{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 6.461538461538462,
"eval_steps": 10,
"global_step": 2100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006153846153846154,
"eval_loss": 1.487821340560913,
"eval_runtime": 1.2374,
"eval_samples_per_second": 92.131,
"eval_steps_per_second": 4.849,
"step": 2
},
{
"epoch": 0.03076923076923077,
"grad_norm": 20.007999420166016,
"learning_rate": 6.153846153846155e-07,
"loss": 1.4703,
"step": 10
},
{
"epoch": 0.03076923076923077,
"eval_loss": 1.4221965074539185,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.431,
"eval_steps_per_second": 5.128,
"step": 10
},
{
"epoch": 0.06153846153846154,
"grad_norm": 29.657033920288086,
"learning_rate": 1.230769230769231e-06,
"loss": 1.4402,
"step": 20
},
{
"epoch": 0.06153846153846154,
"eval_loss": 1.238189935684204,
"eval_runtime": 1.1729,
"eval_samples_per_second": 97.195,
"eval_steps_per_second": 5.116,
"step": 20
},
{
"epoch": 0.09230769230769231,
"grad_norm": 10.80688190460205,
"learning_rate": 1.8461538461538465e-06,
"loss": 1.0957,
"step": 30
},
{
"epoch": 0.09230769230769231,
"eval_loss": 0.908875048160553,
"eval_runtime": 1.1739,
"eval_samples_per_second": 97.111,
"eval_steps_per_second": 5.111,
"step": 30
},
{
"epoch": 0.12307692307692308,
"grad_norm": 12.914667129516602,
"learning_rate": 2.461538461538462e-06,
"loss": 0.8136,
"step": 40
},
{
"epoch": 0.12307692307692308,
"eval_loss": 0.7436326146125793,
"eval_runtime": 1.1712,
"eval_samples_per_second": 97.34,
"eval_steps_per_second": 5.123,
"step": 40
},
{
"epoch": 0.15384615384615385,
"grad_norm": 5.206438064575195,
"learning_rate": 3.0769230769230774e-06,
"loss": 0.7864,
"step": 50
},
{
"epoch": 0.15384615384615385,
"eval_loss": 0.6924600601196289,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.287,
"eval_steps_per_second": 5.12,
"step": 50
},
{
"epoch": 0.18461538461538463,
"grad_norm": 5.670599460601807,
"learning_rate": 3.692307692307693e-06,
"loss": 0.622,
"step": 60
},
{
"epoch": 0.18461538461538463,
"eval_loss": 0.6791077852249146,
"eval_runtime": 1.1678,
"eval_samples_per_second": 97.62,
"eval_steps_per_second": 5.138,
"step": 60
},
{
"epoch": 0.2153846153846154,
"grad_norm": 3.6266868114471436,
"learning_rate": 4.307692307692308e-06,
"loss": 0.6672,
"step": 70
},
{
"epoch": 0.2153846153846154,
"eval_loss": 0.6631948947906494,
"eval_runtime": 1.1626,
"eval_samples_per_second": 98.06,
"eval_steps_per_second": 5.161,
"step": 70
},
{
"epoch": 0.24615384615384617,
"grad_norm": 3.1027884483337402,
"learning_rate": 4.923076923076924e-06,
"loss": 0.6097,
"step": 80
},
{
"epoch": 0.24615384615384617,
"eval_loss": 0.6625475883483887,
"eval_runtime": 1.1717,
"eval_samples_per_second": 97.294,
"eval_steps_per_second": 5.121,
"step": 80
},
{
"epoch": 0.27692307692307694,
"grad_norm": 3.233839511871338,
"learning_rate": 5.538461538461539e-06,
"loss": 0.6957,
"step": 90
},
{
"epoch": 0.27692307692307694,
"eval_loss": 0.6590337157249451,
"eval_runtime": 1.1664,
"eval_samples_per_second": 97.738,
"eval_steps_per_second": 5.144,
"step": 90
},
{
"epoch": 0.3076923076923077,
"grad_norm": 4.932614803314209,
"learning_rate": 6.153846153846155e-06,
"loss": 0.6994,
"step": 100
},
{
"epoch": 0.3076923076923077,
"eval_loss": 0.6592673659324646,
"eval_runtime": 1.1692,
"eval_samples_per_second": 97.502,
"eval_steps_per_second": 5.132,
"step": 100
},
{
"epoch": 0.3384615384615385,
"grad_norm": 2.642303466796875,
"learning_rate": 6.76923076923077e-06,
"loss": 0.6501,
"step": 110
},
{
"epoch": 0.3384615384615385,
"eval_loss": 0.6589992642402649,
"eval_runtime": 1.1669,
"eval_samples_per_second": 97.691,
"eval_steps_per_second": 5.142,
"step": 110
},
{
"epoch": 0.36923076923076925,
"grad_norm": 3.45409893989563,
"learning_rate": 7.384615384615386e-06,
"loss": 0.7512,
"step": 120
},
{
"epoch": 0.36923076923076925,
"eval_loss": 0.6568591594696045,
"eval_runtime": 1.1778,
"eval_samples_per_second": 96.79,
"eval_steps_per_second": 5.094,
"step": 120
},
{
"epoch": 0.4,
"grad_norm": 8.460351943969727,
"learning_rate": 8.000000000000001e-06,
"loss": 0.5996,
"step": 130
},
{
"epoch": 0.4,
"eval_loss": 0.6578297019004822,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.575,
"eval_steps_per_second": 5.136,
"step": 130
},
{
"epoch": 0.4307692307692308,
"grad_norm": 2.4495465755462646,
"learning_rate": 8.615384615384617e-06,
"loss": 0.6498,
"step": 140
},
{
"epoch": 0.4307692307692308,
"eval_loss": 0.6588199138641357,
"eval_runtime": 1.1664,
"eval_samples_per_second": 97.739,
"eval_steps_per_second": 5.144,
"step": 140
},
{
"epoch": 0.46153846153846156,
"grad_norm": 2.6779162883758545,
"learning_rate": 9.230769230769232e-06,
"loss": 0.703,
"step": 150
},
{
"epoch": 0.46153846153846156,
"eval_loss": 0.6626676321029663,
"eval_runtime": 1.1764,
"eval_samples_per_second": 96.904,
"eval_steps_per_second": 5.1,
"step": 150
},
{
"epoch": 0.49230769230769234,
"grad_norm": 8.147880554199219,
"learning_rate": 9.846153846153848e-06,
"loss": 0.6649,
"step": 160
},
{
"epoch": 0.49230769230769234,
"eval_loss": 0.6644802689552307,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.424,
"eval_steps_per_second": 5.128,
"step": 160
},
{
"epoch": 0.5230769230769231,
"grad_norm": 2.7410728931427,
"learning_rate": 1.0461538461538463e-05,
"loss": 0.6704,
"step": 170
},
{
"epoch": 0.5230769230769231,
"eval_loss": 0.6591455340385437,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.253,
"eval_steps_per_second": 5.119,
"step": 170
},
{
"epoch": 0.5538461538461539,
"grad_norm": 6.886598587036133,
"learning_rate": 1.1076923076923079e-05,
"loss": 0.6073,
"step": 180
},
{
"epoch": 0.5538461538461539,
"eval_loss": 0.6617754697799683,
"eval_runtime": 1.1682,
"eval_samples_per_second": 97.588,
"eval_steps_per_second": 5.136,
"step": 180
},
{
"epoch": 0.5846153846153846,
"grad_norm": 4.26249361038208,
"learning_rate": 1.1692307692307694e-05,
"loss": 0.7556,
"step": 190
},
{
"epoch": 0.5846153846153846,
"eval_loss": 0.6659030318260193,
"eval_runtime": 1.1721,
"eval_samples_per_second": 97.265,
"eval_steps_per_second": 5.119,
"step": 190
},
{
"epoch": 0.6153846153846154,
"grad_norm": 4.204222202301025,
"learning_rate": 1.230769230769231e-05,
"loss": 0.6583,
"step": 200
},
{
"epoch": 0.6153846153846154,
"eval_loss": 0.6660778522491455,
"eval_runtime": 1.1728,
"eval_samples_per_second": 97.201,
"eval_steps_per_second": 5.116,
"step": 200
},
{
"epoch": 0.6461538461538462,
"grad_norm": 3.154658079147339,
"learning_rate": 1.2923076923076925e-05,
"loss": 0.7288,
"step": 210
},
{
"epoch": 0.6461538461538462,
"eval_loss": 0.6642422676086426,
"eval_runtime": 1.166,
"eval_samples_per_second": 97.772,
"eval_steps_per_second": 5.146,
"step": 210
},
{
"epoch": 0.676923076923077,
"grad_norm": 3.248359203338623,
"learning_rate": 1.353846153846154e-05,
"loss": 0.6702,
"step": 220
},
{
"epoch": 0.676923076923077,
"eval_loss": 0.6680281758308411,
"eval_runtime": 1.1691,
"eval_samples_per_second": 97.507,
"eval_steps_per_second": 5.132,
"step": 220
},
{
"epoch": 0.7076923076923077,
"grad_norm": 2.5507216453552246,
"learning_rate": 1.4153846153846156e-05,
"loss": 0.66,
"step": 230
},
{
"epoch": 0.7076923076923077,
"eval_loss": 0.6713310480117798,
"eval_runtime": 1.167,
"eval_samples_per_second": 97.683,
"eval_steps_per_second": 5.141,
"step": 230
},
{
"epoch": 0.7384615384615385,
"grad_norm": 2.1180007457733154,
"learning_rate": 1.4769230769230772e-05,
"loss": 0.6172,
"step": 240
},
{
"epoch": 0.7384615384615385,
"eval_loss": 0.6759226322174072,
"eval_runtime": 1.1758,
"eval_samples_per_second": 96.954,
"eval_steps_per_second": 5.103,
"step": 240
},
{
"epoch": 0.7692307692307693,
"grad_norm": 2.8318371772766113,
"learning_rate": 1.5384615384615387e-05,
"loss": 0.5954,
"step": 250
},
{
"epoch": 0.7692307692307693,
"eval_loss": 0.6854314208030701,
"eval_runtime": 1.1698,
"eval_samples_per_second": 97.457,
"eval_steps_per_second": 5.129,
"step": 250
},
{
"epoch": 0.8,
"grad_norm": 3.519829034805298,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.6639,
"step": 260
},
{
"epoch": 0.8,
"eval_loss": 0.6849851012229919,
"eval_runtime": 1.1732,
"eval_samples_per_second": 97.173,
"eval_steps_per_second": 5.114,
"step": 260
},
{
"epoch": 0.8307692307692308,
"grad_norm": 3.7909388542175293,
"learning_rate": 1.6615384615384618e-05,
"loss": 0.6892,
"step": 270
},
{
"epoch": 0.8307692307692308,
"eval_loss": 0.6863855123519897,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.53,
"eval_steps_per_second": 5.133,
"step": 270
},
{
"epoch": 0.8615384615384616,
"grad_norm": 8.278326988220215,
"learning_rate": 1.7230769230769234e-05,
"loss": 0.6014,
"step": 280
},
{
"epoch": 0.8615384615384616,
"eval_loss": 0.6890373826026917,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.251,
"eval_steps_per_second": 5.118,
"step": 280
},
{
"epoch": 0.8923076923076924,
"grad_norm": 2.616001605987549,
"learning_rate": 1.784615384615385e-05,
"loss": 0.6775,
"step": 290
},
{
"epoch": 0.8923076923076924,
"eval_loss": 0.6797110438346863,
"eval_runtime": 1.1888,
"eval_samples_per_second": 95.897,
"eval_steps_per_second": 5.047,
"step": 290
},
{
"epoch": 0.9230769230769231,
"grad_norm": 2.4450528621673584,
"learning_rate": 1.8461538461538465e-05,
"loss": 0.7434,
"step": 300
},
{
"epoch": 0.9230769230769231,
"eval_loss": 0.6779448986053467,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.254,
"eval_steps_per_second": 5.119,
"step": 300
},
{
"epoch": 0.9538461538461539,
"grad_norm": 4.121164798736572,
"learning_rate": 1.907692307692308e-05,
"loss": 0.6413,
"step": 310
},
{
"epoch": 0.9538461538461539,
"eval_loss": 0.6819297671318054,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.574,
"eval_steps_per_second": 5.135,
"step": 310
},
{
"epoch": 0.9846153846153847,
"grad_norm": 2.057229995727539,
"learning_rate": 1.9692307692307696e-05,
"loss": 0.6511,
"step": 320
},
{
"epoch": 0.9846153846153847,
"eval_loss": 0.6861481666564941,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.444,
"eval_steps_per_second": 5.129,
"step": 320
},
{
"epoch": 1.0153846153846153,
"grad_norm": 1.7295210361480713,
"learning_rate": 1.9999855802751384e-05,
"loss": 0.5872,
"step": 330
},
{
"epoch": 1.0153846153846153,
"eval_loss": 0.7481781244277954,
"eval_runtime": 1.172,
"eval_samples_per_second": 97.273,
"eval_steps_per_second": 5.12,
"step": 330
},
{
"epoch": 1.0461538461538462,
"grad_norm": 3.6910336017608643,
"learning_rate": 1.9998702249713747e-05,
"loss": 0.438,
"step": 340
},
{
"epoch": 1.0461538461538462,
"eval_loss": 0.7471169829368591,
"eval_runtime": 1.175,
"eval_samples_per_second": 97.019,
"eval_steps_per_second": 5.106,
"step": 340
},
{
"epoch": 1.0769230769230769,
"grad_norm": 3.9220831394195557,
"learning_rate": 1.9996395276708856e-05,
"loss": 0.4141,
"step": 350
},
{
"epoch": 1.0769230769230769,
"eval_loss": 0.7425897121429443,
"eval_runtime": 1.1888,
"eval_samples_per_second": 95.893,
"eval_steps_per_second": 5.047,
"step": 350
},
{
"epoch": 1.1076923076923078,
"grad_norm": 2.1136879920959473,
"learning_rate": 1.9992935149862116e-05,
"loss": 0.4173,
"step": 360
},
{
"epoch": 1.1076923076923078,
"eval_loss": 0.7477542161941528,
"eval_runtime": 1.1705,
"eval_samples_per_second": 97.398,
"eval_steps_per_second": 5.126,
"step": 360
},
{
"epoch": 1.1384615384615384,
"grad_norm": 2.187299966812134,
"learning_rate": 1.998832226832327e-05,
"loss": 0.4242,
"step": 370
},
{
"epoch": 1.1384615384615384,
"eval_loss": 0.7475994229316711,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.582,
"eval_steps_per_second": 5.136,
"step": 370
},
{
"epoch": 1.1692307692307693,
"grad_norm": 2.683750629425049,
"learning_rate": 1.9982557164220335e-05,
"loss": 0.4413,
"step": 380
},
{
"epoch": 1.1692307692307693,
"eval_loss": 0.7552060484886169,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.529,
"eval_steps_per_second": 5.133,
"step": 380
},
{
"epoch": 1.2,
"grad_norm": 3.312657117843628,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.3361,
"step": 390
},
{
"epoch": 1.2,
"eval_loss": 0.7630029916763306,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.529,
"eval_steps_per_second": 5.133,
"step": 390
},
{
"epoch": 1.2307692307692308,
"grad_norm": 2.3290536403656006,
"learning_rate": 1.9967573081342103e-05,
"loss": 0.4174,
"step": 400
},
{
"epoch": 1.2307692307692308,
"eval_loss": 0.7669293880462646,
"eval_runtime": 1.1711,
"eval_samples_per_second": 97.345,
"eval_steps_per_second": 5.123,
"step": 400
},
{
"epoch": 1.2615384615384615,
"grad_norm": 2.4483697414398193,
"learning_rate": 1.9958355831085155e-05,
"loss": 0.4045,
"step": 410
},
{
"epoch": 1.2615384615384615,
"eval_loss": 0.7652024626731873,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.461,
"eval_steps_per_second": 5.13,
"step": 410
},
{
"epoch": 1.2923076923076924,
"grad_norm": 1.703616976737976,
"learning_rate": 1.9947989815101444e-05,
"loss": 0.4292,
"step": 420
},
{
"epoch": 1.2923076923076924,
"eval_loss": 0.7553891539573669,
"eval_runtime": 1.1719,
"eval_samples_per_second": 97.277,
"eval_steps_per_second": 5.12,
"step": 420
},
{
"epoch": 1.323076923076923,
"grad_norm": 1.6245399713516235,
"learning_rate": 1.9936476229183133e-05,
"loss": 0.4164,
"step": 430
},
{
"epoch": 1.323076923076923,
"eval_loss": 0.7711736559867859,
"eval_runtime": 1.1776,
"eval_samples_per_second": 96.806,
"eval_steps_per_second": 5.095,
"step": 430
},
{
"epoch": 1.353846153846154,
"grad_norm": 1.9511611461639404,
"learning_rate": 1.992381640150257e-05,
"loss": 0.4337,
"step": 440
},
{
"epoch": 1.353846153846154,
"eval_loss": 0.7438372373580933,
"eval_runtime": 1.1692,
"eval_samples_per_second": 97.5,
"eval_steps_per_second": 5.132,
"step": 440
},
{
"epoch": 1.3846153846153846,
"grad_norm": 2.466737747192383,
"learning_rate": 1.9910011792459086e-05,
"loss": 0.3712,
"step": 450
},
{
"epoch": 1.3846153846153846,
"eval_loss": 0.8045614361763,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.53,
"eval_steps_per_second": 5.133,
"step": 450
},
{
"epoch": 1.4153846153846155,
"grad_norm": 2.473851203918457,
"learning_rate": 1.9895063994510512e-05,
"loss": 0.516,
"step": 460
},
{
"epoch": 1.4153846153846155,
"eval_loss": 0.7463352084159851,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.464,
"eval_steps_per_second": 5.13,
"step": 460
},
{
"epoch": 1.4461538461538461,
"grad_norm": 2.757955312728882,
"learning_rate": 1.9878974731989487e-05,
"loss": 0.4965,
"step": 470
},
{
"epoch": 1.4461538461538461,
"eval_loss": 0.7571713924407959,
"eval_runtime": 1.1688,
"eval_samples_per_second": 97.534,
"eval_steps_per_second": 5.133,
"step": 470
},
{
"epoch": 1.476923076923077,
"grad_norm": 3.9884915351867676,
"learning_rate": 1.9861745860904538e-05,
"loss": 0.3973,
"step": 480
},
{
"epoch": 1.476923076923077,
"eval_loss": 0.7763081789016724,
"eval_runtime": 1.168,
"eval_samples_per_second": 97.605,
"eval_steps_per_second": 5.137,
"step": 480
},
{
"epoch": 1.5076923076923077,
"grad_norm": 1.9474631547927856,
"learning_rate": 1.9843379368725978e-05,
"loss": 0.4625,
"step": 490
},
{
"epoch": 1.5076923076923077,
"eval_loss": 0.7595367431640625,
"eval_runtime": 1.1708,
"eval_samples_per_second": 97.37,
"eval_steps_per_second": 5.125,
"step": 490
},
{
"epoch": 1.5384615384615383,
"grad_norm": 2.2907869815826416,
"learning_rate": 1.9823877374156647e-05,
"loss": 0.4371,
"step": 500
},
{
"epoch": 1.5384615384615383,
"eval_loss": 0.762865424156189,
"eval_runtime": 1.1766,
"eval_samples_per_second": 96.891,
"eval_steps_per_second": 5.1,
"step": 500
},
{
"epoch": 1.5692307692307692,
"grad_norm": 3.067608594894409,
"learning_rate": 1.9803242126887496e-05,
"loss": 0.45,
"step": 510
},
{
"epoch": 1.5692307692307692,
"eval_loss": 0.7638718485832214,
"eval_runtime": 1.1684,
"eval_samples_per_second": 97.568,
"eval_steps_per_second": 5.135,
"step": 510
},
{
"epoch": 1.6,
"grad_norm": 2.1113197803497314,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.4524,
"step": 520
},
{
"epoch": 1.6,
"eval_loss": 0.7560412883758545,
"eval_runtime": 1.1719,
"eval_samples_per_second": 97.278,
"eval_steps_per_second": 5.12,
"step": 520
},
{
"epoch": 1.6307692307692307,
"grad_norm": 1.889585256576538,
"learning_rate": 1.9758581526381878e-05,
"loss": 0.4287,
"step": 530
},
{
"epoch": 1.6307692307692307,
"eval_loss": 0.7591350674629211,
"eval_runtime": 1.1693,
"eval_samples_per_second": 97.495,
"eval_steps_per_second": 5.131,
"step": 530
},
{
"epoch": 1.6615384615384614,
"grad_norm": 2.519418716430664,
"learning_rate": 1.973456132505684e-05,
"loss": 0.4603,
"step": 540
},
{
"epoch": 1.6615384615384614,
"eval_loss": 0.7536805868148804,
"eval_runtime": 1.1799,
"eval_samples_per_second": 96.617,
"eval_steps_per_second": 5.085,
"step": 540
},
{
"epoch": 1.6923076923076923,
"grad_norm": 2.7455596923828125,
"learning_rate": 1.9709418174260523e-05,
"loss": 0.4293,
"step": 550
},
{
"epoch": 1.6923076923076923,
"eval_loss": 0.763169527053833,
"eval_runtime": 1.1696,
"eval_samples_per_second": 97.47,
"eval_steps_per_second": 5.13,
"step": 550
},
{
"epoch": 1.7230769230769232,
"grad_norm": 1.9823411703109741,
"learning_rate": 1.9683154974430544e-05,
"loss": 0.4643,
"step": 560
},
{
"epoch": 1.7230769230769232,
"eval_loss": 0.7607314586639404,
"eval_runtime": 1.1712,
"eval_samples_per_second": 97.336,
"eval_steps_per_second": 5.123,
"step": 560
},
{
"epoch": 1.7538461538461538,
"grad_norm": 2.2437684535980225,
"learning_rate": 1.965577475520999e-05,
"loss": 0.452,
"step": 570
},
{
"epoch": 1.7538461538461538,
"eval_loss": 0.7615717649459839,
"eval_runtime": 1.1946,
"eval_samples_per_second": 95.425,
"eval_steps_per_second": 5.022,
"step": 570
},
{
"epoch": 1.7846153846153845,
"grad_norm": 3.0740182399749756,
"learning_rate": 1.962728067509791e-05,
"loss": 0.4885,
"step": 580
},
{
"epoch": 1.7846153846153845,
"eval_loss": 0.7764689326286316,
"eval_runtime": 1.1714,
"eval_samples_per_second": 97.32,
"eval_steps_per_second": 5.122,
"step": 580
},
{
"epoch": 1.8153846153846154,
"grad_norm": 2.050313711166382,
"learning_rate": 1.9597676021084962e-05,
"loss": 0.4992,
"step": 590
},
{
"epoch": 1.8153846153846154,
"eval_loss": 0.7457339763641357,
"eval_runtime": 1.1747,
"eval_samples_per_second": 97.045,
"eval_steps_per_second": 5.108,
"step": 590
},
{
"epoch": 1.8461538461538463,
"grad_norm": 2.349470853805542,
"learning_rate": 1.9566964208274254e-05,
"loss": 0.3963,
"step": 600
},
{
"epoch": 1.8461538461538463,
"eval_loss": 0.7712545990943909,
"eval_runtime": 1.17,
"eval_samples_per_second": 97.433,
"eval_steps_per_second": 5.128,
"step": 600
},
{
"epoch": 1.876923076923077,
"grad_norm": 1.947912573814392,
"learning_rate": 1.9535148779487365e-05,
"loss": 0.4388,
"step": 610
},
{
"epoch": 1.876923076923077,
"eval_loss": 0.772519052028656,
"eval_runtime": 1.1717,
"eval_samples_per_second": 97.296,
"eval_steps_per_second": 5.121,
"step": 610
},
{
"epoch": 1.9076923076923076,
"grad_norm": 1.8837225437164307,
"learning_rate": 1.9502233404855672e-05,
"loss": 0.3965,
"step": 620
},
{
"epoch": 1.9076923076923076,
"eval_loss": 0.7594517469406128,
"eval_runtime": 1.2215,
"eval_samples_per_second": 93.326,
"eval_steps_per_second": 4.912,
"step": 620
},
{
"epoch": 1.9384615384615385,
"grad_norm": 1.8398792743682861,
"learning_rate": 1.946822188139696e-05,
"loss": 0.3898,
"step": 630
},
{
"epoch": 1.9384615384615385,
"eval_loss": 0.7748451828956604,
"eval_runtime": 1.2222,
"eval_samples_per_second": 93.273,
"eval_steps_per_second": 4.909,
"step": 630
},
{
"epoch": 1.9692307692307693,
"grad_norm": 2.3971219062805176,
"learning_rate": 1.9433118132577432e-05,
"loss": 0.4253,
"step": 640
},
{
"epoch": 1.9692307692307693,
"eval_loss": 0.7729252576828003,
"eval_runtime": 1.1743,
"eval_samples_per_second": 97.081,
"eval_steps_per_second": 5.11,
"step": 640
},
{
"epoch": 2.0,
"grad_norm": 1.657328724861145,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.4898,
"step": 650
},
{
"epoch": 2.0,
"eval_loss": 0.7602188587188721,
"eval_runtime": 1.1738,
"eval_samples_per_second": 97.117,
"eval_steps_per_second": 5.111,
"step": 650
},
{
"epoch": 2.0307692307692307,
"grad_norm": 1.3759773969650269,
"learning_rate": 1.935965028223259e-05,
"loss": 0.2182,
"step": 660
},
{
"epoch": 2.0307692307692307,
"eval_loss": 0.8499692678451538,
"eval_runtime": 1.1719,
"eval_samples_per_second": 97.276,
"eval_steps_per_second": 5.12,
"step": 660
},
{
"epoch": 2.0615384615384613,
"grad_norm": 1.4757074117660522,
"learning_rate": 1.932129465573568e-05,
"loss": 0.2196,
"step": 670
},
{
"epoch": 2.0615384615384613,
"eval_loss": 0.866599977016449,
"eval_runtime": 1.1739,
"eval_samples_per_second": 97.115,
"eval_steps_per_second": 5.111,
"step": 670
},
{
"epoch": 2.0923076923076924,
"grad_norm": 1.2500200271606445,
"learning_rate": 1.9281863752957095e-05,
"loss": 0.1772,
"step": 680
},
{
"epoch": 2.0923076923076924,
"eval_loss": 0.8642630577087402,
"eval_runtime": 1.1753,
"eval_samples_per_second": 96.999,
"eval_steps_per_second": 5.105,
"step": 680
},
{
"epoch": 2.123076923076923,
"grad_norm": 2.561828851699829,
"learning_rate": 1.92413621225262e-05,
"loss": 0.2014,
"step": 690
},
{
"epoch": 2.123076923076923,
"eval_loss": 0.8872668743133545,
"eval_runtime": 1.1761,
"eval_samples_per_second": 96.933,
"eval_steps_per_second": 5.102,
"step": 690
},
{
"epoch": 2.1538461538461537,
"grad_norm": 1.8431521654129028,
"learning_rate": 1.9199794436588244e-05,
"loss": 0.2245,
"step": 700
},
{
"epoch": 2.1538461538461537,
"eval_loss": 0.8799911737442017,
"eval_runtime": 1.1766,
"eval_samples_per_second": 96.89,
"eval_steps_per_second": 5.099,
"step": 700
},
{
"epoch": 2.184615384615385,
"grad_norm": 1.4718531370162964,
"learning_rate": 1.915716549026541e-05,
"loss": 0.1911,
"step": 710
},
{
"epoch": 2.184615384615385,
"eval_loss": 0.8806982636451721,
"eval_runtime": 1.1696,
"eval_samples_per_second": 97.473,
"eval_steps_per_second": 5.13,
"step": 710
},
{
"epoch": 2.2153846153846155,
"grad_norm": 2.061520576477051,
"learning_rate": 1.9113480201103658e-05,
"loss": 0.232,
"step": 720
},
{
"epoch": 2.2153846153846155,
"eval_loss": 0.8812103271484375,
"eval_runtime": 1.1676,
"eval_samples_per_second": 97.636,
"eval_steps_per_second": 5.139,
"step": 720
},
{
"epoch": 2.246153846153846,
"grad_norm": 2.530285596847534,
"learning_rate": 1.9068743608505454e-05,
"loss": 0.2232,
"step": 730
},
{
"epoch": 2.246153846153846,
"eval_loss": 0.8682798147201538,
"eval_runtime": 1.1687,
"eval_samples_per_second": 97.545,
"eval_steps_per_second": 5.134,
"step": 730
},
{
"epoch": 2.276923076923077,
"grad_norm": 1.8433207273483276,
"learning_rate": 1.902296087314845e-05,
"loss": 0.2315,
"step": 740
},
{
"epoch": 2.276923076923077,
"eval_loss": 0.8747490048408508,
"eval_runtime": 1.1702,
"eval_samples_per_second": 97.421,
"eval_steps_per_second": 5.127,
"step": 740
},
{
"epoch": 2.3076923076923075,
"grad_norm": 2.1529664993286133,
"learning_rate": 1.8976137276390145e-05,
"loss": 0.239,
"step": 750
},
{
"epoch": 2.3076923076923075,
"eval_loss": 0.8778530359268188,
"eval_runtime": 1.1674,
"eval_samples_per_second": 97.656,
"eval_steps_per_second": 5.14,
"step": 750
},
{
"epoch": 2.3384615384615386,
"grad_norm": 1.839979887008667,
"learning_rate": 1.892827821965864e-05,
"loss": 0.2228,
"step": 760
},
{
"epoch": 2.3384615384615386,
"eval_loss": 0.8769873976707458,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.288,
"eval_steps_per_second": 5.12,
"step": 760
},
{
"epoch": 2.3692307692307693,
"grad_norm": 1.3814674615859985,
"learning_rate": 1.8879389223829592e-05,
"loss": 0.1999,
"step": 770
},
{
"epoch": 2.3692307692307693,
"eval_loss": 0.8731983304023743,
"eval_runtime": 1.1723,
"eval_samples_per_second": 97.242,
"eval_steps_per_second": 5.118,
"step": 770
},
{
"epoch": 2.4,
"grad_norm": 1.5382564067840576,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.1888,
"step": 780
},
{
"epoch": 2.4,
"eval_loss": 0.8794592022895813,
"eval_runtime": 1.1678,
"eval_samples_per_second": 97.622,
"eval_steps_per_second": 5.138,
"step": 780
},
{
"epoch": 2.430769230769231,
"grad_norm": 2.371277093887329,
"learning_rate": 1.8778544091784047e-05,
"loss": 0.2293,
"step": 790
},
{
"epoch": 2.430769230769231,
"eval_loss": 0.8893064260482788,
"eval_runtime": 1.1684,
"eval_samples_per_second": 97.567,
"eval_steps_per_second": 5.135,
"step": 790
},
{
"epoch": 2.4615384615384617,
"grad_norm": 1.4892629384994507,
"learning_rate": 1.8726599588756144e-05,
"loss": 0.2326,
"step": 800
},
{
"epoch": 2.4615384615384617,
"eval_loss": 0.8676194548606873,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.441,
"eval_steps_per_second": 5.128,
"step": 800
},
{
"epoch": 2.4923076923076923,
"grad_norm": 1.6047152280807495,
"learning_rate": 1.8673648411665895e-05,
"loss": 0.1986,
"step": 810
},
{
"epoch": 2.4923076923076923,
"eval_loss": 0.8942193984985352,
"eval_runtime": 1.17,
"eval_samples_per_second": 97.433,
"eval_steps_per_second": 5.128,
"step": 810
},
{
"epoch": 2.523076923076923,
"grad_norm": 2.106461524963379,
"learning_rate": 1.8619696668800494e-05,
"loss": 0.238,
"step": 820
},
{
"epoch": 2.523076923076923,
"eval_loss": 0.8907297849655151,
"eval_runtime": 1.1681,
"eval_samples_per_second": 97.593,
"eval_steps_per_second": 5.136,
"step": 820
},
{
"epoch": 2.5538461538461537,
"grad_norm": 1.7706941366195679,
"learning_rate": 1.8564750583869374e-05,
"loss": 0.2278,
"step": 830
},
{
"epoch": 2.5538461538461537,
"eval_loss": 0.8588199019432068,
"eval_runtime": 1.1709,
"eval_samples_per_second": 97.36,
"eval_steps_per_second": 5.124,
"step": 830
},
{
"epoch": 2.5846153846153848,
"grad_norm": 1.4446183443069458,
"learning_rate": 1.850881649528625e-05,
"loss": 0.2033,
"step": 840
},
{
"epoch": 2.5846153846153848,
"eval_loss": 0.8720155954360962,
"eval_runtime": 1.1716,
"eval_samples_per_second": 97.304,
"eval_steps_per_second": 5.121,
"step": 840
},
{
"epoch": 2.6153846153846154,
"grad_norm": 1.5155885219573975,
"learning_rate": 1.845190085543795e-05,
"loss": 0.2211,
"step": 850
},
{
"epoch": 2.6153846153846154,
"eval_loss": 0.8732415437698364,
"eval_runtime": 1.1708,
"eval_samples_per_second": 97.371,
"eval_steps_per_second": 5.125,
"step": 850
},
{
"epoch": 2.646153846153846,
"grad_norm": 2.934957981109619,
"learning_rate": 1.839401022994006e-05,
"loss": 0.21,
"step": 860
},
{
"epoch": 2.646153846153846,
"eval_loss": 0.8790932297706604,
"eval_runtime": 1.1671,
"eval_samples_per_second": 97.68,
"eval_steps_per_second": 5.141,
"step": 860
},
{
"epoch": 2.676923076923077,
"grad_norm": 2.930457353591919,
"learning_rate": 1.8335151296879576e-05,
"loss": 0.2313,
"step": 870
},
{
"epoch": 2.676923076923077,
"eval_loss": 0.8555310368537903,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.284,
"eval_steps_per_second": 5.12,
"step": 870
},
{
"epoch": 2.707692307692308,
"grad_norm": 2.029303789138794,
"learning_rate": 1.82753308460445e-05,
"loss": 0.2285,
"step": 880
},
{
"epoch": 2.707692307692308,
"eval_loss": 0.870873749256134,
"eval_runtime": 1.1696,
"eval_samples_per_second": 97.47,
"eval_steps_per_second": 5.13,
"step": 880
},
{
"epoch": 2.7384615384615385,
"grad_norm": 1.949743390083313,
"learning_rate": 1.821455577814062e-05,
"loss": 0.2134,
"step": 890
},
{
"epoch": 2.7384615384615385,
"eval_loss": 0.8873838186264038,
"eval_runtime": 1.1688,
"eval_samples_per_second": 97.538,
"eval_steps_per_second": 5.134,
"step": 890
},
{
"epoch": 2.769230769230769,
"grad_norm": 1.6945637464523315,
"learning_rate": 1.8152833103995443e-05,
"loss": 0.218,
"step": 900
},
{
"epoch": 2.769230769230769,
"eval_loss": 0.8798539042472839,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.444,
"eval_steps_per_second": 5.129,
"step": 900
},
{
"epoch": 2.8,
"grad_norm": 2.2784035205841064,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.2426,
"step": 910
},
{
"epoch": 2.8,
"eval_loss": 0.8753871321678162,
"eval_runtime": 1.1723,
"eval_samples_per_second": 97.246,
"eval_steps_per_second": 5.118,
"step": 910
},
{
"epoch": 2.830769230769231,
"grad_norm": 1.750718593597412,
"learning_rate": 1.802657352603483e-05,
"loss": 0.2059,
"step": 920
},
{
"epoch": 2.830769230769231,
"eval_loss": 0.8761821985244751,
"eval_runtime": 1.1762,
"eval_samples_per_second": 96.92,
"eval_steps_per_second": 5.101,
"step": 920
},
{
"epoch": 2.8615384615384616,
"grad_norm": 1.2204371690750122,
"learning_rate": 1.7962051187141377e-05,
"loss": 0.2229,
"step": 930
},
{
"epoch": 2.8615384615384616,
"eval_loss": 0.8772630095481873,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.458,
"eval_steps_per_second": 5.129,
"step": 930
},
{
"epoch": 2.8923076923076922,
"grad_norm": 2.0748281478881836,
"learning_rate": 1.7896610370170452e-05,
"loss": 0.2582,
"step": 940
},
{
"epoch": 2.8923076923076922,
"eval_loss": 0.884286105632782,
"eval_runtime": 1.1773,
"eval_samples_per_second": 96.829,
"eval_steps_per_second": 5.096,
"step": 940
},
{
"epoch": 2.9230769230769234,
"grad_norm": 1.9051027297973633,
"learning_rate": 1.7830258624176224e-05,
"loss": 0.2102,
"step": 950
},
{
"epoch": 2.9230769230769234,
"eval_loss": 0.878767192363739,
"eval_runtime": 1.1684,
"eval_samples_per_second": 97.571,
"eval_steps_per_second": 5.135,
"step": 950
},
{
"epoch": 2.953846153846154,
"grad_norm": 1.8921849727630615,
"learning_rate": 1.776300360329488e-05,
"loss": 0.2331,
"step": 960
},
{
"epoch": 2.953846153846154,
"eval_loss": 0.8681402206420898,
"eval_runtime": 1.1693,
"eval_samples_per_second": 97.497,
"eval_steps_per_second": 5.131,
"step": 960
},
{
"epoch": 2.9846153846153847,
"grad_norm": 1.5593761205673218,
"learning_rate": 1.769485306586166e-05,
"loss": 0.2226,
"step": 970
},
{
"epoch": 2.9846153846153847,
"eval_loss": 0.8726215958595276,
"eval_runtime": 1.1712,
"eval_samples_per_second": 97.336,
"eval_steps_per_second": 5.123,
"step": 970
},
{
"epoch": 3.0153846153846153,
"grad_norm": 1.4075913429260254,
"learning_rate": 1.762581487351587e-05,
"loss": 0.1535,
"step": 980
},
{
"epoch": 3.0153846153846153,
"eval_loss": 0.9113463163375854,
"eval_runtime": 1.1794,
"eval_samples_per_second": 96.661,
"eval_steps_per_second": 5.087,
"step": 980
},
{
"epoch": 3.046153846153846,
"grad_norm": 1.410906434059143,
"learning_rate": 1.7555896990294003e-05,
"loss": 0.1021,
"step": 990
},
{
"epoch": 3.046153846153846,
"eval_loss": 0.9633604288101196,
"eval_runtime": 1.1702,
"eval_samples_per_second": 97.416,
"eval_steps_per_second": 5.127,
"step": 990
},
{
"epoch": 3.076923076923077,
"grad_norm": 1.0470328330993652,
"learning_rate": 1.7485107481711014e-05,
"loss": 0.1118,
"step": 1000
},
{
"epoch": 3.076923076923077,
"eval_loss": 0.9647616744041443,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.574,
"eval_steps_per_second": 5.135,
"step": 1000
},
{
"epoch": 3.1076923076923078,
"grad_norm": 2.098592758178711,
"learning_rate": 1.741345451382992e-05,
"loss": 0.1224,
"step": 1010
},
{
"epoch": 3.1076923076923078,
"eval_loss": 0.9478728175163269,
"eval_runtime": 1.1735,
"eval_samples_per_second": 97.143,
"eval_steps_per_second": 5.113,
"step": 1010
},
{
"epoch": 3.1384615384615384,
"grad_norm": 1.6263700723648071,
"learning_rate": 1.7340946352319795e-05,
"loss": 0.1377,
"step": 1020
},
{
"epoch": 3.1384615384615384,
"eval_loss": 0.922594428062439,
"eval_runtime": 1.1721,
"eval_samples_per_second": 97.262,
"eval_steps_per_second": 5.119,
"step": 1020
},
{
"epoch": 3.169230769230769,
"grad_norm": 1.0413360595703125,
"learning_rate": 1.7267591361502233e-05,
"loss": 0.1249,
"step": 1030
},
{
"epoch": 3.169230769230769,
"eval_loss": 0.9370183944702148,
"eval_runtime": 1.169,
"eval_samples_per_second": 97.519,
"eval_steps_per_second": 5.133,
"step": 1030
},
{
"epoch": 3.2,
"grad_norm": 1.6745877265930176,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.1186,
"step": 1040
},
{
"epoch": 3.2,
"eval_loss": 0.9573487639427185,
"eval_runtime": 1.1681,
"eval_samples_per_second": 97.596,
"eval_steps_per_second": 5.137,
"step": 1040
},
{
"epoch": 3.230769230769231,
"grad_norm": 1.3979861736297607,
"learning_rate": 1.7118374836693407e-05,
"loss": 0.135,
"step": 1050
},
{
"epoch": 3.230769230769231,
"eval_loss": 0.9554935693740845,
"eval_runtime": 1.1664,
"eval_samples_per_second": 97.735,
"eval_steps_per_second": 5.144,
"step": 1050
},
{
"epoch": 3.2615384615384615,
"grad_norm": 1.4017126560211182,
"learning_rate": 1.7042530515867897e-05,
"loss": 0.1102,
"step": 1060
},
{
"epoch": 3.2615384615384615,
"eval_loss": 0.9547311663627625,
"eval_runtime": 1.1685,
"eval_samples_per_second": 97.561,
"eval_steps_per_second": 5.135,
"step": 1060
},
{
"epoch": 3.292307692307692,
"grad_norm": 1.3390568494796753,
"learning_rate": 1.6965873790080806e-05,
"loss": 0.1358,
"step": 1070
},
{
"epoch": 3.292307692307692,
"eval_loss": 0.9695214629173279,
"eval_runtime": 1.1695,
"eval_samples_per_second": 97.474,
"eval_steps_per_second": 5.13,
"step": 1070
},
{
"epoch": 3.3230769230769233,
"grad_norm": 1.9002848863601685,
"learning_rate": 1.6888413502219534e-05,
"loss": 0.1419,
"step": 1080
},
{
"epoch": 3.3230769230769233,
"eval_loss": 0.9528248310089111,
"eval_runtime": 1.174,
"eval_samples_per_second": 97.107,
"eval_steps_per_second": 5.111,
"step": 1080
},
{
"epoch": 3.353846153846154,
"grad_norm": 1.7857142686843872,
"learning_rate": 1.6810158587867973e-05,
"loss": 0.128,
"step": 1090
},
{
"epoch": 3.353846153846154,
"eval_loss": 0.9331367611885071,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.282,
"eval_steps_per_second": 5.12,
"step": 1090
},
{
"epoch": 3.3846153846153846,
"grad_norm": 1.5178117752075195,
"learning_rate": 1.67311180742757e-05,
"loss": 0.112,
"step": 1100
},
{
"epoch": 3.3846153846153846,
"eval_loss": 0.9547919631004333,
"eval_runtime": 1.1695,
"eval_samples_per_second": 97.48,
"eval_steps_per_second": 5.131,
"step": 1100
},
{
"epoch": 3.4153846153846152,
"grad_norm": 1.3151490688323975,
"learning_rate": 1.665130107931666e-05,
"loss": 0.1119,
"step": 1110
},
{
"epoch": 3.4153846153846152,
"eval_loss": 0.9812674522399902,
"eval_runtime": 1.1666,
"eval_samples_per_second": 97.721,
"eval_steps_per_second": 5.143,
"step": 1110
},
{
"epoch": 3.4461538461538463,
"grad_norm": 1.4334938526153564,
"learning_rate": 1.657071681043731e-05,
"loss": 0.1042,
"step": 1120
},
{
"epoch": 3.4461538461538463,
"eval_loss": 0.9601649641990662,
"eval_runtime": 1.1706,
"eval_samples_per_second": 97.388,
"eval_steps_per_second": 5.126,
"step": 1120
},
{
"epoch": 3.476923076923077,
"grad_norm": 1.3328880071640015,
"learning_rate": 1.648937456359451e-05,
"loss": 0.1129,
"step": 1130
},
{
"epoch": 3.476923076923077,
"eval_loss": 0.9889415502548218,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.527,
"eval_steps_per_second": 5.133,
"step": 1130
},
{
"epoch": 3.5076923076923077,
"grad_norm": 1.8561350107192993,
"learning_rate": 1.640728372218317e-05,
"loss": 0.1247,
"step": 1140
},
{
"epoch": 3.5076923076923077,
"eval_loss": 0.9682590961456299,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.461,
"eval_steps_per_second": 5.13,
"step": 1140
},
{
"epoch": 3.5384615384615383,
"grad_norm": 1.1813424825668335,
"learning_rate": 1.6324453755953772e-05,
"loss": 0.1213,
"step": 1150
},
{
"epoch": 3.5384615384615383,
"eval_loss": 0.9461174607276917,
"eval_runtime": 1.1671,
"eval_samples_per_second": 97.674,
"eval_steps_per_second": 5.141,
"step": 1150
},
{
"epoch": 3.569230769230769,
"grad_norm": 1.2129303216934204,
"learning_rate": 1.624089421992003e-05,
"loss": 0.1253,
"step": 1160
},
{
"epoch": 3.569230769230769,
"eval_loss": 0.9655346274375916,
"eval_runtime": 1.1837,
"eval_samples_per_second": 96.312,
"eval_steps_per_second": 5.069,
"step": 1160
},
{
"epoch": 3.6,
"grad_norm": 2.2233059406280518,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.1147,
"step": 1170
},
{
"epoch": 3.6,
"eval_loss": 0.9837222099304199,
"eval_runtime": 1.168,
"eval_samples_per_second": 97.6,
"eval_steps_per_second": 5.137,
"step": 1170
},
{
"epoch": 3.6307692307692307,
"grad_norm": 1.382871389389038,
"learning_rate": 1.6071625078187113e-05,
"loss": 0.1272,
"step": 1180
},
{
"epoch": 3.6307692307692307,
"eval_loss": 0.9973061680793762,
"eval_runtime": 1.1757,
"eval_samples_per_second": 96.964,
"eval_steps_per_second": 5.103,
"step": 1180
},
{
"epoch": 3.6615384615384614,
"grad_norm": 1.2613117694854736,
"learning_rate": 1.5985934998862775e-05,
"loss": 0.1367,
"step": 1190
},
{
"epoch": 3.6615384615384614,
"eval_loss": 0.9685854315757751,
"eval_runtime": 1.1695,
"eval_samples_per_second": 97.474,
"eval_steps_per_second": 5.13,
"step": 1190
},
{
"epoch": 3.6923076923076925,
"grad_norm": 1.0769059658050537,
"learning_rate": 1.5899554400231233e-05,
"loss": 0.1208,
"step": 1200
},
{
"epoch": 3.6923076923076925,
"eval_loss": 0.9511159062385559,
"eval_runtime": 1.1684,
"eval_samples_per_second": 97.566,
"eval_steps_per_second": 5.135,
"step": 1200
},
{
"epoch": 3.723076923076923,
"grad_norm": 2.335827350616455,
"learning_rate": 1.5812493246896368e-05,
"loss": 0.1315,
"step": 1210
},
{
"epoch": 3.723076923076923,
"eval_loss": 0.9450687766075134,
"eval_runtime": 1.1715,
"eval_samples_per_second": 97.309,
"eval_steps_per_second": 5.122,
"step": 1210
},
{
"epoch": 3.753846153846154,
"grad_norm": 1.2259918451309204,
"learning_rate": 1.572476158196879e-05,
"loss": 0.1252,
"step": 1220
},
{
"epoch": 3.753846153846154,
"eval_loss": 0.9482319355010986,
"eval_runtime": 1.1688,
"eval_samples_per_second": 97.537,
"eval_steps_per_second": 5.134,
"step": 1220
},
{
"epoch": 3.7846153846153845,
"grad_norm": 1.4782612323760986,
"learning_rate": 1.5636369525907297e-05,
"loss": 0.1424,
"step": 1230
},
{
"epoch": 3.7846153846153845,
"eval_loss": 0.9551723599433899,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.427,
"eval_steps_per_second": 5.128,
"step": 1230
},
{
"epoch": 3.815384615384615,
"grad_norm": 1.3023622035980225,
"learning_rate": 1.554732727535139e-05,
"loss": 0.121,
"step": 1240
},
{
"epoch": 3.815384615384615,
"eval_loss": 0.9602871537208557,
"eval_runtime": 1.1726,
"eval_samples_per_second": 97.217,
"eval_steps_per_second": 5.117,
"step": 1240
},
{
"epoch": 3.8461538461538463,
"grad_norm": 1.2243452072143555,
"learning_rate": 1.5457645101945046e-05,
"loss": 0.1213,
"step": 1250
},
{
"epoch": 3.8461538461538463,
"eval_loss": 0.9587708711624146,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.253,
"eval_steps_per_second": 5.119,
"step": 1250
},
{
"epoch": 3.876923076923077,
"grad_norm": 1.085587739944458,
"learning_rate": 1.5367333351151804e-05,
"loss": 0.1273,
"step": 1260
},
{
"epoch": 3.876923076923077,
"eval_loss": 0.9537996053695679,
"eval_runtime": 1.1669,
"eval_samples_per_second": 97.696,
"eval_steps_per_second": 5.142,
"step": 1260
},
{
"epoch": 3.9076923076923076,
"grad_norm": 1.8853720426559448,
"learning_rate": 1.527640244106133e-05,
"loss": 0.1411,
"step": 1270
},
{
"epoch": 3.9076923076923076,
"eval_loss": 0.953859269618988,
"eval_runtime": 1.1705,
"eval_samples_per_second": 97.392,
"eval_steps_per_second": 5.126,
"step": 1270
},
{
"epoch": 3.9384615384615387,
"grad_norm": 1.3071255683898926,
"learning_rate": 1.5184862861187637e-05,
"loss": 0.1339,
"step": 1280
},
{
"epoch": 3.9384615384615387,
"eval_loss": 0.9453871846199036,
"eval_runtime": 1.169,
"eval_samples_per_second": 97.523,
"eval_steps_per_second": 5.133,
"step": 1280
},
{
"epoch": 3.9692307692307693,
"grad_norm": 1.2340234518051147,
"learning_rate": 1.5092725171259037e-05,
"loss": 0.1336,
"step": 1290
},
{
"epoch": 3.9692307692307693,
"eval_loss": 0.9494842290878296,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.257,
"eval_steps_per_second": 5.119,
"step": 1290
},
{
"epoch": 4.0,
"grad_norm": 1.270678162574768,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1132,
"step": 1300
},
{
"epoch": 4.0,
"eval_loss": 0.9589464068412781,
"eval_runtime": 1.174,
"eval_samples_per_second": 97.106,
"eval_steps_per_second": 5.111,
"step": 1300
},
{
"epoch": 4.030769230769231,
"grad_norm": 0.9828625321388245,
"learning_rate": 1.4906698043905067e-05,
"loss": 0.0818,
"step": 1310
},
{
"epoch": 4.030769230769231,
"eval_loss": 0.9852768778800964,
"eval_runtime": 1.1812,
"eval_samples_per_second": 96.514,
"eval_steps_per_second": 5.08,
"step": 1310
},
{
"epoch": 4.061538461538461,
"grad_norm": 0.7625837326049805,
"learning_rate": 1.481283006600493e-05,
"loss": 0.0729,
"step": 1320
},
{
"epoch": 4.061538461538461,
"eval_loss": 1.0075314044952393,
"eval_runtime": 1.1741,
"eval_samples_per_second": 97.098,
"eval_steps_per_second": 5.11,
"step": 1320
},
{
"epoch": 4.092307692307692,
"grad_norm": 0.9247090220451355,
"learning_rate": 1.471840689462482e-05,
"loss": 0.0856,
"step": 1330
},
{
"epoch": 4.092307692307692,
"eval_loss": 1.0200262069702148,
"eval_runtime": 1.1737,
"eval_samples_per_second": 97.126,
"eval_steps_per_second": 5.112,
"step": 1330
},
{
"epoch": 4.123076923076923,
"grad_norm": 1.220361351966858,
"learning_rate": 1.4623439422135435e-05,
"loss": 0.0831,
"step": 1340
},
{
"epoch": 4.123076923076923,
"eval_loss": 1.0162899494171143,
"eval_runtime": 1.1736,
"eval_samples_per_second": 97.135,
"eval_steps_per_second": 5.112,
"step": 1340
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.8103403449058533,
"learning_rate": 1.4527938603696376e-05,
"loss": 0.0876,
"step": 1350
},
{
"epoch": 4.153846153846154,
"eval_loss": 1.0209596157073975,
"eval_runtime": 1.1775,
"eval_samples_per_second": 96.818,
"eval_steps_per_second": 5.096,
"step": 1350
},
{
"epoch": 4.184615384615385,
"grad_norm": 1.0180028676986694,
"learning_rate": 1.4431915455992416e-05,
"loss": 0.0895,
"step": 1360
},
{
"epoch": 4.184615384615385,
"eval_loss": 1.0170763731002808,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.427,
"eval_steps_per_second": 5.128,
"step": 1360
},
{
"epoch": 4.2153846153846155,
"grad_norm": 1.0312772989273071,
"learning_rate": 1.4335381055962657e-05,
"loss": 0.0836,
"step": 1370
},
{
"epoch": 4.2153846153846155,
"eval_loss": 1.0018525123596191,
"eval_runtime": 1.171,
"eval_samples_per_second": 97.355,
"eval_steps_per_second": 5.124,
"step": 1370
},
{
"epoch": 4.246153846153846,
"grad_norm": 0.5939295887947083,
"learning_rate": 1.4238346539522716e-05,
"loss": 0.0882,
"step": 1380
},
{
"epoch": 4.246153846153846,
"eval_loss": 1.0118681192398071,
"eval_runtime": 1.1723,
"eval_samples_per_second": 97.246,
"eval_steps_per_second": 5.118,
"step": 1380
},
{
"epoch": 4.276923076923077,
"grad_norm": 0.9801693558692932,
"learning_rate": 1.414082310028012e-05,
"loss": 0.0866,
"step": 1390
},
{
"epoch": 4.276923076923077,
"eval_loss": 1.0159003734588623,
"eval_runtime": 1.1679,
"eval_samples_per_second": 97.608,
"eval_steps_per_second": 5.137,
"step": 1390
},
{
"epoch": 4.3076923076923075,
"grad_norm": 1.014660358428955,
"learning_rate": 1.404282198824305e-05,
"loss": 0.0861,
"step": 1400
},
{
"epoch": 4.3076923076923075,
"eval_loss": 1.0265697240829468,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.458,
"eval_steps_per_second": 5.129,
"step": 1400
},
{
"epoch": 4.338461538461538,
"grad_norm": 0.9535912871360779,
"learning_rate": 1.3944354508522573e-05,
"loss": 0.0904,
"step": 1410
},
{
"epoch": 4.338461538461538,
"eval_loss": 1.028001308441162,
"eval_runtime": 1.1692,
"eval_samples_per_second": 97.502,
"eval_steps_per_second": 5.132,
"step": 1410
},
{
"epoch": 4.36923076923077,
"grad_norm": 0.8744879364967346,
"learning_rate": 1.3845432020028511e-05,
"loss": 0.0841,
"step": 1420
},
{
"epoch": 4.36923076923077,
"eval_loss": 1.0107029676437378,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.463,
"eval_steps_per_second": 5.13,
"step": 1420
},
{
"epoch": 4.4,
"grad_norm": 1.0251070261001587,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0847,
"step": 1430
},
{
"epoch": 4.4,
"eval_loss": 1.0237025022506714,
"eval_runtime": 1.1719,
"eval_samples_per_second": 97.275,
"eval_steps_per_second": 5.12,
"step": 1430
},
{
"epoch": 4.430769230769231,
"grad_norm": 1.2007213830947876,
"learning_rate": 1.3646267713484718e-05,
"loss": 0.0813,
"step": 1440
},
{
"epoch": 4.430769230769231,
"eval_loss": 1.043028712272644,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.464,
"eval_steps_per_second": 5.13,
"step": 1440
},
{
"epoch": 4.461538461538462,
"grad_norm": 1.0892722606658936,
"learning_rate": 1.3546048870425356e-05,
"loss": 0.0813,
"step": 1450
},
{
"epoch": 4.461538461538462,
"eval_loss": 1.0351496934890747,
"eval_runtime": 1.1715,
"eval_samples_per_second": 97.313,
"eval_steps_per_second": 5.122,
"step": 1450
},
{
"epoch": 4.492307692307692,
"grad_norm": 1.0319191217422485,
"learning_rate": 1.3445420965922837e-05,
"loss": 0.0824,
"step": 1460
},
{
"epoch": 4.492307692307692,
"eval_loss": 1.0122759342193604,
"eval_runtime": 1.1695,
"eval_samples_per_second": 97.478,
"eval_steps_per_second": 5.13,
"step": 1460
},
{
"epoch": 4.523076923076923,
"grad_norm": 1.9638742208480835,
"learning_rate": 1.3344395608107032e-05,
"loss": 0.085,
"step": 1470
},
{
"epoch": 4.523076923076923,
"eval_loss": 1.02043616771698,
"eval_runtime": 1.1682,
"eval_samples_per_second": 97.59,
"eval_steps_per_second": 5.136,
"step": 1470
},
{
"epoch": 4.553846153846154,
"grad_norm": 1.267501711845398,
"learning_rate": 1.3242984450956829e-05,
"loss": 0.0926,
"step": 1480
},
{
"epoch": 4.553846153846154,
"eval_loss": 1.0314496755599976,
"eval_runtime": 1.1689,
"eval_samples_per_second": 97.527,
"eval_steps_per_second": 5.133,
"step": 1480
},
{
"epoch": 4.584615384615384,
"grad_norm": 0.779525101184845,
"learning_rate": 1.3141199192955751e-05,
"loss": 0.088,
"step": 1490
},
{
"epoch": 4.584615384615384,
"eval_loss": 1.0142338275909424,
"eval_runtime": 1.1706,
"eval_samples_per_second": 97.388,
"eval_steps_per_second": 5.126,
"step": 1490
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.7838029265403748,
"learning_rate": 1.303905157574247e-05,
"loss": 0.0832,
"step": 1500
},
{
"epoch": 4.615384615384615,
"eval_loss": 1.0125129222869873,
"eval_runtime": 1.1763,
"eval_samples_per_second": 96.915,
"eval_steps_per_second": 5.101,
"step": 1500
},
{
"epoch": 4.6461538461538465,
"grad_norm": 0.85016930103302,
"learning_rate": 1.293655338275631e-05,
"loss": 0.0823,
"step": 1510
},
{
"epoch": 4.6461538461538465,
"eval_loss": 1.0058341026306152,
"eval_runtime": 1.1706,
"eval_samples_per_second": 97.385,
"eval_steps_per_second": 5.126,
"step": 1510
},
{
"epoch": 4.676923076923077,
"grad_norm": 0.9694855809211731,
"learning_rate": 1.2833716437877952e-05,
"loss": 0.0961,
"step": 1520
},
{
"epoch": 4.676923076923077,
"eval_loss": 1.011164665222168,
"eval_runtime": 1.1696,
"eval_samples_per_second": 97.467,
"eval_steps_per_second": 5.13,
"step": 1520
},
{
"epoch": 4.707692307692308,
"grad_norm": 1.1941697597503662,
"learning_rate": 1.2730552604065477e-05,
"loss": 0.0934,
"step": 1530
},
{
"epoch": 4.707692307692308,
"eval_loss": 1.0244613885879517,
"eval_runtime": 1.1676,
"eval_samples_per_second": 97.636,
"eval_steps_per_second": 5.139,
"step": 1530
},
{
"epoch": 4.7384615384615385,
"grad_norm": 0.946321427822113,
"learning_rate": 1.262707378198587e-05,
"loss": 0.0868,
"step": 1540
},
{
"epoch": 4.7384615384615385,
"eval_loss": 1.0241734981536865,
"eval_runtime": 1.1705,
"eval_samples_per_second": 97.393,
"eval_steps_per_second": 5.126,
"step": 1540
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.9640923142433167,
"learning_rate": 1.2523291908642219e-05,
"loss": 0.0834,
"step": 1550
},
{
"epoch": 4.769230769230769,
"eval_loss": 1.026370644569397,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.283,
"eval_steps_per_second": 5.12,
"step": 1550
},
{
"epoch": 4.8,
"grad_norm": 1.2922747135162354,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.0867,
"step": 1560
},
{
"epoch": 4.8,
"eval_loss": 1.0209637880325317,
"eval_runtime": 1.1729,
"eval_samples_per_second": 97.195,
"eval_steps_per_second": 5.116,
"step": 1560
},
{
"epoch": 4.8307692307692305,
"grad_norm": 1.8624268770217896,
"learning_rate": 1.2314866929589434e-05,
"loss": 0.0805,
"step": 1570
},
{
"epoch": 4.8307692307692305,
"eval_loss": 1.0203014612197876,
"eval_runtime": 1.1728,
"eval_samples_per_second": 97.199,
"eval_steps_per_second": 5.116,
"step": 1570
},
{
"epoch": 4.861538461538462,
"grad_norm": 0.5643526911735535,
"learning_rate": 1.2210247867153763e-05,
"loss": 0.0909,
"step": 1580
},
{
"epoch": 4.861538461538462,
"eval_loss": 1.014268159866333,
"eval_runtime": 1.1727,
"eval_samples_per_second": 97.213,
"eval_steps_per_second": 5.116,
"step": 1580
},
{
"epoch": 4.892307692307693,
"grad_norm": 1.2319228649139404,
"learning_rate": 1.2105373837227425e-05,
"loss": 0.0872,
"step": 1590
},
{
"epoch": 4.892307692307693,
"eval_loss": 1.0142085552215576,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.424,
"eval_steps_per_second": 5.128,
"step": 1590
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.7346972823143005,
"learning_rate": 1.2000256937760446e-05,
"loss": 0.0812,
"step": 1600
},
{
"epoch": 4.923076923076923,
"eval_loss": 1.011918544769287,
"eval_runtime": 1.1706,
"eval_samples_per_second": 97.388,
"eval_steps_per_second": 5.126,
"step": 1600
},
{
"epoch": 4.953846153846154,
"grad_norm": 0.6954300403594971,
"learning_rate": 1.1894909294719547e-05,
"loss": 0.0874,
"step": 1610
},
{
"epoch": 4.953846153846154,
"eval_loss": 1.0086393356323242,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.444,
"eval_steps_per_second": 5.129,
"step": 1610
},
{
"epoch": 4.984615384615385,
"grad_norm": 0.7410770058631897,
"learning_rate": 1.178934306068933e-05,
"loss": 0.0762,
"step": 1620
},
{
"epoch": 4.984615384615385,
"eval_loss": 1.0001431703567505,
"eval_runtime": 1.1682,
"eval_samples_per_second": 97.586,
"eval_steps_per_second": 5.136,
"step": 1620
},
{
"epoch": 5.015384615384615,
"grad_norm": 0.5212005376815796,
"learning_rate": 1.1683570413470384e-05,
"loss": 0.0659,
"step": 1630
},
{
"epoch": 5.015384615384615,
"eval_loss": 1.0165449380874634,
"eval_runtime": 1.167,
"eval_samples_per_second": 97.683,
"eval_steps_per_second": 5.141,
"step": 1630
},
{
"epoch": 5.046153846153846,
"grad_norm": 1.2066491842269897,
"learning_rate": 1.1577603554674515e-05,
"loss": 0.0611,
"step": 1640
},
{
"epoch": 5.046153846153846,
"eval_loss": 1.0587912797927856,
"eval_runtime": 1.173,
"eval_samples_per_second": 97.19,
"eval_steps_per_second": 5.115,
"step": 1640
},
{
"epoch": 5.076923076923077,
"grad_norm": 0.6969051361083984,
"learning_rate": 1.1471454708317163e-05,
"loss": 0.061,
"step": 1650
},
{
"epoch": 5.076923076923077,
"eval_loss": 1.0719058513641357,
"eval_runtime": 1.1708,
"eval_samples_per_second": 97.372,
"eval_steps_per_second": 5.125,
"step": 1650
},
{
"epoch": 5.107692307692307,
"grad_norm": 0.7933226227760315,
"learning_rate": 1.1365136119407318e-05,
"loss": 0.063,
"step": 1660
},
{
"epoch": 5.107692307692307,
"eval_loss": 1.0709913969039917,
"eval_runtime": 1.1672,
"eval_samples_per_second": 97.668,
"eval_steps_per_second": 5.14,
"step": 1660
},
{
"epoch": 5.138461538461539,
"grad_norm": 0.6501964926719666,
"learning_rate": 1.125866005253495e-05,
"loss": 0.0626,
"step": 1670
},
{
"epoch": 5.138461538461539,
"eval_loss": 1.0556234121322632,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.447,
"eval_steps_per_second": 5.129,
"step": 1670
},
{
"epoch": 5.1692307692307695,
"grad_norm": 0.6732635498046875,
"learning_rate": 1.1152038790456212e-05,
"loss": 0.0606,
"step": 1680
},
{
"epoch": 5.1692307692307695,
"eval_loss": 1.0559837818145752,
"eval_runtime": 1.168,
"eval_samples_per_second": 97.604,
"eval_steps_per_second": 5.137,
"step": 1680
},
{
"epoch": 5.2,
"grad_norm": 0.9151493310928345,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0664,
"step": 1690
},
{
"epoch": 5.2,
"eval_loss": 1.0514063835144043,
"eval_runtime": 1.1711,
"eval_samples_per_second": 97.347,
"eval_steps_per_second": 5.124,
"step": 1690
},
{
"epoch": 5.230769230769231,
"grad_norm": 0.5691230893135071,
"learning_rate": 1.0938409894031793e-05,
"loss": 0.0588,
"step": 1700
},
{
"epoch": 5.230769230769231,
"eval_loss": 1.0502656698226929,
"eval_runtime": 1.1718,
"eval_samples_per_second": 97.287,
"eval_steps_per_second": 5.12,
"step": 1700
},
{
"epoch": 5.2615384615384615,
"grad_norm": 0.972598135471344,
"learning_rate": 1.0831426903267706e-05,
"loss": 0.0647,
"step": 1710
},
{
"epoch": 5.2615384615384615,
"eval_loss": 1.0546636581420898,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.252,
"eval_steps_per_second": 5.119,
"step": 1710
},
{
"epoch": 5.292307692307692,
"grad_norm": 0.36528125405311584,
"learning_rate": 1.0724348001617626e-05,
"loss": 0.0559,
"step": 1720
},
{
"epoch": 5.292307692307692,
"eval_loss": 1.0595762729644775,
"eval_runtime": 1.1798,
"eval_samples_per_second": 96.63,
"eval_steps_per_second": 5.086,
"step": 1720
},
{
"epoch": 5.323076923076923,
"grad_norm": 0.8283866047859192,
"learning_rate": 1.0617185541378895e-05,
"loss": 0.0632,
"step": 1730
},
{
"epoch": 5.323076923076923,
"eval_loss": 1.0664962530136108,
"eval_runtime": 1.1688,
"eval_samples_per_second": 97.54,
"eval_steps_per_second": 5.134,
"step": 1730
},
{
"epoch": 5.3538461538461535,
"grad_norm": 0.8648726940155029,
"learning_rate": 1.0509951884487926e-05,
"loss": 0.0596,
"step": 1740
},
{
"epoch": 5.3538461538461535,
"eval_loss": 1.068932056427002,
"eval_runtime": 1.1698,
"eval_samples_per_second": 97.449,
"eval_steps_per_second": 5.129,
"step": 1740
},
{
"epoch": 5.384615384615385,
"grad_norm": 0.4078614115715027,
"learning_rate": 1.0402659401094154e-05,
"loss": 0.0614,
"step": 1750
},
{
"epoch": 5.384615384615385,
"eval_loss": 1.072813630104065,
"eval_runtime": 1.1737,
"eval_samples_per_second": 97.133,
"eval_steps_per_second": 5.112,
"step": 1750
},
{
"epoch": 5.415384615384616,
"grad_norm": 0.7758128046989441,
"learning_rate": 1.0295320468133066e-05,
"loss": 0.0665,
"step": 1760
},
{
"epoch": 5.415384615384616,
"eval_loss": 1.0686752796173096,
"eval_runtime": 1.1717,
"eval_samples_per_second": 97.292,
"eval_steps_per_second": 5.121,
"step": 1760
},
{
"epoch": 5.446153846153846,
"grad_norm": 0.5912731885910034,
"learning_rate": 1.0187947467898425e-05,
"loss": 0.0652,
"step": 1770
},
{
"epoch": 5.446153846153846,
"eval_loss": 1.0550626516342163,
"eval_runtime": 1.173,
"eval_samples_per_second": 97.19,
"eval_steps_per_second": 5.115,
"step": 1770
},
{
"epoch": 5.476923076923077,
"grad_norm": 0.7513710856437683,
"learning_rate": 1.0080552786613899e-05,
"loss": 0.056,
"step": 1780
},
{
"epoch": 5.476923076923077,
"eval_loss": 1.0569382905960083,
"eval_runtime": 1.1739,
"eval_samples_per_second": 97.109,
"eval_steps_per_second": 5.111,
"step": 1780
},
{
"epoch": 5.507692307692308,
"grad_norm": 0.4815368354320526,
"learning_rate": 9.973148813004216e-06,
"loss": 0.0604,
"step": 1790
},
{
"epoch": 5.507692307692308,
"eval_loss": 1.060049295425415,
"eval_runtime": 1.1763,
"eval_samples_per_second": 96.912,
"eval_steps_per_second": 5.101,
"step": 1790
},
{
"epoch": 5.538461538461538,
"grad_norm": 0.6192671060562134,
"learning_rate": 9.865747936866027e-06,
"loss": 0.0673,
"step": 1800
},
{
"epoch": 5.538461538461538,
"eval_loss": 1.0688085556030273,
"eval_runtime": 1.1719,
"eval_samples_per_second": 97.274,
"eval_steps_per_second": 5.12,
"step": 1800
},
{
"epoch": 5.569230769230769,
"grad_norm": 0.9673665165901184,
"learning_rate": 9.75836254763868e-06,
"loss": 0.0626,
"step": 1810
},
{
"epoch": 5.569230769230769,
"eval_loss": 1.0702043771743774,
"eval_runtime": 1.1759,
"eval_samples_per_second": 96.946,
"eval_steps_per_second": 5.102,
"step": 1810
},
{
"epoch": 5.6,
"grad_norm": 0.7563217878341675,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0616,
"step": 1820
},
{
"epoch": 5.6,
"eval_loss": 1.07481849193573,
"eval_runtime": 1.1687,
"eval_samples_per_second": 97.546,
"eval_steps_per_second": 5.134,
"step": 1820
},
{
"epoch": 5.63076923076923,
"grad_norm": 1.084849238395691,
"learning_rate": 9.543687777312264e-06,
"loss": 0.0708,
"step": 1830
},
{
"epoch": 5.63076923076923,
"eval_loss": 1.0818299055099487,
"eval_runtime": 1.1698,
"eval_samples_per_second": 97.449,
"eval_steps_per_second": 5.129,
"step": 1830
},
{
"epoch": 5.661538461538462,
"grad_norm": 0.505901038646698,
"learning_rate": 9.436423160443625e-06,
"loss": 0.0579,
"step": 1840
},
{
"epoch": 5.661538461538462,
"eval_loss": 1.078920841217041,
"eval_runtime": 1.1793,
"eval_samples_per_second": 96.666,
"eval_steps_per_second": 5.088,
"step": 1840
},
{
"epoch": 5.6923076923076925,
"grad_norm": 0.44022971391677856,
"learning_rate": 9.329223556089976e-06,
"loss": 0.0629,
"step": 1850
},
{
"epoch": 5.6923076923076925,
"eval_loss": 1.0715224742889404,
"eval_runtime": 1.1676,
"eval_samples_per_second": 97.638,
"eval_steps_per_second": 5.139,
"step": 1850
},
{
"epoch": 5.723076923076923,
"grad_norm": 1.5358976125717163,
"learning_rate": 9.222101330472553e-06,
"loss": 0.0686,
"step": 1860
},
{
"epoch": 5.723076923076923,
"eval_loss": 1.0757378339767456,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.58,
"eval_steps_per_second": 5.136,
"step": 1860
},
{
"epoch": 5.753846153846154,
"grad_norm": 0.84912109375,
"learning_rate": 9.115068840886418e-06,
"loss": 0.0644,
"step": 1870
},
{
"epoch": 5.753846153846154,
"eval_loss": 1.0791435241699219,
"eval_runtime": 1.1703,
"eval_samples_per_second": 97.409,
"eval_steps_per_second": 5.127,
"step": 1870
},
{
"epoch": 5.7846153846153845,
"grad_norm": 0.7800810933113098,
"learning_rate": 9.00813843427496e-06,
"loss": 0.0588,
"step": 1880
},
{
"epoch": 5.7846153846153845,
"eval_loss": 1.082220435142517,
"eval_runtime": 1.1687,
"eval_samples_per_second": 97.544,
"eval_steps_per_second": 5.134,
"step": 1880
},
{
"epoch": 5.815384615384615,
"grad_norm": 0.6433857083320618,
"learning_rate": 8.901322445805586e-06,
"loss": 0.0636,
"step": 1890
},
{
"epoch": 5.815384615384615,
"eval_loss": 1.079559087753296,
"eval_runtime": 1.171,
"eval_samples_per_second": 97.353,
"eval_steps_per_second": 5.124,
"step": 1890
},
{
"epoch": 5.846153846153846,
"grad_norm": 0.6224117279052734,
"learning_rate": 8.79463319744677e-06,
"loss": 0.0609,
"step": 1900
},
{
"epoch": 5.846153846153846,
"eval_loss": 1.0735894441604614,
"eval_runtime": 1.1743,
"eval_samples_per_second": 97.083,
"eval_steps_per_second": 5.11,
"step": 1900
},
{
"epoch": 5.876923076923077,
"grad_norm": 0.6254037618637085,
"learning_rate": 8.688082996546639e-06,
"loss": 0.0673,
"step": 1910
},
{
"epoch": 5.876923076923077,
"eval_loss": 1.0738794803619385,
"eval_runtime": 1.1747,
"eval_samples_per_second": 97.048,
"eval_steps_per_second": 5.108,
"step": 1910
},
{
"epoch": 5.907692307692308,
"grad_norm": 0.45801621675491333,
"learning_rate": 8.581684134413216e-06,
"loss": 0.0614,
"step": 1920
},
{
"epoch": 5.907692307692308,
"eval_loss": 1.0641692876815796,
"eval_runtime": 1.1691,
"eval_samples_per_second": 97.508,
"eval_steps_per_second": 5.132,
"step": 1920
},
{
"epoch": 5.938461538461539,
"grad_norm": 1.3301217555999756,
"learning_rate": 8.475448884896546e-06,
"loss": 0.0582,
"step": 1930
},
{
"epoch": 5.938461538461539,
"eval_loss": 1.0678892135620117,
"eval_runtime": 1.1724,
"eval_samples_per_second": 97.238,
"eval_steps_per_second": 5.118,
"step": 1930
},
{
"epoch": 5.969230769230769,
"grad_norm": 0.40689530968666077,
"learning_rate": 8.36938950297283e-06,
"loss": 0.0626,
"step": 1940
},
{
"epoch": 5.969230769230769,
"eval_loss": 1.078120470046997,
"eval_runtime": 1.1684,
"eval_samples_per_second": 97.571,
"eval_steps_per_second": 5.135,
"step": 1940
},
{
"epoch": 6.0,
"grad_norm": 0.8460783362388611,
"learning_rate": 8.263518223330698e-06,
"loss": 0.0688,
"step": 1950
},
{
"epoch": 6.0,
"eval_loss": 1.0826774835586548,
"eval_runtime": 1.1888,
"eval_samples_per_second": 95.897,
"eval_steps_per_second": 5.047,
"step": 1950
},
{
"epoch": 6.030769230769231,
"grad_norm": 0.547752320766449,
"learning_rate": 8.157847258959885e-06,
"loss": 0.0495,
"step": 1960
},
{
"epoch": 6.030769230769231,
"eval_loss": 1.103231430053711,
"eval_runtime": 1.1725,
"eval_samples_per_second": 97.23,
"eval_steps_per_second": 5.117,
"step": 1960
},
{
"epoch": 6.061538461538461,
"grad_norm": 0.7027488350868225,
"learning_rate": 8.05238879974236e-06,
"loss": 0.0436,
"step": 1970
},
{
"epoch": 6.061538461538461,
"eval_loss": 1.123745322227478,
"eval_runtime": 1.1696,
"eval_samples_per_second": 97.47,
"eval_steps_per_second": 5.13,
"step": 1970
},
{
"epoch": 6.092307692307692,
"grad_norm": 0.7175533175468445,
"learning_rate": 7.947155011046144e-06,
"loss": 0.0533,
"step": 1980
},
{
"epoch": 6.092307692307692,
"eval_loss": 1.128766417503357,
"eval_runtime": 1.1697,
"eval_samples_per_second": 97.461,
"eval_steps_per_second": 5.13,
"step": 1980
},
{
"epoch": 6.123076923076923,
"grad_norm": 0.6020351648330688,
"learning_rate": 7.84215803232194e-06,
"loss": 0.0506,
"step": 1990
},
{
"epoch": 6.123076923076923,
"eval_loss": 1.1175923347473145,
"eval_runtime": 1.1699,
"eval_samples_per_second": 97.442,
"eval_steps_per_second": 5.129,
"step": 1990
},
{
"epoch": 6.153846153846154,
"grad_norm": 11.60201358795166,
"learning_rate": 7.73740997570278e-06,
"loss": 0.0494,
"step": 2000
},
{
"epoch": 6.153846153846154,
"eval_loss": 1.124334692955017,
"eval_runtime": 1.173,
"eval_samples_per_second": 97.184,
"eval_steps_per_second": 5.115,
"step": 2000
},
{
"epoch": 6.184615384615385,
"grad_norm": 0.6954211592674255,
"learning_rate": 7.632922924606794e-06,
"loss": 0.0484,
"step": 2010
},
{
"epoch": 6.184615384615385,
"eval_loss": 1.112605094909668,
"eval_runtime": 1.1728,
"eval_samples_per_second": 97.203,
"eval_steps_per_second": 5.116,
"step": 2010
},
{
"epoch": 6.2153846153846155,
"grad_norm": 0.700815737247467,
"learning_rate": 7.5287089323433035e-06,
"loss": 0.0455,
"step": 2020
},
{
"epoch": 6.2153846153846155,
"eval_loss": 1.1227997541427612,
"eval_runtime": 1.1686,
"eval_samples_per_second": 97.553,
"eval_steps_per_second": 5.134,
"step": 2020
},
{
"epoch": 6.246153846153846,
"grad_norm": 0.7772124409675598,
"learning_rate": 7.4247800207223975e-06,
"loss": 0.0423,
"step": 2030
},
{
"epoch": 6.246153846153846,
"eval_loss": 1.1330678462982178,
"eval_runtime": 1.1704,
"eval_samples_per_second": 97.405,
"eval_steps_per_second": 5.127,
"step": 2030
},
{
"epoch": 6.276923076923077,
"grad_norm": 0.4515056908130646,
"learning_rate": 7.32114817866812e-06,
"loss": 0.0506,
"step": 2040
},
{
"epoch": 6.276923076923077,
"eval_loss": 1.1395100355148315,
"eval_runtime": 1.1683,
"eval_samples_per_second": 97.576,
"eval_steps_per_second": 5.136,
"step": 2040
},
{
"epoch": 6.3076923076923075,
"grad_norm": 0.24425525963306427,
"learning_rate": 7.217825360835475e-06,
"loss": 0.0537,
"step": 2050
},
{
"epoch": 6.3076923076923075,
"eval_loss": 1.1437585353851318,
"eval_runtime": 1.1701,
"eval_samples_per_second": 97.429,
"eval_steps_per_second": 5.128,
"step": 2050
},
{
"epoch": 6.338461538461538,
"grad_norm": 0.8167008757591248,
"learning_rate": 7.114823486231366e-06,
"loss": 0.0494,
"step": 2060
},
{
"epoch": 6.338461538461538,
"eval_loss": 1.151004433631897,
"eval_runtime": 1.1733,
"eval_samples_per_second": 97.16,
"eval_steps_per_second": 5.114,
"step": 2060
},
{
"epoch": 6.36923076923077,
"grad_norm": 0.6093922257423401,
"learning_rate": 7.012154436839664e-06,
"loss": 0.0476,
"step": 2070
},
{
"epoch": 6.36923076923077,
"eval_loss": 1.1457868814468384,
"eval_runtime": 1.174,
"eval_samples_per_second": 97.106,
"eval_steps_per_second": 5.111,
"step": 2070
},
{
"epoch": 6.4,
"grad_norm": 0.6646329164505005,
"learning_rate": 6.909830056250527e-06,
"loss": 0.047,
"step": 2080
},
{
"epoch": 6.4,
"eval_loss": 1.1414220333099365,
"eval_runtime": 1.1714,
"eval_samples_per_second": 97.316,
"eval_steps_per_second": 5.122,
"step": 2080
},
{
"epoch": 6.430769230769231,
"grad_norm": 0.612014889717102,
"learning_rate": 6.807862148294171e-06,
"loss": 0.0482,
"step": 2090
},
{
"epoch": 6.430769230769231,
"eval_loss": 1.1401718854904175,
"eval_runtime": 1.1722,
"eval_samples_per_second": 97.257,
"eval_steps_per_second": 5.119,
"step": 2090
},
{
"epoch": 6.461538461538462,
"grad_norm": 1.0283567905426025,
"learning_rate": 6.706262475679205e-06,
"loss": 0.0498,
"step": 2100
},
{
"epoch": 6.461538461538462,
"eval_loss": 1.1398248672485352,
"eval_runtime": 1.1747,
"eval_samples_per_second": 97.046,
"eval_steps_per_second": 5.108,
"step": 2100
}
],
"logging_steps": 10,
"max_steps": 3250,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 50,
"total_flos": 1.3125679576619418e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}