54805755615234,
"learning_rate": 3.631140852646355e-05,
"loss": 243.8812,
"step": 3740
},
{
"epoch": 0.749,
"grad_norm": 55.502037048339844,
"learning_rate": 3.604268195453421e-05,
"loss": 248.7211,
"step": 3745
},
{
"epoch": 0.75,
"grad_norm": 60.66739273071289,
"learning_rate": 3.577473468234156e-05,
"loss": 237.7394,
"step": 3750
},
{
"epoch": 0.751,
"grad_norm": 73.82022857666016,
"learning_rate": 3.5507569974738574e-05,
"loss": 246.7511,
"step": 3755
},
{
"epoch": 0.752,
"grad_norm": 62.20229721069336,
"learning_rate": 3.524119108704286e-05,
"loss": 238.2229,
"step": 3760
},
{
"epoch": 0.753,
"grad_norm": 56.83375930786133,
"learning_rate": 3.497560126499709e-05,
"loss": 228.7374,
"step": 3765
},
{
"epoch": 0.754,
"grad_norm": 57.045745849609375,
"learning_rate": 3.4710803744729515e-05,
"loss": 237.407,
"step": 3770
},
{
"epoch": 0.755,
"grad_norm": 60.47734832763672,
"learning_rate": 3.444680175271428e-05,
"loss": 241.6857,
"step": 3775
},
{
"epoch": 0.756,
"grad_norm": 58.87367630004883,
"learning_rate": 3.418359850573234e-05,
"loss": 237.9792,
"step": 3780
},
{
"epoch": 0.757,
"grad_norm": 58.008602142333984,
"learning_rate": 3.3921197210832235e-05,
"loss": 229.2696,
"step": 3785
},
{
"epoch": 0.758,
"grad_norm": 56.2470588684082,
"learning_rate": 3.3659601065290893e-05,
"loss": 249.7396,
"step": 3790
},
{
"epoch": 0.759,
"grad_norm": 69.34009552001953,
"learning_rate": 3.339881325657484e-05,
"loss": 240.9706,
"step": 3795
},
{
"epoch": 0.76,
"grad_norm": 62.69635772705078,
"learning_rate": 3.313883696230119e-05,
"loss": 226.9666,
"step": 3800
},
{
"epoch": 0.761,
"grad_norm": 60.74796676635742,
"learning_rate": 3.287967535019908e-05,
"loss": 223.5108,
"step": 3805
},
{
"epoch": 0.762,
"grad_norm": 64.01351928710938,
"learning_rate": 3.2621331578070934e-05,
"loss": 250.8177,
"step": 3810
},
{
"epoch": 0.763,
"grad_norm": 60.843807220458984,
"learning_rate": 3.236380879375408e-05,
"loss": 239.6298,
"step": 3815
},
{
"epoch": 0.764,
"grad_norm": 58.698387145996094,
"learning_rate": 3.210711013508242e-05,
"loss": 240.155,
"step": 3820
},
{
"epoch": 0.765,
"grad_norm": 54.22844314575195,
"learning_rate": 3.1851238729848034e-05,
"loss": 232.3402,
"step": 3825
},
{
"epoch": 0.766,
"grad_norm": 67.27794647216797,
"learning_rate": 3.159619769576333e-05,
"loss": 212.4046,
"step": 3830
},
{
"epoch": 0.767,
"grad_norm": 64.98332977294922,
"learning_rate": 3.134199014042274e-05,
"loss": 221.7149,
"step": 3835
},
{
"epoch": 0.768,
"grad_norm": 57.6930046081543,
"learning_rate": 3.108861916126518e-05,
"loss": 228.1428,
"step": 3840
},
{
"epoch": 0.769,
"grad_norm": 64.71436309814453,
"learning_rate": 3.0836087845536e-05,
"loss": 229.4941,
"step": 3845
},
{
"epoch": 0.77,
"grad_norm": 59.681800842285156,
"learning_rate": 3.058439927024962e-05,
"loss": 213.9398,
"step": 3850
},
{
"epoch": 0.771,
"grad_norm": 55.54667282104492,
"learning_rate": 3.0333556502151926e-05,
"loss": 236.1587,
"step": 3855
},
{
"epoch": 0.772,
"grad_norm": 61.9726448059082,
"learning_rate": 3.0083562597682847e-05,
"loss": 246.4536,
"step": 3860
},
{
"epoch": 0.773,
"grad_norm": 66.04378509521484,
"learning_rate": 2.9834420602939263e-05,
"loss": 232.1891,
"step": 3865
},
{
"epoch": 0.774,
"grad_norm": 66.83670043945312,
"learning_rate": 2.9586133553637683e-05,
"loss": 232.6448,
"step": 3870
},
{
"epoch": 0.775,
"grad_norm": 61.186588287353516,
"learning_rate": 2.933870447507753e-05,
"loss": 230.2862,
"step": 3875
},
{
"epoch": 0.776,
"grad_norm": 54.34227752685547,
"learning_rate": 2.9092136382103973e-05,
"loss": 233.7255,
"step": 3880
},
{
"epoch": 0.777,
"grad_norm": 56.309364318847656,
"learning_rate": 2.8846432279071467e-05,
"loss": 220.2837,
"step": 3885
},
{
"epoch": 0.778,
"grad_norm": 62.381900787353516,
"learning_rate": 2.860159515980695e-05,
"loss": 240.8807,
"step": 3890
},
{
"epoch": 0.779,
"grad_norm": 62.585479736328125,
"learning_rate": 2.835762800757341e-05,
"loss": 227.0986,
"step": 3895
},
{
"epoch": 0.78,
"grad_norm": 65.24736785888672,
"learning_rate": 2.8114533795033683e-05,
"loss": 234.3742,
"step": 3900
},
{
"epoch": 0.781,
"grad_norm": 100.97679138183594,
"learning_rate": 2.7872315484213952e-05,
"loss": 216.9641,
"step": 3905
},
{
"epoch": 0.782,
"grad_norm": 52.50382614135742,
"learning_rate": 2.7630976026467968e-05,
"loss": 224.2065,
"step": 3910
},
{
"epoch": 0.783,
"grad_norm": 77.02485656738281,
"learning_rate": 2.7390518362440808e-05,
"loss": 238.4704,
"step": 3915
},
{
"epoch": 0.784,
"grad_norm": 57.394798278808594,
"learning_rate": 2.715094542203327e-05,
"loss": 210.1653,
"step": 3920
},
{
"epoch": 0.785,
"grad_norm": 60.25807571411133,
"learning_rate": 2.6912260124366006e-05,
"loss": 240.164,
"step": 3925
},
{
"epoch": 0.786,
"grad_norm": 55.640377044677734,
"learning_rate": 2.6674465377744017e-05,
"loss": 233.1223,
"step": 3930
},
{
"epoch": 0.787,
"grad_norm": 54.346595764160156,
"learning_rate": 2.6437564079621267e-05,
"loss": 239.7916,
"step": 3935
},
{
"epoch": 0.788,
"grad_norm": 59.0637092590332,
"learning_rate": 2.6201559116565345e-05,
"loss": 217.0782,
"step": 3940
},
{
"epoch": 0.789,
"grad_norm": 56.549293518066406,
"learning_rate": 2.5966453364222186e-05,
"loss": 234.8658,
"step": 3945
},
{
"epoch": 0.79,
"grad_norm": 58.505035400390625,
"learning_rate": 2.573224968728123e-05,
"loss": 239.6084,
"step": 3950
},
{
"epoch": 0.791,
"grad_norm": 59.75248336791992,
"learning_rate": 2.549895093944039e-05,
"loss": 220.2904,
"step": 3955
},
{
"epoch": 0.792,
"grad_norm": 55.047000885009766,
"learning_rate": 2.5266559963371216e-05,
"loss": 238.7349,
"step": 3960
},
{
"epoch": 0.793,
"grad_norm": 67.48583984375,
"learning_rate": 2.5035079590684497e-05,
"loss": 233.7797,
"step": 3965
},
{
"epoch": 0.794,
"grad_norm": 66.99308776855469,
"learning_rate": 2.480451264189546e-05,
"loss": 232.5435,
"step": 3970
},
{
"epoch": 0.795,
"grad_norm": 58.150306701660156,
"learning_rate": 2.4574861926389615e-05,
"loss": 238.8692,
"step": 3975
},
{
"epoch": 0.796,
"grad_norm": 57.683746337890625,
"learning_rate": 2.43461302423885e-05,
"loss": 240.2565,
"step": 3980
},
{
"epoch": 0.797,
"grad_norm": 74.0052261352539,
"learning_rate": 2.411832037691545e-05,
"loss": 221.1994,
"step": 3985
},
{
"epoch": 0.798,
"grad_norm": 52.258670806884766,
"learning_rate": 2.389143510576184e-05,
"loss": 223.9109,
"step": 3990
},
{
"epoch": 0.799,
"grad_norm": 64.43135833740234,
"learning_rate": 2.3665477193453034e-05,
"loss": 229.5644,
"step": 3995
},
{
"epoch": 0.8,
"grad_norm": 55.17902755737305,
"learning_rate": 2.3440449393214948e-05,
"loss": 227.9993,
"step": 4000
},
{
"epoch": 0.801,
"grad_norm": 52.14693069458008,
"learning_rate": 2.321635444694028e-05,
"loss": 224.2135,
"step": 4005
},
{
"epoch": 0.802,
"grad_norm": 123.734375,
"learning_rate": 2.2993195085155205e-05,
"loss": 245.3664,
"step": 4010
},
{
"epoch": 0.803,
"grad_norm": 57.668914794921875,
"learning_rate": 2.277097402698619e-05,
"loss": 229.9262,
"step": 4015
},
{
"epoch": 0.804,
"grad_norm": 111.40909576416016,
"learning_rate": 2.254969398012663e-05,
"loss": 232.9373,
"step": 4020
},
{
"epoch": 0.805,
"grad_norm": 61.80019760131836,
"learning_rate": 2.2329357640804117e-05,
"loss": 237.4311,
"step": 4025
},
{
"epoch": 0.806,
"grad_norm": 57.69606018066406,
"learning_rate": 2.210996769374737e-05,
"loss": 232.8376,
"step": 4030
},
{
"epoch": 0.807,
"grad_norm": 60.3912467956543,
"learning_rate": 2.1891526812153672e-05,
"loss": 228.6421,
"step": 4035
},
{
"epoch": 0.808,
"grad_norm": 55.70375442504883,
"learning_rate": 2.1674037657656266e-05,
"loss": 235.4651,
"step": 4040
},
{
"epoch": 0.809,
"grad_norm": 63.516334533691406,
"learning_rate": 2.1457502880291812e-05,
"loss": 228.4013,
"step": 4045
},
{
"epoch": 0.81,
"grad_norm": 61.39441680908203,
"learning_rate": 2.1241925118468287e-05,
"loss": 226.7115,
"step": 4050
},
{
"epoch": 0.811,
"grad_norm": 61.74851608276367,
"learning_rate": 2.1027306998932627e-05,
"loss": 241.7483,
"step": 4055
},
{
"epoch": 0.812,
"grad_norm": 52.33977127075195,
"learning_rate": 2.0813651136738954e-05,
"loss": 243.618,
"step": 4060
},
{
"epoch": 0.813,
"grad_norm": 60.82252502441406,
"learning_rate": 2.0600960135216462e-05,
"loss": 223.3947,
"step": 4065
},
{
"epoch": 0.814,
"grad_norm": 55.754234313964844,
"learning_rate": 2.0389236585937945e-05,
"loss": 217.4923,
"step": 4070
},
{
"epoch": 0.815,
"grad_norm": 54.083370208740234,
"learning_rate": 2.017848306868797e-05,
"loss": 242.984,
"step": 4075
},
{
"epoch": 0.816,
"grad_norm": 57.90782928466797,
"learning_rate": 1.9968702151431696e-05,
"loss": 237.8171,
"step": 4080
},
{
"epoch": 0.817,
"grad_norm": 66.3219985961914,
"learning_rate": 1.9759896390283362e-05,
"loss": 232.888,
"step": 4085
},
{
"epoch": 0.818,
"grad_norm": 65.33077239990234,
"learning_rate": 1.955206832947526e-05,
"loss": 230.6478,
"step": 4090
},
{
"epoch": 0.819,
"grad_norm": 68.88883209228516,
"learning_rate": 1.9345220501326777e-05,
"loss": 236.5641,
"step": 4095
},
{
"epoch": 0.82,
"grad_norm": 57.845664978027344,
"learning_rate": 1.9139355426213347e-05,
"loss": 226.5857,
"step": 4100
},
{
"epoch": 0.821,
"grad_norm": 61.666664123535156,
"learning_rate": 1.8934475612536017e-05,
"loss": 232.615,
"step": 4105
},
{
"epoch": 0.822,
"grad_norm": 58.66468048095703,
"learning_rate": 1.8730583556690605e-05,
"loss": 208.3849,
"step": 4110
},
{
"epoch": 0.823,
"grad_norm": 59.37729263305664,
"learning_rate": 1.852768174303752e-05,
"loss": 233.1701,
"step": 4115
},
{
"epoch": 0.824,
"grad_norm": 69.13996124267578,
"learning_rate": 1.8325772643871265e-05,
"loss": 226.5468,
"step": 4120
},
{
"epoch": 0.825,
"grad_norm": 55.20663833618164,
"learning_rate": 1.812485871939056e-05,
"loss": 245.2576,
"step": 4125
},
{
"epoch": 0.826,
"grad_norm": 68.27503204345703,
"learning_rate": 1.792494241766811e-05,
"loss": 238.4567,
"step": 4130
},
{
"epoch": 0.827,
"grad_norm": 64.22474670410156,
"learning_rate": 1.7726026174621003e-05,
"loss": 235.5188,
"step": 4135
},
{
"epoch": 0.828,
"grad_norm": 65.86067962646484,
"learning_rate": 1.752811241398089e-05,
"loss": 219.3787,
"step": 4140
},
{
"epoch": 0.829,
"grad_norm": 59.22248458862305,
"learning_rate": 1.7331203547264453e-05,
"loss": 228.8221,
"step": 4145
},
{
"epoch": 0.83,
"grad_norm": 64.88153839111328,
"learning_rate": 1.7135301973744124e-05,
"loss": 224.6133,
"step": 4150
},
{
"epoch": 0.831,
"grad_norm": 73.89867401123047,
"learning_rate": 1.6940410080418723e-05,
"loss": 235.9304,
"step": 4155
},
{
"epoch": 0.832,
"grad_norm": 60.56105422973633,
"learning_rate": 1.6746530241984504e-05,
"loss": 219.4839,
"step": 4160
},
{
"epoch": 0.833,
"grad_norm": 61.200008392333984,
"learning_rate": 1.6553664820806102e-05,
"loss": 223.4815,
"step": 4165
},
{
"epoch": 0.834,
"grad_norm": 112.37931823730469,
"learning_rate": 1.6361816166887768e-05,
"loss": 239.3813,
"step": 4170
},
{
"epoch": 0.835,
"grad_norm": 86.23741149902344,
"learning_rate": 1.6170986617844863e-05,
"loss": 221.3643,
"step": 4175
},
{
"epoch": 0.836,
"grad_norm": 53.07310485839844,
"learning_rate": 1.598117849887518e-05,
"loss": 242.8684,
"step": 4180
},
{
"epoch": 0.837,
"grad_norm": 88.25235748291016,
"learning_rate": 1.579239412273078e-05,
"loss": 236.7191,
"step": 4185
},
{
"epoch": 0.838,
"grad_norm": 66.78803253173828,
"learning_rate": 1.560463578968967e-05,
"loss": 236.2187,
"step": 4190
},
{
"epoch": 0.839,
"grad_norm": 53.144500732421875,
"learning_rate": 1.541790578752794e-05,
"loss": 235.4785,
"step": 4195
},
{
"epoch": 0.84,
"grad_norm": 78.45801544189453,
"learning_rate": 1.5232206391491699e-05,
"loss": 236.1022,
"step": 4200
},
{
"epoch": 0.841,
"grad_norm": 59.565711975097656,
"learning_rate": 1.5047539864269478e-05,
"loss": 223.2776,
"step": 4205
},
{
"epoch": 0.842,
"grad_norm": 70.48817443847656,
"learning_rate": 1.486390845596466e-05,
"loss": 232.9711,
"step": 4210
},
{
"epoch": 0.843,
"grad_norm": 55.6562385559082,
"learning_rate": 1.468131440406798e-05,
"loss": 225.7114,
"step": 4215
},
{
"epoch": 0.844,
"grad_norm": 57.754940032958984,
"learning_rate": 1.4499759933430346e-05,
"loss": 236.1438,
"step": 4220
},
{
"epoch": 0.845,
"grad_norm": 52.195316314697266,
"learning_rate": 1.4319247256235714e-05,
"loss": 224.9649,
"step": 4225
},
{
"epoch": 0.846,
"grad_norm": 65.36671447753906,
"learning_rate": 1.4139778571974049e-05,
"loss": 220.7384,
"step": 4230
},
{
"epoch": 0.847,
"grad_norm": 58.99745559692383,
"learning_rate": 1.3961356067414666e-05,
"loss": 221.4328,
"step": 4235
},
{
"epoch": 0.848,
"grad_norm": 55.73463439941406,
"learning_rate": 1.3783981916579446e-05,
"loss": 230.3899,
"step": 4240
},
{
"epoch": 0.849,
"grad_norm": 58.30274963378906,
"learning_rate": 1.3607658280716473e-05,
"loss": 233.4058,
"step": 4245
},
{
"epoch": 0.85,
"grad_norm": 55.33646774291992,
"learning_rate": 1.3432387308273575e-05,
"loss": 234.6749,
"step": 4250
},
{
"epoch": 0.851,
"grad_norm": 54.992279052734375,
"learning_rate": 1.3258171134872265e-05,
"loss": 229.7397,
"step": 4255
},
{
"epoch": 0.852,
"grad_norm": 53.384925842285156,
"learning_rate": 1.3085011883281606e-05,
"loss": 223.5759,
"step": 4260
},
{
"epoch": 0.853,
"grad_norm": 55.697025299072266,
"learning_rate": 1.2912911663392469e-05,
"loss": 231.2667,
"step": 4265
},
{
"epoch": 0.854,
"grad_norm": 54.866600036621094,
"learning_rate": 1.2741872572191682e-05,
"loss": 201.2136,
"step": 4270
},
{
"epoch": 0.855,
"grad_norm": 52.937129974365234,
"learning_rate": 1.257189669373664e-05,
"loss": 233.4738,
"step": 4275
},
{
"epoch": 0.856,
"grad_norm": 69.88152313232422,
"learning_rate": 1.2402986099129765e-05,
"loss": 242.4458,
"step": 4280
},
{
"epoch": 0.857,
"grad_norm": 52.642433166503906,
"learning_rate": 1.2235142846493308e-05,
"loss": 226.7807,
"step": 4285
},
{
"epoch": 0.858,
"grad_norm": 54.44631576538086,
"learning_rate": 1.206836898094439e-05,
"loss": 237.4488,
"step": 4290
},
{
"epoch": 0.859,
"grad_norm": 57.18129348754883,
"learning_rate": 1.1902666534569883e-05,
"loss": 244.0537,
"step": 4295
},
{
"epoch": 0.86,
"grad_norm": 56.92108917236328,
"learning_rate": 1.1738037526401857e-05,
"loss": 228.7178,
"step": 4300
},
{
"epoch": 0.861,
"grad_norm": 53.47532272338867,
"learning_rate": 1.1574483962392767e-05,
"loss": 241.2409,
"step": 4305
},
{
"epoch": 0.862,
"grad_norm": 78.33045196533203,
"learning_rate": 1.1412007835391236e-05,
"loss": 225.2705,
"step": 4310
},
{
"epoch": 0.863,
"grad_norm": 61.042747497558594,
"learning_rate": 1.1250611125117527e-05,
"loss": 235.5464,
"step": 4315
},
{
"epoch": 0.864,
"grad_norm": 57.8859977722168,
"learning_rate": 1.1090295798139672e-05,
"loss": 243.981,
"step": 4320
},
{
"epoch": 0.865,
"grad_norm": 119.81159210205078,
"learning_rate": 1.093106380784934e-05,
"loss": 216.4435,
"step": 4325
},
{
"epoch": 0.866,
"grad_norm": 60.223514556884766,
"learning_rate": 1.0772917094438051e-05,
"loss": 241.4054,
"step": 4330
},
{
"epoch": 0.867,
"grad_norm": 62.8787956237793,
"learning_rate": 1.0615857584873623e-05,
"loss": 231.1311,
"step": 4335
},
{
"epoch": 0.868,
"grad_norm": 57.91008758544922,
"learning_rate": 1.0459887192876594e-05,
"loss": 224.3055,
"step": 4340
},
{
"epoch": 0.869,
"grad_norm": 67.75902557373047,
"learning_rate": 1.0305007818897006e-05,
"loss": 213.4049,
"step": 4345
},
{
"epoch": 0.87,
"grad_norm": 63.86601638793945,
"learning_rate": 1.0151221350091134e-05,
"loss": 229.5152,
"step": 4350
},
{
"epoch": 0.871,
"grad_norm": 56.40724182128906,
"learning_rate": 9.998529660298539e-06,
"loss": 232.2357,
"step": 4355
},
{
"epoch": 0.872,
"grad_norm": 56.3475227355957,
"learning_rate": 9.84693461001932e-06,
"loss": 240.0751,
"step": 4360
},
{
"epoch": 0.873,
"grad_norm": 76.1172866821289,
"learning_rate": 9.696438046391288e-06,
"loss": 225.4039,
"step": 4365
},
{
"epoch": 0.874,
"grad_norm": 60.91194152832031,
"learning_rate": 9.5470418031676e-06,
"loss": 233.809,
"step": 4370
},
{
"epoch": 0.875,
"grad_norm": 73.00634765625,
"learning_rate": 9.398747700694322e-06,
"loss": 221.946,
"step": 4375
},
{
"epoch": 0.876,
"grad_norm": 61.197269439697266,
"learning_rate": 9.251557545888312e-06,
"loss": 220.0909,
"step": 4380
},
{
"epoch": 0.877,
"grad_norm": 59.198219299316406,
"learning_rate": 9.105473132215125e-06,
"loss": 222.3656,
"step": 4385
},
{
"epoch": 0.878,
"grad_norm": 54.44852066040039,
"learning_rate": 8.960496239667282e-06,
"loss": 238.4998,
"step": 4390
},
{
"epoch": 0.879,
"grad_norm": 63.17068862915039,
"learning_rate": 8.816628634742441e-06,
"loss": 230.8824,
"step": 4395
},
{
"epoch": 0.88,
"grad_norm": 66.57695007324219,
"learning_rate": 8.67387207042194e-06,
"loss": 224.5499,
"step": 4400
},
{
"epoch": 0.881,
"grad_norm": 58.868507385253906,
"learning_rate": 8.532228286149501e-06,
"loss": 242.6571,
"step": 4405
},
{
"epoch": 0.882,
"grad_norm": 54.390377044677734,
"learning_rate": 8.39169900780995e-06,
"loss": 232.6081,
"step": 4410
},
{
"epoch": 0.883,
"grad_norm": 56.776519775390625,
"learning_rate": 8.252285947708139e-06,
"loss": 230.3372,
"step": 4415
},
{
"epoch": 0.884,
"grad_norm": 74.6689453125,
"learning_rate": 8.113990804548244e-06,
"loss": 242.9896,
"step": 4420
},
{
"epoch": 0.885,
"grad_norm": 69.21456146240234,
"learning_rate": 7.976815263412963e-06,
"loss": 234.1071,
"step": 4425
},
{
"epoch": 0.886,
"grad_norm": 68.24514770507812,
"learning_rate": 7.840760995742946e-06,
"loss": 245.8178,
"step": 4430
},
{
"epoch": 0.887,
"grad_norm": 74.81890106201172,
"learning_rate": 7.7058296593165e-06,
"loss": 242.7295,
"step": 4435
},
{
"epoch": 0.888,
"grad_norm": 54.274539947509766,
"learning_rate": 7.572022898229403e-06,
"loss": 238.5827,
"step": 4440
},
{
"epoch": 0.889,
"grad_norm": 52.99814224243164,
"learning_rate": 7.439342342874789e-06,
"loss": 241.2665,
"step": 4445
},
{
"epoch": 0.89,
"grad_norm": 57.67208480834961,
"learning_rate": 7.3077896099233765e-06,
"loss": 225.8318,
"step": 4450
},
{
"epoch": 0.891,
"grad_norm": 57.720699310302734,
"learning_rate": 7.177366302303667e-06,
"loss": 241.7749,
"step": 4455
},
{
"epoch": 0.892,
"grad_norm": 95.16434478759766,
"learning_rate": 7.048074009182548e-06,
"loss": 225.8215,
"step": 4460
},
{
"epoch": 0.893,
"grad_norm": 58.85953140258789,
"learning_rate": 6.919914305945774e-06,
"loss": 233.0086,
"step": 4465
},
{
"epoch": 0.894,
"grad_norm": 65.57076263427734,
"learning_rate": 6.7928887541789055e-06,
"loss": 240.5957,
"step": 4470
},
{
"epoch": 0.895,
"grad_norm": 76.61201477050781,
"learning_rate": 6.666998901648203e-06,
"loss": 250.1538,
"step": 4475
},
{
"epoch": 0.896,
"grad_norm": 61.74240493774414,
"learning_rate": 6.542246282281772e-06,
"loss": 230.3959,
"step": 4480
},
{
"epoch": 0.897,
"grad_norm": 62.60547637939453,
"learning_rate": 6.418632416150927e-06,
"loss": 228.3141,
"step": 4485
},
{
"epoch": 0.898,
"grad_norm": 52.61183547973633,
"learning_rate": 6.296158809451602e-06,
"loss": 235.3186,
"step": 4490
},
{
"epoch": 0.899,
"grad_norm": 82.08222961425781,
"learning_rate": 6.174826954486068e-06,
"loss": 218.2079,
"step": 4495
},
{
"epoch": 0.9,
"grad_norm": 52.42107391357422,
"learning_rate": 6.054638329644657e-06,
"loss": 226.4754,
"step": 4500
},
{
"epoch": 0.901,
"grad_norm": 67.4814682006836,
"learning_rate": 5.935594399387856e-06,
"loss": 234.8855,
"step": 4505
},
{
"epoch": 0.902,
"grad_norm": 67.26934814453125,
"learning_rate": 5.8176966142283965e-06,
"loss": 219.8597,
"step": 4510
},
{
"epoch": 0.903,
"grad_norm": 60.23142623901367,
"learning_rate": 5.700946410713548e-06,
"loss": 231.3427,
"step": 4515
},
{
"epoch": 0.904,
"grad_norm": 58.828922271728516,
"learning_rate": 5.585345211407733e-06,
"loss": 185.6134,
"step": 4520
},
{
"epoch": 0.905,
"grad_norm": 128.30523681640625,
"learning_rate": 5.470894424875062e-06,
"loss": 238.9794,
"step": 4525
},
{
"epoch": 0.906,
"grad_norm": 55.108795166015625,
"learning_rate": 5.357595445662267e-06,
"loss": 246.1672,
"step": 4530
},
{
"epoch": 0.907,
"grad_norm": 55.24848556518555,
"learning_rate": 5.245449654281631e-06,
"loss": 234.3741,
"step": 4535
},
{
"epoch": 0.908,
"grad_norm": 53.884220123291016,
"learning_rate": 5.134458417194254e-06,
"loss": 228.2846,
"step": 4540
},
{
"epoch": 0.909,
"grad_norm": 63.71116638183594,
"learning_rate": 5.024623086793323e-06,
"loss": 247.432,
"step": 4545
},
{
"epoch": 0.91,
"grad_norm": 59.29132843017578,
"learning_rate": 4.915945001387667e-06,
"loss": 231.0043,
"step": 4550
},
{
"epoch": 0.911,
"grad_norm": 61.93382263183594,
"learning_rate": 4.808425485185486e-06,
"loss": 239.5254,
"step": 4555
},
{
"epoch": 0.912,
"grad_norm": 66.62710571289062,
"learning_rate": 4.702065848278126e-06,
"loss": 237.1388,
"step": 4560
},
{
"epoch": 0.913,
"grad_norm": 75.43902587890625,
"learning_rate": 4.596867386624215e-06,
"loss": 228.6524,
"step": 4565
},
{
"epoch": 0.914,
"grad_norm": 58.848175048828125,
"learning_rate": 4.492831382033791e-06,
"loss": 229.1175,
"step": 4570
},
{
"epoch": 0.915,
"grad_norm": 56.25382995605469,
"learning_rate": 4.389959102152774e-06,
"loss": 246.6734,
"step": 4575
},
{
"epoch": 0.916,
"grad_norm": 60.79210662841797,
"learning_rate": 4.288251800447385e-06,
"loss": 217.5788,
"step": 4580
},
{
"epoch": 0.917,
"grad_norm": 63.43537902832031,
"learning_rate": 4.187710716189042e-06,
"loss": 234.6046,
"step": 4585
},
{
"epoch": 0.918,
"grad_norm": 51.950069427490234,
"learning_rate": 4.088337074439097e-06,
"loss": 228.8915,
"step": 4590
},
{
"epoch": 0.919,
"grad_norm": 57.15205383300781,
"learning_rate": 3.990132086034026e-06,
"loss": 233.0669,
"step": 4595
},
{
"epoch": 0.92,
"grad_norm": 71.64669799804688,
"learning_rate": 3.893096947570618e-06,
"loss": 233.7483,
"step": 4600
},
{
"epoch": 0.921,
"grad_norm": 73.67546081542969,
"learning_rate": 3.797232841391407e-06,
"loss": 230.8494,
"step": 4605
},
{
"epoch": 0.922,
"grad_norm": 55.27510070800781,
"learning_rate": 3.7025409355702976e-06,
"loss": 226.8931,
"step": 4610
},
{
"epoch": 0.923,
"grad_norm": 79.46499633789062,
"learning_rate": 3.6090223838982417e-06,
"loss": 231.0974,
"step": 4615
},
{
"epoch": 0.924,
"grad_norm": 57.17405700683594,
"learning_rate": 3.516678325869316e-06,
"loss": 240.2524,
"step": 4620
},
{
"epoch": 0.925,
"grad_norm": 54.9896240234375,
"learning_rate": 3.425509886666711e-06,
"loss": 235.579,
"step": 4625
},
{
"epoch": 0.926,
"grad_norm": 56.310646057128906,
"learning_rate": 3.3355181771490772e-06,
"loss": 238.8575,
"step": 4630
},
{
"epoch": 0.927,
"grad_norm": 54.55952835083008,
"learning_rate": 3.246704293837011e-06,
"loss": 224.249,
"step": 4635
},
{
"epoch": 0.928,
"grad_norm": 64.51632690429688,
"learning_rate": 3.1590693188996323e-06,
"loss": 203.2584,
"step": 4640
},
{
"epoch": 0.929,
"grad_norm": 58.42237091064453,
"learning_rate": 3.072614320141487e-06,
"loss": 228.6576,
"step": 4645
},
{
"epoch": 0.93,
"grad_norm": 61.63466262817383,
"learning_rate": 2.9873403509894203e-06,
"loss": 226.0415,
"step": 4650
},
{
"epoch": 0.931,
"grad_norm": 451.0778503417969,
"learning_rate": 2.9032484504798452e-06,
"loss": 222.5971,
"step": 4655
},
{
"epoch": 0.932,
"grad_norm": 55.56150436401367,
"learning_rate": 2.8203396432460506e-06,
"loss": 226.4209,
"step": 4660
},
{
"epoch": 0.933,
"grad_norm": 60.220184326171875,
"learning_rate": 2.738614939505646e-06,
"loss": 222.4756,
"step": 4665
},
{
"epoch": 0.934,
"grad_norm": 70.02185821533203,
"learning_rate": 2.6580753350484046e-06,
"loss": 218.7253,
"step": 4670
},
{
"epoch": 0.935,
"grad_norm": 55.44175720214844,
"learning_rate": 2.5787218112239496e-06,
"loss": 233.2217,
"step": 4675
},
{
"epoch": 0.936,
"grad_norm": 63.98387908935547,
"learning_rate": 2.5005553349299547e-06,
"loss": 235.4435,
"step": 4680
},
{
"epoch": 0.937,
"grad_norm": 57.520328521728516,
"learning_rate": 2.4235768586002517e-06,
"loss": 236.6912,
"step": 4685
},
{
"epoch": 0.938,
"grad_norm": 62.835716247558594,
"learning_rate": 2.3477873201932734e-06,
"loss": 236.4981,
"step": 4690
},
{
"epoch": 0.939,
"grad_norm": 58.75897979736328,
"learning_rate": 2.273187643180652e-06,
"loss": 235.2504,
"step": 4695
},
{
"epoch": 0.94,
"grad_norm": 61.33835220336914,
"learning_rate": 2.1997787365358958e-06,
"loss": 237.4497,
"step": 4700
},
{
"epoch": 0.941,
"grad_norm": 66.11475372314453,
"learning_rate": 2.1275614947233624e-06,
"loss": 234.8359,
"step": 4705
},
{
"epoch": 0.942,
"grad_norm": 57.40705871582031,
"learning_rate": 2.0565367976873584e-06,
"loss": 241.3148,
"step": 4710
},
{
"epoch": 0.943,
"grad_norm": 56.83675765991211,
"learning_rate": 1.986705510841402e-06,
"loss": 230.5031,
"step": 4715
},
{
"epoch": 0.944,
"grad_norm": 107.03742980957031,
"learning_rate": 1.918068485057689e-06,
"loss": 229.3515,
"step": 4720
},
{
"epoch": 0.945,
"grad_norm": 60.75922775268555,
"learning_rate": 1.8506265566567094e-06,
"loss": 228.9578,
"step": 4725
},
{
"epoch": 0.946,
"grad_norm": 84.7714614868164,
"learning_rate": 1.7843805473970798e-06,
"loss": 235.0035,
"step": 4730
},
{
"epoch": 0.947,
"grad_norm": 57.533287048339844,
"learning_rate": 1.719331264465529e-06,
"loss": 248.2758,
"step": 4735
},
{
"epoch": 0.948,
"grad_norm": 67.88099670410156,
"learning_rate": 1.6554795004670388e-06,
"loss": 237.8643,
"step": 4740
},
{
"epoch": 0.949,
"grad_norm": 61.89854431152344,
"learning_rate": 1.5928260334151845e-06,
"loss": 248.7147,
"step": 4745
},
{
"epoch": 0.95,
"grad_norm": 69.54090881347656,
"learning_rate": 1.5313716267226996e-06,
"loss": 232.0517,
"step": 4750
},
{
"epoch": 0.951,
"grad_norm": 55.21658706665039,
"learning_rate": 1.4711170291921484e-06,
"loss": 231.2092,
"step": 4755
},
{
"epoch": 0.952,
"grad_norm": 60.80591583251953,
"learning_rate": 1.412062975006767e-06,
"loss": 232.7167,
"step": 4760
},
{
"epoch": 0.953,
"grad_norm": 62.32032775878906,
"learning_rate": 1.3542101837215826e-06,
"loss": 243.0689,
"step": 4765
},
{
"epoch": 0.954,
"grad_norm": 73.29032135009766,
"learning_rate": 1.2975593602545965e-06,
"loss": 230.3041,
"step": 4770
},
{
"epoch": 0.955,
"grad_norm": 71.43152618408203,
"learning_rate": 1.2421111948782149e-06,
"loss": 239.0136,
"step": 4775
},
{
"epoch": 0.956,
"grad_norm": 53.471073150634766,
"learning_rate": 1.1878663632108322e-06,
"loss": 236.1855,
"step": 4780
},
{
"epoch": 0.957,
"grad_norm": 59.07845687866211,
"learning_rate": 1.134825526208605e-06,
"loss": 229.7462,
"step": 4785
},
{
"epoch": 0.958,
"grad_norm": 54.04808807373047,
"learning_rate": 1.0829893301573913e-06,
"loss": 225.9276,
"step": 4790
},
{
"epoch": 0.959,
"grad_norm": 63.05385971069336,
"learning_rate": 1.0323584066648795e-06,
"loss": 211.4304,
"step": 4795
},
{
"epoch": 0.96,
"grad_norm": 56.7470817565918,
"learning_rate": 9.829333726529056e-07,
"loss": 218.8612,
"step": 4800
},
{
"epoch": 0.961,
"grad_norm": 57.52992248535156,
"learning_rate": 9.347148303499142e-07,
"loss": 221.4465,
"step": 4805
},
{
"epoch": 0.962,
"grad_norm": 70.68370056152344,
"learning_rate": 8.877033672835988e-07,
"loss": 197.8579,
"step": 4810
},
{
"epoch": 0.963,
"grad_norm": 85.1485366821289,
"learning_rate": 8.418995562738285e-07,
"loss": 149.6401,
"step": 4815
},
{
"epoch": 0.964,
"grad_norm": 112.24050903320312,
"learning_rate": 7.973039554255768e-07,
"loss": 157.1447,
"step": 4820
},
{
"epoch": 0.965,
"grad_norm": 59.06969451904297,
"learning_rate": 7.539171081221596e-07,
"loss": 221.8144,
"step": 4825
},
{
"epoch": 0.966,
"grad_norm": 61.88059616088867,
"learning_rate": 7.117395430186414e-07,
"loss": 238.3345,
"step": 4830
},
{
"epoch": 0.967,
"grad_norm": 76.91974639892578,
"learning_rate": 6.707717740353059e-07,
"loss": 219.8431,
"step": 4835
},
{
"epoch": 0.968,
"grad_norm": 69.49444580078125,
"learning_rate": 6.310143003515179e-07,
"loss": 230.4514,
"step": 4840
},
{
"epoch": 0.969,
"grad_norm": 65.30635070800781,
"learning_rate": 5.924676063995382e-07,
"loss": 226.072,
"step": 4845
},
{
"epoch": 0.97,
"grad_norm": 65.86157989501953,
"learning_rate": 5.551321618586736e-07,
"loss": 233.9377,
"step": 4850
},
{
"epoch": 0.971,
"grad_norm": 57.465084075927734,
"learning_rate": 5.190084216495361e-07,
"loss": 230.2241,
"step": 4855
},
{
"epoch": 0.972,
"grad_norm": 57.69072723388672,
"learning_rate": 4.840968259284817e-07,
"loss": 210.2076,
"step": 4860
},
{
"epoch": 0.973,
"grad_norm": 59.57201385498047,
"learning_rate": 4.503978000823028e-07,
"loss": 235.0914,
"step": 4865
},
{
"epoch": 0.974,
"grad_norm": 68.75829315185547,
"learning_rate": 4.179117547229883e-07,
"loss": 226.2878,
"step": 4870
},
{
"epoch": 0.975,
"grad_norm": 60.78236389160156,
"learning_rate": 3.866390856827495e-07,
"loss": 225.6675,
"step": 4875
},
{
"epoch": 0.976,
"grad_norm": 58.79360580444336,
"learning_rate": 3.565801740092023e-07,
"loss": 223.148,
"step": 4880
},
{
"epoch": 0.977,
"grad_norm": 61.948543548583984,
"learning_rate": 3.277353859606813e-07,
"loss": 237.9903,
"step": 4885
},
{
"epoch": 0.978,
"grad_norm": 55.32079315185547,
"learning_rate": 3.001050730018218e-07,
"loss": 238.0726,
"step": 4890
},
{
"epoch": 0.979,
"grad_norm": 60.94178771972656,
"learning_rate": 2.73689571799296e-07,
"loss": 229.8969,
"step": 4895
},
{
"epoch": 0.98,
"grad_norm": 66.44815063476562,
"learning_rate": 2.484892042176279e-07,
"loss": 234.8636,
"step": 4900
},
{
"epoch": 0.981,
"grad_norm": 63.311954498291016,
"learning_rate": 2.2450427731534053e-07,
"loss": 244.0813,
"step": 4905
},
{
"epoch": 0.982,
"grad_norm": 61.87632369995117,
"learning_rate": 2.017350833412146e-07,
"loss": 254.649,
"step": 4910
},
{
"epoch": 0.983,
"grad_norm": 57.35133361816406,
"learning_rate": 1.8018189973069143e-07,
"loss": 233.5629,
"step": 4915
},
{
"epoch": 0.984,
"grad_norm": 59.498714447021484,
"learning_rate": 1.598449891024978e-07,
"loss": 239.7885,
"step": 4920
},
{
"epoch": 0.985,
"grad_norm": 59.527198791503906,
"learning_rate": 1.4072459925548177e-07,
"loss": 237.1914,
"step": 4925
},
{
"epoch": 0.986,
"grad_norm": 58.62847137451172,
"learning_rate": 1.2282096316554858e-07,
"loss": 240.0035,
"step": 4930
},
{
"epoch": 0.987,
"grad_norm": 58.177547454833984,
"learning_rate": 1.0613429898287398e-07,
"loss": 242.8458,
"step": 4935
},
{
"epoch": 0.988,
"grad_norm": 64.59917449951172,
"learning_rate": 9.066481002918403e-08,
"loss": 254.8642,
"step": 4940
},
{
"epoch": 0.989,
"grad_norm": 55.97276306152344,
"learning_rate": 7.641268479531283e-08,
"loss": 231.4949,
"step": 4945
},
{
"epoch": 0.99,
"grad_norm": 60.24343490600586,
"learning_rate": 6.337809693891527e-08,
"loss": 224.2755,
"step": 4950
},
{
"epoch": 0.991,
"grad_norm": 72.64891052246094,
"learning_rate": 5.1561205282335547e-08,
"loss": 221.7732,
"step": 4955
},
{
"epoch": 0.992,
"grad_norm": 59.544986724853516,
"learning_rate": 4.096215381066415e-08,
"loss": 202.4212,
"step": 4960
},
{
"epoch": 0.993,
"grad_norm": 60.605430603027344,
"learning_rate": 3.1581071670006015e-08,
"loss": 232.7862,
"step": 4965
},
{
"epoch": 0.994,
"grad_norm": 54.5761604309082,
"learning_rate": 2.3418073165870636e-08,
"loss": 224.3857,
"step": 4970
},
{
"epoch": 0.995,
"grad_norm": 58.323360443115234,
"learning_rate": 1.6473257761828732e-08,
"loss": 232.395,
"step": 4975
},
{
"epoch": 0.996,
"grad_norm": 86.96874237060547,
"learning_rate": 1.074671007825767e-08,
"loss": 217.5409,
"step": 4980
},
{
"epoch": 0.997,
"grad_norm": 74.31871032714844,
"learning_rate": 6.238499891353389e-09,
"loss": 233.2456,
"step": 4985
},
{
"epoch": 0.998,
"grad_norm": 99.23565673828125,
"learning_rate": 2.948682132208891e-09,
"loss": 207.9297,
"step": 4990
},
{
"epoch": 0.999,
"grad_norm": 91.24501037597656,
"learning_rate": 8.772968862369446e-10,
"loss": 204.9811,
"step": 4995
},
{
"epoch": 1.0,
"grad_norm": 59.23467254638672,
"learning_rate": 2.4369392592760164e-11,
"loss": 238.1345,
"step": 5000
}
],
"logging_steps": 5,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.4177763418112e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001,
"grad_norm": 133.26564025878906,
"learning_rate": 1.6000000000000001e-06,
"loss": 319.5714,
"step": 5
},
{
"epoch": 0.002,
"grad_norm": 116.60221099853516,
"learning_rate": 3.6e-06,
"loss": 321.6409,
"step": 10
},
{
"epoch": 0.003,
"grad_norm": 99.86441040039062,
"learning_rate": 5.600000000000001e-06,
"loss": 321.4713,
"step": 15
},
{
"epoch": 0.004,
"grad_norm": 120.4101333618164,
"learning_rate": 7.6e-06,
"loss": 326.6742,
"step": 20
},
{
"epoch": 0.005,
"grad_norm": 131.2949676513672,
"learning_rate": 9.600000000000001e-06,
"loss": 325.8993,
"step": 25
},
{
"epoch": 0.006,
"grad_norm": 109.72811889648438,
"learning_rate": 1.16e-05,
"loss": 313.0747,
"step": 30
},
{
"epoch": 0.007,
"grad_norm": 93.34612274169922,
"learning_rate": 1.3600000000000002e-05,
"loss": 315.1154,
"step": 35
},
{
"epoch": 0.008,
"grad_norm": 89.763427734375,
"learning_rate": 1.56e-05,
"loss": 307.5251,
"step": 40
},
{
"epoch": 0.009,
"grad_norm": 79.69110870361328,
"learning_rate": 1.76e-05,
"loss": 310.1657,
"step": 45
},
{
"epoch": 0.01,
"grad_norm": 83.45806121826172,
"learning_rate": 1.9600000000000002e-05,
"loss": 322.3456,
"step": 50
},
{
"epoch": 0.011,
"grad_norm": 83.1032943725586,
"learning_rate": 2.16e-05,
"loss": 307.5482,
"step": 55
},
{
"epoch": 0.012,
"grad_norm": 100.54994201660156,
"learning_rate": 2.36e-05,
"loss": 308.4983,
"step": 60
},
{
"epoch": 0.013,
"grad_norm": 74.47264099121094,
"learning_rate": 2.5600000000000002e-05,
"loss": 313.1681,