2773 lines
62 KiB
JSON
2773 lines
62 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 4.992,
|
|
"eval_steps": 500,
|
|
"global_step": 390,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0128,
|
|
"grad_norm": 7.692362944768503,
|
|
"learning_rate": 1.0256410256410257e-06,
|
|
"loss": 1.5558,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.0256,
|
|
"grad_norm": 7.649452387526613,
|
|
"learning_rate": 2.0512820512820513e-06,
|
|
"loss": 1.5299,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.0384,
|
|
"grad_norm": 7.671202171277273,
|
|
"learning_rate": 3.0769230769230774e-06,
|
|
"loss": 1.5814,
|
|
"step": 3
|
|
},
|
|
{
|
|
"epoch": 0.0512,
|
|
"grad_norm": 7.036877104775377,
|
|
"learning_rate": 4.102564102564103e-06,
|
|
"loss": 1.54,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.064,
|
|
"grad_norm": 5.292988182086679,
|
|
"learning_rate": 5.128205128205128e-06,
|
|
"loss": 1.508,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.0768,
|
|
"grad_norm": 2.9047254740447,
|
|
"learning_rate": 6.153846153846155e-06,
|
|
"loss": 1.4687,
|
|
"step": 6
|
|
},
|
|
{
|
|
"epoch": 0.0896,
|
|
"grad_norm": 2.5469693991424136,
|
|
"learning_rate": 7.17948717948718e-06,
|
|
"loss": 1.4354,
|
|
"step": 7
|
|
},
|
|
{
|
|
"epoch": 0.1024,
|
|
"grad_norm": 5.3394378940218505,
|
|
"learning_rate": 8.205128205128205e-06,
|
|
"loss": 1.448,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.1152,
|
|
"grad_norm": 5.237578063138974,
|
|
"learning_rate": 9.230769230769232e-06,
|
|
"loss": 1.4113,
|
|
"step": 9
|
|
},
|
|
{
|
|
"epoch": 0.128,
|
|
"grad_norm": 4.878229367796371,
|
|
"learning_rate": 1.0256410256410256e-05,
|
|
"loss": 1.3674,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.1408,
|
|
"grad_norm": 4.389136620576146,
|
|
"learning_rate": 1.1282051282051283e-05,
|
|
"loss": 1.3963,
|
|
"step": 11
|
|
},
|
|
{
|
|
"epoch": 0.1536,
|
|
"grad_norm": 3.5209608092779057,
|
|
"learning_rate": 1.230769230769231e-05,
|
|
"loss": 1.3873,
|
|
"step": 12
|
|
},
|
|
{
|
|
"epoch": 0.1664,
|
|
"grad_norm": 2.556831853477579,
|
|
"learning_rate": 1.3333333333333333e-05,
|
|
"loss": 1.3816,
|
|
"step": 13
|
|
},
|
|
{
|
|
"epoch": 0.1792,
|
|
"grad_norm": 1.9791181807597111,
|
|
"learning_rate": 1.435897435897436e-05,
|
|
"loss": 1.2973,
|
|
"step": 14
|
|
},
|
|
{
|
|
"epoch": 0.192,
|
|
"grad_norm": 2.0496832639611804,
|
|
"learning_rate": 1.5384615384615387e-05,
|
|
"loss": 1.3467,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.2048,
|
|
"grad_norm": 1.846599835683718,
|
|
"learning_rate": 1.641025641025641e-05,
|
|
"loss": 1.281,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.2176,
|
|
"grad_norm": 1.2550930412415549,
|
|
"learning_rate": 1.7435897435897438e-05,
|
|
"loss": 1.246,
|
|
"step": 17
|
|
},
|
|
{
|
|
"epoch": 0.2304,
|
|
"grad_norm": 1.3283067811292284,
|
|
"learning_rate": 1.8461538461538465e-05,
|
|
"loss": 1.2626,
|
|
"step": 18
|
|
},
|
|
{
|
|
"epoch": 0.2432,
|
|
"grad_norm": 39.870803667289174,
|
|
"learning_rate": 1.9487179487179488e-05,
|
|
"loss": 1.2497,
|
|
"step": 19
|
|
},
|
|
{
|
|
"epoch": 0.256,
|
|
"grad_norm": 2.5286555965354727,
|
|
"learning_rate": 2.0512820512820512e-05,
|
|
"loss": 1.2911,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.2688,
|
|
"grad_norm": 1.714418881557214,
|
|
"learning_rate": 2.153846153846154e-05,
|
|
"loss": 1.2629,
|
|
"step": 21
|
|
},
|
|
{
|
|
"epoch": 0.2816,
|
|
"grad_norm": 1.7498893430574376,
|
|
"learning_rate": 2.2564102564102566e-05,
|
|
"loss": 1.2377,
|
|
"step": 22
|
|
},
|
|
{
|
|
"epoch": 0.2944,
|
|
"grad_norm": 1.3309715455897657,
|
|
"learning_rate": 2.3589743589743593e-05,
|
|
"loss": 1.2138,
|
|
"step": 23
|
|
},
|
|
{
|
|
"epoch": 0.3072,
|
|
"grad_norm": 1.52541650624061,
|
|
"learning_rate": 2.461538461538462e-05,
|
|
"loss": 1.2404,
|
|
"step": 24
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 1.2104236338613845,
|
|
"learning_rate": 2.5641025641025646e-05,
|
|
"loss": 1.2288,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.3328,
|
|
"grad_norm": 1.3973043387015034,
|
|
"learning_rate": 2.6666666666666667e-05,
|
|
"loss": 1.2305,
|
|
"step": 26
|
|
},
|
|
{
|
|
"epoch": 0.3456,
|
|
"grad_norm": 1.1078267886615953,
|
|
"learning_rate": 2.7692307692307694e-05,
|
|
"loss": 1.2068,
|
|
"step": 27
|
|
},
|
|
{
|
|
"epoch": 0.3584,
|
|
"grad_norm": 1.026639796736863,
|
|
"learning_rate": 2.871794871794872e-05,
|
|
"loss": 1.2025,
|
|
"step": 28
|
|
},
|
|
{
|
|
"epoch": 0.3712,
|
|
"grad_norm": 0.9799675004987618,
|
|
"learning_rate": 2.9743589743589747e-05,
|
|
"loss": 1.2084,
|
|
"step": 29
|
|
},
|
|
{
|
|
"epoch": 0.384,
|
|
"grad_norm": 0.9046839423459286,
|
|
"learning_rate": 3.0769230769230774e-05,
|
|
"loss": 1.1879,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.3968,
|
|
"grad_norm": 1.1989336642506447,
|
|
"learning_rate": 3.1794871794871795e-05,
|
|
"loss": 1.1922,
|
|
"step": 31
|
|
},
|
|
{
|
|
"epoch": 0.4096,
|
|
"grad_norm": 1.0124085415657604,
|
|
"learning_rate": 3.282051282051282e-05,
|
|
"loss": 1.1835,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.4224,
|
|
"grad_norm": 0.9033624835929998,
|
|
"learning_rate": 3.384615384615385e-05,
|
|
"loss": 1.1759,
|
|
"step": 33
|
|
},
|
|
{
|
|
"epoch": 0.4352,
|
|
"grad_norm": 1.2517048919372702,
|
|
"learning_rate": 3.4871794871794875e-05,
|
|
"loss": 1.1863,
|
|
"step": 34
|
|
},
|
|
{
|
|
"epoch": 0.448,
|
|
"grad_norm": 0.9796274817199719,
|
|
"learning_rate": 3.58974358974359e-05,
|
|
"loss": 1.2153,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.4608,
|
|
"grad_norm": 1.0512758150854558,
|
|
"learning_rate": 3.692307692307693e-05,
|
|
"loss": 1.1459,
|
|
"step": 36
|
|
},
|
|
{
|
|
"epoch": 0.4736,
|
|
"grad_norm": 1.1258623830729644,
|
|
"learning_rate": 3.794871794871795e-05,
|
|
"loss": 1.1484,
|
|
"step": 37
|
|
},
|
|
{
|
|
"epoch": 0.4864,
|
|
"grad_norm": 1.0547902344575502,
|
|
"learning_rate": 3.8974358974358976e-05,
|
|
"loss": 1.1469,
|
|
"step": 38
|
|
},
|
|
{
|
|
"epoch": 0.4992,
|
|
"grad_norm": 0.9916194724908646,
|
|
"learning_rate": 4e-05,
|
|
"loss": 1.1893,
|
|
"step": 39
|
|
},
|
|
{
|
|
"epoch": 0.512,
|
|
"grad_norm": 1.3302384012766393,
|
|
"learning_rate": 3.9999198907597046e-05,
|
|
"loss": 1.15,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.5248,
|
|
"grad_norm": 1.963593689910703,
|
|
"learning_rate": 3.9996795694563096e-05,
|
|
"loss": 1.1328,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 0.5376,
|
|
"grad_norm": 1.1715690222051025,
|
|
"learning_rate": 3.999279055341771e-05,
|
|
"loss": 1.1853,
|
|
"step": 42
|
|
},
|
|
{
|
|
"epoch": 0.5504,
|
|
"grad_norm": 2.705845620833423,
|
|
"learning_rate": 3.998718380500971e-05,
|
|
"loss": 1.1834,
|
|
"step": 43
|
|
},
|
|
{
|
|
"epoch": 0.5632,
|
|
"grad_norm": 1.9066020723868575,
|
|
"learning_rate": 3.997997589849145e-05,
|
|
"loss": 1.1458,
|
|
"step": 44
|
|
},
|
|
{
|
|
"epoch": 0.576,
|
|
"grad_norm": 2.423499618492123,
|
|
"learning_rate": 3.9971167411282835e-05,
|
|
"loss": 1.1943,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.5888,
|
|
"grad_norm": 1.7171312136021033,
|
|
"learning_rate": 3.99607590490251e-05,
|
|
"loss": 1.1983,
|
|
"step": 46
|
|
},
|
|
{
|
|
"epoch": 0.6016,
|
|
"grad_norm": 1.6842292096597014,
|
|
"learning_rate": 3.9948751645524235e-05,
|
|
"loss": 1.1626,
|
|
"step": 47
|
|
},
|
|
{
|
|
"epoch": 0.6144,
|
|
"grad_norm": 1.7529623346658405,
|
|
"learning_rate": 3.9935146162684206e-05,
|
|
"loss": 1.1529,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 0.6272,
|
|
"grad_norm": 1.5913373949703382,
|
|
"learning_rate": 3.9919943690429906e-05,
|
|
"loss": 1.1494,
|
|
"step": 49
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 1.6507604461840728,
|
|
"learning_rate": 3.9903145446619837e-05,
|
|
"loss": 1.157,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.6528,
|
|
"grad_norm": 1.5140962844534174,
|
|
"learning_rate": 3.9884752776948564e-05,
|
|
"loss": 1.1729,
|
|
"step": 51
|
|
},
|
|
{
|
|
"epoch": 0.6656,
|
|
"grad_norm": 1.4554166804065727,
|
|
"learning_rate": 3.9864767154838864e-05,
|
|
"loss": 1.1861,
|
|
"step": 52
|
|
},
|
|
{
|
|
"epoch": 0.6784,
|
|
"grad_norm": 1.4204636015581693,
|
|
"learning_rate": 3.9843190181323744e-05,
|
|
"loss": 1.1113,
|
|
"step": 53
|
|
},
|
|
{
|
|
"epoch": 0.6912,
|
|
"grad_norm": 1.2192992190178875,
|
|
"learning_rate": 3.982002358491817e-05,
|
|
"loss": 1.149,
|
|
"step": 54
|
|
},
|
|
{
|
|
"epoch": 0.704,
|
|
"grad_norm": 1.356919138180013,
|
|
"learning_rate": 3.979526922148058e-05,
|
|
"loss": 1.1464,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.7168,
|
|
"grad_norm": 1.1783179174778675,
|
|
"learning_rate": 3.9768929074064206e-05,
|
|
"loss": 1.1468,
|
|
"step": 56
|
|
},
|
|
{
|
|
"epoch": 0.7296,
|
|
"grad_norm": 1.520213069204608,
|
|
"learning_rate": 3.9741005252758255e-05,
|
|
"loss": 1.1694,
|
|
"step": 57
|
|
},
|
|
{
|
|
"epoch": 0.7424,
|
|
"grad_norm": 1.0871527275379298,
|
|
"learning_rate": 3.971149999451886e-05,
|
|
"loss": 1.1939,
|
|
"step": 58
|
|
},
|
|
{
|
|
"epoch": 0.7552,
|
|
"grad_norm": 1.3271862854340502,
|
|
"learning_rate": 3.9680415662989806e-05,
|
|
"loss": 1.1963,
|
|
"step": 59
|
|
},
|
|
{
|
|
"epoch": 0.768,
|
|
"grad_norm": 1.1360261466381845,
|
|
"learning_rate": 3.9647754748313294e-05,
|
|
"loss": 1.1659,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.7808,
|
|
"grad_norm": 1.0465126532970044,
|
|
"learning_rate": 3.96135198669304e-05,
|
|
"loss": 1.135,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 0.7936,
|
|
"grad_norm": 1.104046357543538,
|
|
"learning_rate": 3.957771376137144e-05,
|
|
"loss": 1.1045,
|
|
"step": 62
|
|
},
|
|
{
|
|
"epoch": 0.8064,
|
|
"grad_norm": 1.0076419389768252,
|
|
"learning_rate": 3.954033930003634e-05,
|
|
"loss": 1.1306,
|
|
"step": 63
|
|
},
|
|
{
|
|
"epoch": 0.8192,
|
|
"grad_norm": 1.1595281316865143,
|
|
"learning_rate": 3.9501399476964806e-05,
|
|
"loss": 1.1248,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.832,
|
|
"grad_norm": 0.8082273276318717,
|
|
"learning_rate": 3.946089741159648e-05,
|
|
"loss": 1.1091,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.8448,
|
|
"grad_norm": 1.1783712398913984,
|
|
"learning_rate": 3.9418836348521045e-05,
|
|
"loss": 1.1273,
|
|
"step": 66
|
|
},
|
|
{
|
|
"epoch": 0.8576,
|
|
"grad_norm": 1.0881445998587171,
|
|
"learning_rate": 3.937521965721831e-05,
|
|
"loss": 1.1285,
|
|
"step": 67
|
|
},
|
|
{
|
|
"epoch": 0.8704,
|
|
"grad_norm": 1.068066620388015,
|
|
"learning_rate": 3.933005083178828e-05,
|
|
"loss": 1.1028,
|
|
"step": 68
|
|
},
|
|
{
|
|
"epoch": 0.8832,
|
|
"grad_norm": 0.9707816485302323,
|
|
"learning_rate": 3.928333349067125e-05,
|
|
"loss": 1.139,
|
|
"step": 69
|
|
},
|
|
{
|
|
"epoch": 0.896,
|
|
"grad_norm": 1.213728444765552,
|
|
"learning_rate": 3.923507137635792e-05,
|
|
"loss": 1.1002,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.9088,
|
|
"grad_norm": 0.8551346906494776,
|
|
"learning_rate": 3.9185268355089606e-05,
|
|
"loss": 1.1475,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 0.9216,
|
|
"grad_norm": 0.9204857406831569,
|
|
"learning_rate": 3.913392841654851e-05,
|
|
"loss": 1.141,
|
|
"step": 72
|
|
},
|
|
{
|
|
"epoch": 0.9344,
|
|
"grad_norm": 0.7091199038237178,
|
|
"learning_rate": 3.9081055673538093e-05,
|
|
"loss": 1.1598,
|
|
"step": 73
|
|
},
|
|
{
|
|
"epoch": 0.9472,
|
|
"grad_norm": 0.7199232905668526,
|
|
"learning_rate": 3.902665436165364e-05,
|
|
"loss": 1.1478,
|
|
"step": 74
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.7892379678981329,
|
|
"learning_rate": 3.897072883894291e-05,
|
|
"loss": 1.0939,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.9728,
|
|
"grad_norm": 0.6539783456455385,
|
|
"learning_rate": 3.8913283585557054e-05,
|
|
"loss": 1.1523,
|
|
"step": 76
|
|
},
|
|
{
|
|
"epoch": 0.9856,
|
|
"grad_norm": 1.1995844140662193,
|
|
"learning_rate": 3.885432320339167e-05,
|
|
"loss": 1.1584,
|
|
"step": 77
|
|
},
|
|
{
|
|
"epoch": 0.9984,
|
|
"grad_norm": 0.7584414357624537,
|
|
"learning_rate": 3.879385241571817e-05,
|
|
"loss": 1.1579,
|
|
"step": 78
|
|
},
|
|
{
|
|
"epoch": 1.0112,
|
|
"grad_norm": 0.9838022924748918,
|
|
"learning_rate": 3.873187606680543e-05,
|
|
"loss": 1.0571,
|
|
"step": 79
|
|
},
|
|
{
|
|
"epoch": 1.024,
|
|
"grad_norm": 0.7417507430304172,
|
|
"learning_rate": 3.866839912153168e-05,
|
|
"loss": 1.0308,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 1.0368,
|
|
"grad_norm": 0.7466250162315038,
|
|
"learning_rate": 3.860342666498677e-05,
|
|
"loss": 1.0528,
|
|
"step": 81
|
|
},
|
|
{
|
|
"epoch": 1.0496,
|
|
"grad_norm": 0.9225979437233178,
|
|
"learning_rate": 3.853696390206484e-05,
|
|
"loss": 1.0203,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 1.0624,
|
|
"grad_norm": 0.948559988635711,
|
|
"learning_rate": 3.846901615704734e-05,
|
|
"loss": 1.0205,
|
|
"step": 83
|
|
},
|
|
{
|
|
"epoch": 1.0752,
|
|
"grad_norm": 0.7814874004462734,
|
|
"learning_rate": 3.839958887317649e-05,
|
|
"loss": 1.0338,
|
|
"step": 84
|
|
},
|
|
{
|
|
"epoch": 1.088,
|
|
"grad_norm": 1.0628469023322726,
|
|
"learning_rate": 3.832868761221926e-05,
|
|
"loss": 1.0165,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 1.1008,
|
|
"grad_norm": 0.8615739732498551,
|
|
"learning_rate": 3.825631805402182e-05,
|
|
"loss": 1.0115,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 1.1136,
|
|
"grad_norm": 1.060468292175149,
|
|
"learning_rate": 3.818248599605448e-05,
|
|
"loss": 1.0427,
|
|
"step": 87
|
|
},
|
|
{
|
|
"epoch": 1.1264,
|
|
"grad_norm": 0.7379667374485601,
|
|
"learning_rate": 3.810719735294731e-05,
|
|
"loss": 1.0209,
|
|
"step": 88
|
|
},
|
|
{
|
|
"epoch": 1.1392,
|
|
"grad_norm": 0.9750295810079974,
|
|
"learning_rate": 3.8030458156016326e-05,
|
|
"loss": 1.0251,
|
|
"step": 89
|
|
},
|
|
{
|
|
"epoch": 1.152,
|
|
"grad_norm": 0.7206293262576626,
|
|
"learning_rate": 3.795227455278029e-05,
|
|
"loss": 1.0148,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 1.1648,
|
|
"grad_norm": 0.6855434098356388,
|
|
"learning_rate": 3.787265280646825e-05,
|
|
"loss": 1.0425,
|
|
"step": 91
|
|
},
|
|
{
|
|
"epoch": 1.1776,
|
|
"grad_norm": 0.7383257680403411,
|
|
"learning_rate": 3.7791599295517825e-05,
|
|
"loss": 1.0261,
|
|
"step": 92
|
|
},
|
|
{
|
|
"epoch": 1.1904,
|
|
"grad_norm": 0.6653630981167292,
|
|
"learning_rate": 3.7709120513064196e-05,
|
|
"loss": 1.0227,
|
|
"step": 93
|
|
},
|
|
{
|
|
"epoch": 1.2032,
|
|
"grad_norm": 0.7475014250369564,
|
|
"learning_rate": 3.762522306641998e-05,
|
|
"loss": 1.0252,
|
|
"step": 94
|
|
},
|
|
{
|
|
"epoch": 1.216,
|
|
"grad_norm": 0.7639763136720108,
|
|
"learning_rate": 3.7539913676545874e-05,
|
|
"loss": 1.0313,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 1.2288000000000001,
|
|
"grad_norm": 0.8071040559697056,
|
|
"learning_rate": 3.745319917751229e-05,
|
|
"loss": 1.031,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 1.2416,
|
|
"grad_norm": 0.8100115595913084,
|
|
"learning_rate": 3.736508651595188e-05,
|
|
"loss": 1.0342,
|
|
"step": 97
|
|
},
|
|
{
|
|
"epoch": 1.2544,
|
|
"grad_norm": 0.8707679082472297,
|
|
"learning_rate": 3.727558275050301e-05,
|
|
"loss": 1.0101,
|
|
"step": 98
|
|
},
|
|
{
|
|
"epoch": 1.2671999999999999,
|
|
"grad_norm": 0.7153434835330477,
|
|
"learning_rate": 3.718469505124434e-05,
|
|
"loss": 1.0489,
|
|
"step": 99
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"grad_norm": 1.0870536021354325,
|
|
"learning_rate": 3.709243069912041e-05,
|
|
"loss": 1.022,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 1.2928,
|
|
"grad_norm": 0.7019182618113695,
|
|
"learning_rate": 3.699879708535838e-05,
|
|
"loss": 1.0637,
|
|
"step": 101
|
|
},
|
|
{
|
|
"epoch": 1.3056,
|
|
"grad_norm": 1.1038428470341881,
|
|
"learning_rate": 3.69038017108759e-05,
|
|
"loss": 1.0546,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 1.3184,
|
|
"grad_norm": 0.7627783095736863,
|
|
"learning_rate": 3.680745218568026e-05,
|
|
"loss": 1.0747,
|
|
"step": 103
|
|
},
|
|
{
|
|
"epoch": 1.3312,
|
|
"grad_norm": 0.8547971875954213,
|
|
"learning_rate": 3.6709756228258735e-05,
|
|
"loss": 1.0375,
|
|
"step": 104
|
|
},
|
|
{
|
|
"epoch": 1.3439999999999999,
|
|
"grad_norm": 0.6853435289526536,
|
|
"learning_rate": 3.6610721664960236e-05,
|
|
"loss": 1.0611,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 1.3568,
|
|
"grad_norm": 0.7396875892641765,
|
|
"learning_rate": 3.65103564293684e-05,
|
|
"loss": 1.0217,
|
|
"step": 106
|
|
},
|
|
{
|
|
"epoch": 1.3696,
|
|
"grad_norm": 0.7146716921415344,
|
|
"learning_rate": 3.640866856166601e-05,
|
|
"loss": 1.0274,
|
|
"step": 107
|
|
},
|
|
{
|
|
"epoch": 1.3824,
|
|
"grad_norm": 0.6443857661088517,
|
|
"learning_rate": 3.6305666207990886e-05,
|
|
"loss": 1.0363,
|
|
"step": 108
|
|
},
|
|
{
|
|
"epoch": 1.3952,
|
|
"grad_norm": 0.7944296996171112,
|
|
"learning_rate": 3.6201357619783336e-05,
|
|
"loss": 1.0105,
|
|
"step": 109
|
|
},
|
|
{
|
|
"epoch": 1.408,
|
|
"grad_norm": 0.7727329978152716,
|
|
"learning_rate": 3.609575115312511e-05,
|
|
"loss": 1.0696,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 1.4208,
|
|
"grad_norm": 0.725530680479781,
|
|
"learning_rate": 3.598885526807003e-05,
|
|
"loss": 1.0446,
|
|
"step": 111
|
|
},
|
|
{
|
|
"epoch": 1.4336,
|
|
"grad_norm": 0.8330011282670016,
|
|
"learning_rate": 3.5880678527966224e-05,
|
|
"loss": 1.059,
|
|
"step": 112
|
|
},
|
|
{
|
|
"epoch": 1.4464000000000001,
|
|
"grad_norm": 0.6869528205827169,
|
|
"learning_rate": 3.577122959877017e-05,
|
|
"loss": 1.0335,
|
|
"step": 113
|
|
},
|
|
{
|
|
"epoch": 1.4592,
|
|
"grad_norm": 0.8217643443783395,
|
|
"learning_rate": 3.566051724835245e-05,
|
|
"loss": 1.065,
|
|
"step": 114
|
|
},
|
|
{
|
|
"epoch": 1.472,
|
|
"grad_norm": 0.6722402567252438,
|
|
"learning_rate": 3.554855034579532e-05,
|
|
"loss": 1.006,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 1.4848,
|
|
"grad_norm": 0.6930068885172014,
|
|
"learning_rate": 3.5435337860682304e-05,
|
|
"loss": 1.0279,
|
|
"step": 116
|
|
},
|
|
{
|
|
"epoch": 1.4976,
|
|
"grad_norm": 0.7143189571312937,
|
|
"learning_rate": 3.532088886237956e-05,
|
|
"loss": 1.0442,
|
|
"step": 117
|
|
},
|
|
{
|
|
"epoch": 1.5104,
|
|
"grad_norm": 0.7110452324914206,
|
|
"learning_rate": 3.520521251930941e-05,
|
|
"loss": 1.0209,
|
|
"step": 118
|
|
},
|
|
{
|
|
"epoch": 1.5232,
|
|
"grad_norm": 0.5656132958430365,
|
|
"learning_rate": 3.5088318098215805e-05,
|
|
"loss": 1.0614,
|
|
"step": 119
|
|
},
|
|
{
|
|
"epoch": 1.536,
|
|
"grad_norm": 0.6375980220575151,
|
|
"learning_rate": 3.497021496342203e-05,
|
|
"loss": 1.0298,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 1.5488,
|
|
"grad_norm": 0.6237522040232807,
|
|
"learning_rate": 3.485091257608047e-05,
|
|
"loss": 1.0221,
|
|
"step": 121
|
|
},
|
|
{
|
|
"epoch": 1.5615999999999999,
|
|
"grad_norm": 0.7275575334218716,
|
|
"learning_rate": 3.473042049341474e-05,
|
|
"loss": 0.9958,
|
|
"step": 122
|
|
},
|
|
{
|
|
"epoch": 1.5744,
|
|
"grad_norm": 0.5772814699418052,
|
|
"learning_rate": 3.4608748367954064e-05,
|
|
"loss": 1.042,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 1.5872000000000002,
|
|
"grad_norm": 0.7294639491913409,
|
|
"learning_rate": 3.4485905946759965e-05,
|
|
"loss": 1.005,
|
|
"step": 124
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"grad_norm": 0.6949491354569357,
|
|
"learning_rate": 3.4361903070645484e-05,
|
|
"loss": 1.0019,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 1.6128,
|
|
"grad_norm": 0.6372941005851039,
|
|
"learning_rate": 3.423674967338681e-05,
|
|
"loss": 1.0474,
|
|
"step": 126
|
|
},
|
|
{
|
|
"epoch": 1.6256,
|
|
"grad_norm": 0.6336897873068353,
|
|
"learning_rate": 3.411045578092754e-05,
|
|
"loss": 1.0099,
|
|
"step": 127
|
|
},
|
|
{
|
|
"epoch": 1.6383999999999999,
|
|
"grad_norm": 0.6741231455881623,
|
|
"learning_rate": 3.398303151057543e-05,
|
|
"loss": 1.0727,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 1.6512,
|
|
"grad_norm": 0.6003120067299178,
|
|
"learning_rate": 3.385448707019199e-05,
|
|
"loss": 1.0243,
|
|
"step": 129
|
|
},
|
|
{
|
|
"epoch": 1.6640000000000001,
|
|
"grad_norm": 0.6187596067016436,
|
|
"learning_rate": 3.372483275737468e-05,
|
|
"loss": 0.9836,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 1.6768,
|
|
"grad_norm": 0.6929089905088957,
|
|
"learning_rate": 3.359407895863199e-05,
|
|
"loss": 1.0474,
|
|
"step": 131
|
|
},
|
|
{
|
|
"epoch": 1.6896,
|
|
"grad_norm": 0.6233898973907259,
|
|
"learning_rate": 3.34622361485514e-05,
|
|
"loss": 1.0421,
|
|
"step": 132
|
|
},
|
|
{
|
|
"epoch": 1.7024,
|
|
"grad_norm": 0.6612632537378574,
|
|
"learning_rate": 3.332931488896029e-05,
|
|
"loss": 1.05,
|
|
"step": 133
|
|
},
|
|
{
|
|
"epoch": 1.7151999999999998,
|
|
"grad_norm": 0.6619560427219947,
|
|
"learning_rate": 3.319532582807977e-05,
|
|
"loss": 1.0253,
|
|
"step": 134
|
|
},
|
|
{
|
|
"epoch": 1.728,
|
|
"grad_norm": 0.6888633993579613,
|
|
"learning_rate": 3.30602796996717e-05,
|
|
"loss": 1.0422,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 1.7408000000000001,
|
|
"grad_norm": 0.738563107538913,
|
|
"learning_rate": 3.2924187322178865e-05,
|
|
"loss": 1.0177,
|
|
"step": 136
|
|
},
|
|
{
|
|
"epoch": 1.7536,
|
|
"grad_norm": 0.6960055265279931,
|
|
"learning_rate": 3.278705959785821e-05,
|
|
"loss": 1.0394,
|
|
"step": 137
|
|
},
|
|
{
|
|
"epoch": 1.7664,
|
|
"grad_norm": 0.5909931615279079,
|
|
"learning_rate": 3.2648907511907544e-05,
|
|
"loss": 1.024,
|
|
"step": 138
|
|
},
|
|
{
|
|
"epoch": 1.7792,
|
|
"grad_norm": 0.7575500923597764,
|
|
"learning_rate": 3.250974213158555e-05,
|
|
"loss": 1.0111,
|
|
"step": 139
|
|
},
|
|
{
|
|
"epoch": 1.792,
|
|
"grad_norm": 0.6900873453498054,
|
|
"learning_rate": 3.23695746053251e-05,
|
|
"loss": 0.9901,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 1.8048,
|
|
"grad_norm": 0.8149304511025252,
|
|
"learning_rate": 3.222841616184025e-05,
|
|
"loss": 1.0533,
|
|
"step": 141
|
|
},
|
|
{
|
|
"epoch": 1.8176,
|
|
"grad_norm": 0.6306865685658272,
|
|
"learning_rate": 3.208627810922665e-05,
|
|
"loss": 1.0669,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 1.8304,
|
|
"grad_norm": 0.7790307023573653,
|
|
"learning_rate": 3.194317183405573e-05,
|
|
"loss": 0.9949,
|
|
"step": 143
|
|
},
|
|
{
|
|
"epoch": 1.8432,
|
|
"grad_norm": 0.956777617494605,
|
|
"learning_rate": 3.1799108800462466e-05,
|
|
"loss": 1.0575,
|
|
"step": 144
|
|
},
|
|
{
|
|
"epoch": 1.8559999999999999,
|
|
"grad_norm": 0.7101708684119322,
|
|
"learning_rate": 3.1654100549227024e-05,
|
|
"loss": 1.0161,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 1.8688,
|
|
"grad_norm": 0.6667058828503257,
|
|
"learning_rate": 3.1508158696850275e-05,
|
|
"loss": 1.021,
|
|
"step": 146
|
|
},
|
|
{
|
|
"epoch": 1.8816000000000002,
|
|
"grad_norm": 0.6841101091301254,
|
|
"learning_rate": 3.136129493462312e-05,
|
|
"loss": 1.0205,
|
|
"step": 147
|
|
},
|
|
{
|
|
"epoch": 1.8944,
|
|
"grad_norm": 0.7362428220666088,
|
|
"learning_rate": 3.121352102768998e-05,
|
|
"loss": 1.0128,
|
|
"step": 148
|
|
},
|
|
{
|
|
"epoch": 1.9072,
|
|
"grad_norm": 0.931468433753366,
|
|
"learning_rate": 3.106484881410628e-05,
|
|
"loss": 1.0624,
|
|
"step": 149
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"grad_norm": 0.6514868217300596,
|
|
"learning_rate": 3.091529020389009e-05,
|
|
"loss": 1.045,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 1.9327999999999999,
|
|
"grad_norm": 0.6737780389378739,
|
|
"learning_rate": 3.076485717806808e-05,
|
|
"loss": 1.0603,
|
|
"step": 151
|
|
},
|
|
{
|
|
"epoch": 1.9456,
|
|
"grad_norm": 0.7971764922901946,
|
|
"learning_rate": 3.061356178771564e-05,
|
|
"loss": 1.047,
|
|
"step": 152
|
|
},
|
|
{
|
|
"epoch": 1.9584000000000001,
|
|
"grad_norm": 0.6672968881704991,
|
|
"learning_rate": 3.0461416152991555e-05,
|
|
"loss": 1.0348,
|
|
"step": 153
|
|
},
|
|
{
|
|
"epoch": 1.9712,
|
|
"grad_norm": 0.6285599621316675,
|
|
"learning_rate": 3.0308432462167045e-05,
|
|
"loss": 1.0388,
|
|
"step": 154
|
|
},
|
|
{
|
|
"epoch": 1.984,
|
|
"grad_norm": 0.6323443787495369,
|
|
"learning_rate": 3.015462297064936e-05,
|
|
"loss": 1.0082,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 1.9968,
|
|
"grad_norm": 0.6282469553541248,
|
|
"learning_rate": 3.0000000000000004e-05,
|
|
"loss": 1.0165,
|
|
"step": 156
|
|
},
|
|
{
|
|
"epoch": 2.0096,
|
|
"grad_norm": 0.7920449181589374,
|
|
"learning_rate": 2.98445759369477e-05,
|
|
"loss": 0.9535,
|
|
"step": 157
|
|
},
|
|
{
|
|
"epoch": 2.0224,
|
|
"grad_norm": 0.9354237907431909,
|
|
"learning_rate": 2.9688363232396056e-05,
|
|
"loss": 0.8855,
|
|
"step": 158
|
|
},
|
|
{
|
|
"epoch": 2.0352,
|
|
"grad_norm": 0.7393414908433603,
|
|
"learning_rate": 2.9531374400426158e-05,
|
|
"loss": 0.9174,
|
|
"step": 159
|
|
},
|
|
{
|
|
"epoch": 2.048,
|
|
"grad_norm": 0.7456456430643493,
|
|
"learning_rate": 2.9373622017294075e-05,
|
|
"loss": 0.9103,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 2.0608,
|
|
"grad_norm": 0.7289060958205632,
|
|
"learning_rate": 2.9215118720423375e-05,
|
|
"loss": 0.8442,
|
|
"step": 161
|
|
},
|
|
{
|
|
"epoch": 2.0736,
|
|
"grad_norm": 0.7886441991873184,
|
|
"learning_rate": 2.9055877207392752e-05,
|
|
"loss": 0.8601,
|
|
"step": 162
|
|
},
|
|
{
|
|
"epoch": 2.0864,
|
|
"grad_norm": 0.7164863165828993,
|
|
"learning_rate": 2.8895910234918828e-05,
|
|
"loss": 0.8757,
|
|
"step": 163
|
|
},
|
|
{
|
|
"epoch": 2.0992,
|
|
"grad_norm": 1.0509175010011313,
|
|
"learning_rate": 2.873523061783426e-05,
|
|
"loss": 0.8921,
|
|
"step": 164
|
|
},
|
|
{
|
|
"epoch": 2.112,
|
|
"grad_norm": 0.7083554410944634,
|
|
"learning_rate": 2.8573851228061084e-05,
|
|
"loss": 0.8993,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 2.1248,
|
|
"grad_norm": 0.8584638658175244,
|
|
"learning_rate": 2.8411784993579633e-05,
|
|
"loss": 0.8283,
|
|
"step": 166
|
|
},
|
|
{
|
|
"epoch": 2.1376,
|
|
"grad_norm": 0.8694855263573216,
|
|
"learning_rate": 2.8249044897392814e-05,
|
|
"loss": 0.9054,
|
|
"step": 167
|
|
},
|
|
{
|
|
"epoch": 2.1504,
|
|
"grad_norm": 0.7269472881341665,
|
|
"learning_rate": 2.80856439764861e-05,
|
|
"loss": 0.8712,
|
|
"step": 168
|
|
},
|
|
{
|
|
"epoch": 2.1632,
|
|
"grad_norm": 0.76003993102072,
|
|
"learning_rate": 2.792159532078314e-05,
|
|
"loss": 0.8566,
|
|
"step": 169
|
|
},
|
|
{
|
|
"epoch": 2.176,
|
|
"grad_norm": 0.8819665165882664,
|
|
"learning_rate": 2.77569120720971e-05,
|
|
"loss": 0.888,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 2.1888,
|
|
"grad_norm": 0.8216730545600411,
|
|
"learning_rate": 2.7591607423077932e-05,
|
|
"loss": 0.8974,
|
|
"step": 171
|
|
},
|
|
{
|
|
"epoch": 2.2016,
|
|
"grad_norm": 0.8543194781416534,
|
|
"learning_rate": 2.7425694616155474e-05,
|
|
"loss": 0.8638,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 2.2144,
|
|
"grad_norm": 0.9629014774410137,
|
|
"learning_rate": 2.7259186942478656e-05,
|
|
"loss": 0.902,
|
|
"step": 173
|
|
},
|
|
{
|
|
"epoch": 2.2272,
|
|
"grad_norm": 0.9390773048771573,
|
|
"learning_rate": 2.7092097740850712e-05,
|
|
"loss": 0.9118,
|
|
"step": 174
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"grad_norm": 0.6586425638342351,
|
|
"learning_rate": 2.692444039666066e-05,
|
|
"loss": 0.8978,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 2.2528,
|
|
"grad_norm": 0.8543406916199504,
|
|
"learning_rate": 2.6756228340810946e-05,
|
|
"loss": 0.8812,
|
|
"step": 176
|
|
},
|
|
{
|
|
"epoch": 2.2656,
|
|
"grad_norm": 0.5996784440113992,
|
|
"learning_rate": 2.6587475048641596e-05,
|
|
"loss": 0.881,
|
|
"step": 177
|
|
},
|
|
{
|
|
"epoch": 2.2784,
|
|
"grad_norm": 0.6456129989687627,
|
|
"learning_rate": 2.6418194038850634e-05,
|
|
"loss": 0.8689,
|
|
"step": 178
|
|
},
|
|
{
|
|
"epoch": 2.2912,
|
|
"grad_norm": 0.5918832046032291,
|
|
"learning_rate": 2.624839887241115e-05,
|
|
"loss": 0.8778,
|
|
"step": 179
|
|
},
|
|
{
|
|
"epoch": 2.304,
|
|
"grad_norm": 0.6766626196792827,
|
|
"learning_rate": 2.607810315148494e-05,
|
|
"loss": 0.9127,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 2.3168,
|
|
"grad_norm": 0.7642068892955837,
|
|
"learning_rate": 2.5907320518332827e-05,
|
|
"loss": 0.8929,
|
|
"step": 181
|
|
},
|
|
{
|
|
"epoch": 2.3296,
|
|
"grad_norm": 0.6765160458989914,
|
|
"learning_rate": 2.5736064654221808e-05,
|
|
"loss": 0.887,
|
|
"step": 182
|
|
},
|
|
{
|
|
"epoch": 2.3424,
|
|
"grad_norm": 0.6298926463574855,
|
|
"learning_rate": 2.5564349278329056e-05,
|
|
"loss": 0.8847,
|
|
"step": 183
|
|
},
|
|
{
|
|
"epoch": 2.3552,
|
|
"grad_norm": 0.7069104046493812,
|
|
"learning_rate": 2.539218814664288e-05,
|
|
"loss": 0.8564,
|
|
"step": 184
|
|
},
|
|
{
|
|
"epoch": 2.368,
|
|
"grad_norm": 0.6670546688917147,
|
|
"learning_rate": 2.521959505086075e-05,
|
|
"loss": 0.8604,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 2.3808,
|
|
"grad_norm": 0.6163876874249468,
|
|
"learning_rate": 2.5046583817284437e-05,
|
|
"loss": 0.8296,
|
|
"step": 186
|
|
},
|
|
{
|
|
"epoch": 2.3936,
|
|
"grad_norm": 0.6901896113367271,
|
|
"learning_rate": 2.487316830571244e-05,
|
|
"loss": 0.8715,
|
|
"step": 187
|
|
},
|
|
{
|
|
"epoch": 2.4064,
|
|
"grad_norm": 0.6047440816959818,
|
|
"learning_rate": 2.4699362408329646e-05,
|
|
"loss": 0.8767,
|
|
"step": 188
|
|
},
|
|
{
|
|
"epoch": 2.4192,
|
|
"grad_norm": 0.6099555402580358,
|
|
"learning_rate": 2.4525180048594452e-05,
|
|
"loss": 0.8868,
|
|
"step": 189
|
|
},
|
|
{
|
|
"epoch": 2.432,
|
|
"grad_norm": 0.5881337606800934,
|
|
"learning_rate": 2.435063518012335e-05,
|
|
"loss": 0.8933,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 2.4448,
|
|
"grad_norm": 0.6671370070980503,
|
|
"learning_rate": 2.4175741785573177e-05,
|
|
"loss": 0.8951,
|
|
"step": 191
|
|
},
|
|
{
|
|
"epoch": 2.4576000000000002,
|
|
"grad_norm": 0.7155466500803334,
|
|
"learning_rate": 2.4000513875520892e-05,
|
|
"loss": 0.8865,
|
|
"step": 192
|
|
},
|
|
{
|
|
"epoch": 2.4704,
|
|
"grad_norm": 0.5579261507137053,
|
|
"learning_rate": 2.3824965487341247e-05,
|
|
"loss": 0.8794,
|
|
"step": 193
|
|
},
|
|
{
|
|
"epoch": 2.4832,
|
|
"grad_norm": 0.6713643030866266,
|
|
"learning_rate": 2.3649110684082258e-05,
|
|
"loss": 0.9061,
|
|
"step": 194
|
|
},
|
|
{
|
|
"epoch": 2.496,
|
|
"grad_norm": 0.4896026221778859,
|
|
"learning_rate": 2.3472963553338614e-05,
|
|
"loss": 0.8858,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 2.5088,
|
|
"grad_norm": 0.5533236722587165,
|
|
"learning_rate": 2.3296538206123134e-05,
|
|
"loss": 0.8667,
|
|
"step": 196
|
|
},
|
|
{
|
|
"epoch": 2.5216,
|
|
"grad_norm": 0.47090995094313093,
|
|
"learning_rate": 2.311984877573636e-05,
|
|
"loss": 0.8812,
|
|
"step": 197
|
|
},
|
|
{
|
|
"epoch": 2.5343999999999998,
|
|
"grad_norm": 0.4833307508846615,
|
|
"learning_rate": 2.2942909416634326e-05,
|
|
"loss": 0.8887,
|
|
"step": 198
|
|
},
|
|
{
|
|
"epoch": 2.5472,
|
|
"grad_norm": 0.48979876153574464,
|
|
"learning_rate": 2.2765734303294666e-05,
|
|
"loss": 0.882,
|
|
"step": 199
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"grad_norm": 0.5158802399073401,
|
|
"learning_rate": 2.2588337629081107e-05,
|
|
"loss": 0.8557,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 2.5728,
|
|
"grad_norm": 0.5010025084567198,
|
|
"learning_rate": 2.2410733605106462e-05,
|
|
"loss": 0.881,
|
|
"step": 201
|
|
},
|
|
{
|
|
"epoch": 2.5856,
|
|
"grad_norm": 0.5454774120190347,
|
|
"learning_rate": 2.2232936459094158e-05,
|
|
"loss": 0.8844,
|
|
"step": 202
|
|
},
|
|
{
|
|
"epoch": 2.5984,
|
|
"grad_norm": 0.5464916386754389,
|
|
"learning_rate": 2.205496043423849e-05,
|
|
"loss": 0.8555,
|
|
"step": 203
|
|
},
|
|
{
|
|
"epoch": 2.6112,
|
|
"grad_norm": 0.5833801851333361,
|
|
"learning_rate": 2.1876819788063586e-05,
|
|
"loss": 0.9151,
|
|
"step": 204
|
|
},
|
|
{
|
|
"epoch": 2.624,
|
|
"grad_norm": 0.5920875611002528,
|
|
"learning_rate": 2.16985287912813e-05,
|
|
"loss": 0.8532,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 2.6368,
|
|
"grad_norm": 0.5470965335334187,
|
|
"learning_rate": 2.1520101726647922e-05,
|
|
"loss": 0.9498,
|
|
"step": 206
|
|
},
|
|
{
|
|
"epoch": 2.6496,
|
|
"grad_norm": 0.6307772012829592,
|
|
"learning_rate": 2.1341552887820048e-05,
|
|
"loss": 0.8701,
|
|
"step": 207
|
|
},
|
|
{
|
|
"epoch": 2.6624,
|
|
"grad_norm": 0.48514304019394006,
|
|
"learning_rate": 2.1162896578209517e-05,
|
|
"loss": 0.8511,
|
|
"step": 208
|
|
},
|
|
{
|
|
"epoch": 2.6752000000000002,
|
|
"grad_norm": 0.6282389120924936,
|
|
"learning_rate": 2.0984147109837564e-05,
|
|
"loss": 0.8587,
|
|
"step": 209
|
|
},
|
|
{
|
|
"epoch": 2.6879999999999997,
|
|
"grad_norm": 0.44617040466325647,
|
|
"learning_rate": 2.0805318802188307e-05,
|
|
"loss": 0.8273,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 2.7008,
|
|
"grad_norm": 0.5636627707513929,
|
|
"learning_rate": 2.0626425981061608e-05,
|
|
"loss": 0.8766,
|
|
"step": 211
|
|
},
|
|
{
|
|
"epoch": 2.7136,
|
|
"grad_norm": 0.4446555464176329,
|
|
"learning_rate": 2.0447482977425465e-05,
|
|
"loss": 0.8733,
|
|
"step": 212
|
|
},
|
|
{
|
|
"epoch": 2.7264,
|
|
"grad_norm": 0.4456934329218442,
|
|
"learning_rate": 2.0268504126267952e-05,
|
|
"loss": 0.8684,
|
|
"step": 213
|
|
},
|
|
{
|
|
"epoch": 2.7392,
|
|
"grad_norm": 0.5148410533257054,
|
|
"learning_rate": 2.008950376544887e-05,
|
|
"loss": 0.8812,
|
|
"step": 214
|
|
},
|
|
{
|
|
"epoch": 2.752,
|
|
"grad_norm": 0.4639599429519105,
|
|
"learning_rate": 1.9910496234551132e-05,
|
|
"loss": 0.8658,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 2.7648,
|
|
"grad_norm": 0.46951766652022153,
|
|
"learning_rate": 1.9731495873732055e-05,
|
|
"loss": 0.8324,
|
|
"step": 216
|
|
},
|
|
{
|
|
"epoch": 2.7776,
|
|
"grad_norm": 0.44882362079406707,
|
|
"learning_rate": 1.9552517022574542e-05,
|
|
"loss": 0.8876,
|
|
"step": 217
|
|
},
|
|
{
|
|
"epoch": 2.7904,
|
|
"grad_norm": 0.52018784813985,
|
|
"learning_rate": 1.93735740189384e-05,
|
|
"loss": 0.873,
|
|
"step": 218
|
|
},
|
|
{
|
|
"epoch": 2.8032,
|
|
"grad_norm": 0.4709442382740269,
|
|
"learning_rate": 1.9194681197811703e-05,
|
|
"loss": 0.9178,
|
|
"step": 219
|
|
},
|
|
{
|
|
"epoch": 2.816,
|
|
"grad_norm": 0.5299421391031691,
|
|
"learning_rate": 1.901585289016244e-05,
|
|
"loss": 0.9269,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 2.8288,
|
|
"grad_norm": 0.4897808540168492,
|
|
"learning_rate": 1.8837103421790486e-05,
|
|
"loss": 0.9416,
|
|
"step": 221
|
|
},
|
|
{
|
|
"epoch": 2.8416,
|
|
"grad_norm": 0.46425335582121,
|
|
"learning_rate": 1.8658447112179952e-05,
|
|
"loss": 0.9035,
|
|
"step": 222
|
|
},
|
|
{
|
|
"epoch": 2.8544,
|
|
"grad_norm": 0.518932487390929,
|
|
"learning_rate": 1.8479898273352084e-05,
|
|
"loss": 0.8947,
|
|
"step": 223
|
|
},
|
|
{
|
|
"epoch": 2.8672,
|
|
"grad_norm": 0.5666166959233204,
|
|
"learning_rate": 1.83014712087187e-05,
|
|
"loss": 0.8635,
|
|
"step": 224
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"grad_norm": 0.487924819853616,
|
|
"learning_rate": 1.8123180211936417e-05,
|
|
"loss": 0.9178,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 2.8928000000000003,
|
|
"grad_norm": 0.46873872119200616,
|
|
"learning_rate": 1.794503956576152e-05,
|
|
"loss": 0.8915,
|
|
"step": 226
|
|
},
|
|
{
|
|
"epoch": 2.9055999999999997,
|
|
"grad_norm": 0.5407830565438971,
|
|
"learning_rate": 1.776706354090585e-05,
|
|
"loss": 0.8862,
|
|
"step": 227
|
|
},
|
|
{
|
|
"epoch": 2.9184,
|
|
"grad_norm": 0.5886391967590532,
|
|
"learning_rate": 1.758926639489354e-05,
|
|
"loss": 0.8611,
|
|
"step": 228
|
|
},
|
|
{
|
|
"epoch": 2.9312,
|
|
"grad_norm": 0.47630443340352235,
|
|
"learning_rate": 1.7411662370918893e-05,
|
|
"loss": 0.861,
|
|
"step": 229
|
|
},
|
|
{
|
|
"epoch": 2.944,
|
|
"grad_norm": 0.6098771136390113,
|
|
"learning_rate": 1.7234265696705344e-05,
|
|
"loss": 0.92,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 2.9568,
|
|
"grad_norm": 0.45692835759900857,
|
|
"learning_rate": 1.7057090583365678e-05,
|
|
"loss": 0.9233,
|
|
"step": 231
|
|
},
|
|
{
|
|
"epoch": 2.9696,
|
|
"grad_norm": 0.5576678894833484,
|
|
"learning_rate": 1.6880151224263646e-05,
|
|
"loss": 0.8436,
|
|
"step": 232
|
|
},
|
|
{
|
|
"epoch": 2.9824,
|
|
"grad_norm": 0.4685258580438374,
|
|
"learning_rate": 1.6703461793876876e-05,
|
|
"loss": 0.9097,
|
|
"step": 233
|
|
},
|
|
{
|
|
"epoch": 2.9952,
|
|
"grad_norm": 0.504819149781262,
|
|
"learning_rate": 1.6527036446661396e-05,
|
|
"loss": 0.8545,
|
|
"step": 234
|
|
},
|
|
{
|
|
"epoch": 3.008,
|
|
"grad_norm": 0.6519811830810548,
|
|
"learning_rate": 1.635088931591775e-05,
|
|
"loss": 0.8374,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 3.0208,
|
|
"grad_norm": 0.4874725917892601,
|
|
"learning_rate": 1.6175034512658753e-05,
|
|
"loss": 0.7423,
|
|
"step": 236
|
|
},
|
|
{
|
|
"epoch": 3.0336,
|
|
"grad_norm": 0.7398932139222312,
|
|
"learning_rate": 1.5999486124479115e-05,
|
|
"loss": 0.7227,
|
|
"step": 237
|
|
},
|
|
{
|
|
"epoch": 3.0464,
|
|
"grad_norm": 0.7191644498891802,
|
|
"learning_rate": 1.5824258214426833e-05,
|
|
"loss": 0.7677,
|
|
"step": 238
|
|
},
|
|
{
|
|
"epoch": 3.0592,
|
|
"grad_norm": 0.6196994167588507,
|
|
"learning_rate": 1.5649364819876655e-05,
|
|
"loss": 0.7768,
|
|
"step": 239
|
|
},
|
|
{
|
|
"epoch": 3.072,
|
|
"grad_norm": 0.6894402364689628,
|
|
"learning_rate": 1.547481995140556e-05,
|
|
"loss": 0.7745,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 3.0848,
|
|
"grad_norm": 0.5705946275968563,
|
|
"learning_rate": 1.5300637591670357e-05,
|
|
"loss": 0.7342,
|
|
"step": 241
|
|
},
|
|
{
|
|
"epoch": 3.0976,
|
|
"grad_norm": 0.611007325572401,
|
|
"learning_rate": 1.5126831694287564e-05,
|
|
"loss": 0.7096,
|
|
"step": 242
|
|
},
|
|
{
|
|
"epoch": 3.1104,
|
|
"grad_norm": 0.6354279224520238,
|
|
"learning_rate": 1.4953416182715566e-05,
|
|
"loss": 0.7988,
|
|
"step": 243
|
|
},
|
|
{
|
|
"epoch": 3.1232,
|
|
"grad_norm": 0.557072613263889,
|
|
"learning_rate": 1.478040494913926e-05,
|
|
"loss": 0.7481,
|
|
"step": 244
|
|
},
|
|
{
|
|
"epoch": 3.136,
|
|
"grad_norm": 0.6886549270766755,
|
|
"learning_rate": 1.460781185335713e-05,
|
|
"loss": 0.757,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 3.1488,
|
|
"grad_norm": 0.5311903167356413,
|
|
"learning_rate": 1.443565072167095e-05,
|
|
"loss": 0.7826,
|
|
"step": 246
|
|
},
|
|
{
|
|
"epoch": 3.1616,
|
|
"grad_norm": 0.5878784893665593,
|
|
"learning_rate": 1.4263935345778202e-05,
|
|
"loss": 0.7886,
|
|
"step": 247
|
|
},
|
|
{
|
|
"epoch": 3.1744,
|
|
"grad_norm": 0.5497592443294768,
|
|
"learning_rate": 1.409267948166718e-05,
|
|
"loss": 0.7768,
|
|
"step": 248
|
|
},
|
|
{
|
|
"epoch": 3.1872,
|
|
"grad_norm": 0.499403040556143,
|
|
"learning_rate": 1.3921896848515064e-05,
|
|
"loss": 0.7398,
|
|
"step": 249
|
|
},
|
|
{
|
|
"epoch": 3.2,
|
|
"grad_norm": 0.5176994890224423,
|
|
"learning_rate": 1.3751601127588849e-05,
|
|
"loss": 0.7344,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 3.2128,
|
|
"grad_norm": 0.513937352982502,
|
|
"learning_rate": 1.3581805961149371e-05,
|
|
"loss": 0.7289,
|
|
"step": 251
|
|
},
|
|
{
|
|
"epoch": 3.2256,
|
|
"grad_norm": 0.5085230133863774,
|
|
"learning_rate": 1.341252495135841e-05,
|
|
"loss": 0.7285,
|
|
"step": 252
|
|
},
|
|
{
|
|
"epoch": 3.2384,
|
|
"grad_norm": 0.48502274403674484,
|
|
"learning_rate": 1.324377165918906e-05,
|
|
"loss": 0.7188,
|
|
"step": 253
|
|
},
|
|
{
|
|
"epoch": 3.2512,
|
|
"grad_norm": 0.44755314341811114,
|
|
"learning_rate": 1.3075559603339354e-05,
|
|
"loss": 0.7723,
|
|
"step": 254
|
|
},
|
|
{
|
|
"epoch": 3.2640000000000002,
|
|
"grad_norm": 0.4865070203678609,
|
|
"learning_rate": 1.2907902259149287e-05,
|
|
"loss": 0.7881,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 3.2768,
|
|
"grad_norm": 0.4313048125863158,
|
|
"learning_rate": 1.274081305752135e-05,
|
|
"loss": 0.7191,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 3.2896,
|
|
"grad_norm": 0.4240472484996708,
|
|
"learning_rate": 1.2574305383844528e-05,
|
|
"loss": 0.7464,
|
|
"step": 257
|
|
},
|
|
{
|
|
"epoch": 3.3024,
|
|
"grad_norm": 0.4325166990775412,
|
|
"learning_rate": 1.2408392576922075e-05,
|
|
"loss": 0.7621,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 3.3152,
|
|
"grad_norm": 0.4566343829800513,
|
|
"learning_rate": 1.2243087927902905e-05,
|
|
"loss": 0.7178,
|
|
"step": 259
|
|
},
|
|
{
|
|
"epoch": 3.328,
|
|
"grad_norm": 0.41356618644878995,
|
|
"learning_rate": 1.2078404679216864e-05,
|
|
"loss": 0.8272,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 3.3407999999999998,
|
|
"grad_norm": 0.4076080621370411,
|
|
"learning_rate": 1.1914356023513904e-05,
|
|
"loss": 0.7726,
|
|
"step": 261
|
|
},
|
|
{
|
|
"epoch": 3.3536,
|
|
"grad_norm": 0.45042673694474916,
|
|
"learning_rate": 1.1750955102607193e-05,
|
|
"loss": 0.7578,
|
|
"step": 262
|
|
},
|
|
{
|
|
"epoch": 3.3664,
|
|
"grad_norm": 0.3948818114710141,
|
|
"learning_rate": 1.1588215006420374e-05,
|
|
"loss": 0.7829,
|
|
"step": 263
|
|
},
|
|
{
|
|
"epoch": 3.3792,
|
|
"grad_norm": 0.40820928538282475,
|
|
"learning_rate": 1.1426148771938915e-05,
|
|
"loss": 0.7438,
|
|
"step": 264
|
|
},
|
|
{
|
|
"epoch": 3.392,
|
|
"grad_norm": 0.40130937332636657,
|
|
"learning_rate": 1.1264769382165748e-05,
|
|
"loss": 0.7771,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 3.4048,
|
|
"grad_norm": 0.43505495772614156,
|
|
"learning_rate": 1.110408976508118e-05,
|
|
"loss": 0.7672,
|
|
"step": 266
|
|
},
|
|
{
|
|
"epoch": 3.4176,
|
|
"grad_norm": 0.40695836943831887,
|
|
"learning_rate": 1.094412279260726e-05,
|
|
"loss": 0.7912,
|
|
"step": 267
|
|
},
|
|
{
|
|
"epoch": 3.4304,
|
|
"grad_norm": 0.42588542701909704,
|
|
"learning_rate": 1.0784881279576635e-05,
|
|
"loss": 0.784,
|
|
"step": 268
|
|
},
|
|
{
|
|
"epoch": 3.4432,
|
|
"grad_norm": 0.43094699386668583,
|
|
"learning_rate": 1.0626377982705929e-05,
|
|
"loss": 0.7489,
|
|
"step": 269
|
|
},
|
|
{
|
|
"epoch": 3.456,
|
|
"grad_norm": 0.39680827424605447,
|
|
"learning_rate": 1.0468625599573842e-05,
|
|
"loss": 0.7784,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 3.4688,
|
|
"grad_norm": 0.3970909365963024,
|
|
"learning_rate": 1.0311636767603952e-05,
|
|
"loss": 0.7935,
|
|
"step": 271
|
|
},
|
|
{
|
|
"epoch": 3.4816,
|
|
"grad_norm": 0.4458156807009302,
|
|
"learning_rate": 1.0155424063052306e-05,
|
|
"loss": 0.7564,
|
|
"step": 272
|
|
},
|
|
{
|
|
"epoch": 3.4944,
|
|
"grad_norm": 0.4207458966216561,
|
|
"learning_rate": 1.0000000000000006e-05,
|
|
"loss": 0.753,
|
|
"step": 273
|
|
},
|
|
{
|
|
"epoch": 3.5072,
|
|
"grad_norm": 0.3970972025429375,
|
|
"learning_rate": 9.84537702935065e-06,
|
|
"loss": 0.7325,
|
|
"step": 274
|
|
},
|
|
{
|
|
"epoch": 3.52,
|
|
"grad_norm": 0.39976069168626505,
|
|
"learning_rate": 9.691567537832964e-06,
|
|
"loss": 0.7293,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 3.5328,
|
|
"grad_norm": 0.421956049762003,
|
|
"learning_rate": 9.538583847008452e-06,
|
|
"loss": 0.7106,
|
|
"step": 276
|
|
},
|
|
{
|
|
"epoch": 3.5456,
|
|
"grad_norm": 0.4237565677721429,
|
|
"learning_rate": 9.386438212284372e-06,
|
|
"loss": 0.7289,
|
|
"step": 277
|
|
},
|
|
{
|
|
"epoch": 3.5584,
|
|
"grad_norm": 0.38997648773116933,
|
|
"learning_rate": 9.235142821931928e-06,
|
|
"loss": 0.7199,
|
|
"step": 278
|
|
},
|
|
{
|
|
"epoch": 3.5712,
|
|
"grad_norm": 0.3857372505711672,
|
|
"learning_rate": 9.084709796109907e-06,
|
|
"loss": 0.7905,
|
|
"step": 279
|
|
},
|
|
{
|
|
"epoch": 3.584,
|
|
"grad_norm": 0.41795353699534166,
|
|
"learning_rate": 8.93515118589373e-06,
|
|
"loss": 0.7608,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 3.5968,
|
|
"grad_norm": 0.37577239930421513,
|
|
"learning_rate": 8.786478972310023e-06,
|
|
"loss": 0.7573,
|
|
"step": 281
|
|
},
|
|
{
|
|
"epoch": 3.6096,
|
|
"grad_norm": 0.3644839989291009,
|
|
"learning_rate": 8.638705065376887e-06,
|
|
"loss": 0.7468,
|
|
"step": 282
|
|
},
|
|
{
|
|
"epoch": 3.6224,
|
|
"grad_norm": 0.3803277128946228,
|
|
"learning_rate": 8.491841303149728e-06,
|
|
"loss": 0.7759,
|
|
"step": 283
|
|
},
|
|
{
|
|
"epoch": 3.6352,
|
|
"grad_norm": 0.371863808115335,
|
|
"learning_rate": 8.345899450772975e-06,
|
|
"loss": 0.7509,
|
|
"step": 284
|
|
},
|
|
{
|
|
"epoch": 3.648,
|
|
"grad_norm": 0.3644754129069241,
|
|
"learning_rate": 8.200891199537549e-06,
|
|
"loss": 0.7356,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 3.6608,
|
|
"grad_norm": 0.40268320556516357,
|
|
"learning_rate": 8.056828165944282e-06,
|
|
"loss": 0.7068,
|
|
"step": 286
|
|
},
|
|
{
|
|
"epoch": 3.6736,
|
|
"grad_norm": 0.394716063369627,
|
|
"learning_rate": 7.913721890773354e-06,
|
|
"loss": 0.759,
|
|
"step": 287
|
|
},
|
|
{
|
|
"epoch": 3.6864,
|
|
"grad_norm": 0.38812622954106013,
|
|
"learning_rate": 7.771583838159756e-06,
|
|
"loss": 0.7596,
|
|
"step": 288
|
|
},
|
|
{
|
|
"epoch": 3.6992000000000003,
|
|
"grad_norm": 0.37922553085096566,
|
|
"learning_rate": 7.630425394674903e-06,
|
|
"loss": 0.7434,
|
|
"step": 289
|
|
},
|
|
{
|
|
"epoch": 3.7119999999999997,
|
|
"grad_norm": 0.38003748075383165,
|
|
"learning_rate": 7.49025786841445e-06,
|
|
"loss": 0.7631,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 3.7248,
|
|
"grad_norm": 0.35758289241851654,
|
|
"learning_rate": 7.3510924880924575e-06,
|
|
"loss": 0.7464,
|
|
"step": 291
|
|
},
|
|
{
|
|
"epoch": 3.7376,
|
|
"grad_norm": 0.3954391374993535,
|
|
"learning_rate": 7.212940402141808e-06,
|
|
"loss": 0.7257,
|
|
"step": 292
|
|
},
|
|
{
|
|
"epoch": 3.7504,
|
|
"grad_norm": 0.3695155475353515,
|
|
"learning_rate": 7.075812677821145e-06,
|
|
"loss": 0.7306,
|
|
"step": 293
|
|
},
|
|
{
|
|
"epoch": 3.7632,
|
|
"grad_norm": 0.3481154912845101,
|
|
"learning_rate": 6.939720300328303e-06,
|
|
"loss": 0.7637,
|
|
"step": 294
|
|
},
|
|
{
|
|
"epoch": 3.776,
|
|
"grad_norm": 0.36757114334729113,
|
|
"learning_rate": 6.8046741719202385e-06,
|
|
"loss": 0.7364,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 3.7888,
|
|
"grad_norm": 0.36766231492812096,
|
|
"learning_rate": 6.67068511103971e-06,
|
|
"loss": 0.7154,
|
|
"step": 296
|
|
},
|
|
{
|
|
"epoch": 3.8016,
|
|
"grad_norm": 0.3594126418739973,
|
|
"learning_rate": 6.537763851448593e-06,
|
|
"loss": 0.739,
|
|
"step": 297
|
|
},
|
|
{
|
|
"epoch": 3.8144,
|
|
"grad_norm": 0.3695278652729132,
|
|
"learning_rate": 6.4059210413680175e-06,
|
|
"loss": 0.7404,
|
|
"step": 298
|
|
},
|
|
{
|
|
"epoch": 3.8272,
|
|
"grad_norm": 0.3469218747542019,
|
|
"learning_rate": 6.275167242625331e-06,
|
|
"loss": 0.775,
|
|
"step": 299
|
|
},
|
|
{
|
|
"epoch": 3.84,
|
|
"grad_norm": 0.3530007874211548,
|
|
"learning_rate": 6.145512929808013e-06,
|
|
"loss": 0.806,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 3.8528000000000002,
|
|
"grad_norm": 0.34044311854526643,
|
|
"learning_rate": 6.016968489424572e-06,
|
|
"loss": 0.7129,
|
|
"step": 301
|
|
},
|
|
{
|
|
"epoch": 3.8656,
|
|
"grad_norm": 0.3410813237685364,
|
|
"learning_rate": 5.889544219072465e-06,
|
|
"loss": 0.7298,
|
|
"step": 302
|
|
},
|
|
{
|
|
"epoch": 3.8784,
|
|
"grad_norm": 0.34537139147728846,
|
|
"learning_rate": 5.7632503266131925e-06,
|
|
"loss": 0.7621,
|
|
"step": 303
|
|
},
|
|
{
|
|
"epoch": 3.8912,
|
|
"grad_norm": 0.34471319900197667,
|
|
"learning_rate": 5.638096929354522e-06,
|
|
"loss": 0.7537,
|
|
"step": 304
|
|
},
|
|
{
|
|
"epoch": 3.904,
|
|
"grad_norm": 0.3472159637835492,
|
|
"learning_rate": 5.514094053240035e-06,
|
|
"loss": 0.7452,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 3.9168,
|
|
"grad_norm": 0.3447595885558001,
|
|
"learning_rate": 5.39125163204594e-06,
|
|
"loss": 0.803,
|
|
"step": 306
|
|
},
|
|
{
|
|
"epoch": 3.9295999999999998,
|
|
"grad_norm": 0.34569322352863874,
|
|
"learning_rate": 5.269579506585259e-06,
|
|
"loss": 0.7645,
|
|
"step": 307
|
|
},
|
|
{
|
|
"epoch": 3.9424,
|
|
"grad_norm": 0.3536645907641703,
|
|
"learning_rate": 5.149087423919541e-06,
|
|
"loss": 0.7452,
|
|
"step": 308
|
|
},
|
|
{
|
|
"epoch": 3.9552,
|
|
"grad_norm": 0.3455657100117534,
|
|
"learning_rate": 5.029785036577976e-06,
|
|
"loss": 0.739,
|
|
"step": 309
|
|
},
|
|
{
|
|
"epoch": 3.968,
|
|
"grad_norm": 0.3377694367318897,
|
|
"learning_rate": 4.911681901784198e-06,
|
|
"loss": 0.7621,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 3.9808,
|
|
"grad_norm": 0.331860075985349,
|
|
"learning_rate": 4.794787480690597e-06,
|
|
"loss": 0.7688,
|
|
"step": 311
|
|
},
|
|
{
|
|
"epoch": 3.9936,
|
|
"grad_norm": 0.35761985983691535,
|
|
"learning_rate": 4.679111137620442e-06,
|
|
"loss": 0.7227,
|
|
"step": 312
|
|
},
|
|
{
|
|
"epoch": 4.0064,
|
|
"grad_norm": 0.37198317152723936,
|
|
"learning_rate": 4.5646621393177e-06,
|
|
"loss": 0.7252,
|
|
"step": 313
|
|
},
|
|
{
|
|
"epoch": 4.0192,
|
|
"grad_norm": 0.6561852689389754,
|
|
"learning_rate": 4.451449654204685e-06,
|
|
"loss": 0.6061,
|
|
"step": 314
|
|
},
|
|
{
|
|
"epoch": 4.032,
|
|
"grad_norm": 0.36651131968397777,
|
|
"learning_rate": 4.339482751647557e-06,
|
|
"loss": 0.7034,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 4.0448,
|
|
"grad_norm": 0.37514756639768854,
|
|
"learning_rate": 4.228770401229824e-06,
|
|
"loss": 0.6067,
|
|
"step": 316
|
|
},
|
|
{
|
|
"epoch": 4.0576,
|
|
"grad_norm": 0.5566002695565736,
|
|
"learning_rate": 4.119321472033779e-06,
|
|
"loss": 0.6513,
|
|
"step": 317
|
|
},
|
|
{
|
|
"epoch": 4.0704,
|
|
"grad_norm": 0.6058517265401333,
|
|
"learning_rate": 4.011144731929981e-06,
|
|
"loss": 0.7156,
|
|
"step": 318
|
|
},
|
|
{
|
|
"epoch": 4.0832,
|
|
"grad_norm": 0.4381743948996236,
|
|
"learning_rate": 3.904248846874894e-06,
|
|
"loss": 0.6634,
|
|
"step": 319
|
|
},
|
|
{
|
|
"epoch": 4.096,
|
|
"grad_norm": 0.3907006773329518,
|
|
"learning_rate": 3.7986423802166705e-06,
|
|
"loss": 0.6652,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 4.1088,
|
|
"grad_norm": 0.475663431057593,
|
|
"learning_rate": 3.694333792009115e-06,
|
|
"loss": 0.6529,
|
|
"step": 321
|
|
},
|
|
{
|
|
"epoch": 4.1216,
|
|
"grad_norm": 0.4781882268473551,
|
|
"learning_rate": 3.5913314383339937e-06,
|
|
"loss": 0.7027,
|
|
"step": 322
|
|
},
|
|
{
|
|
"epoch": 4.1344,
|
|
"grad_norm": 0.40575808944524244,
|
|
"learning_rate": 3.4896435706316e-06,
|
|
"loss": 0.7221,
|
|
"step": 323
|
|
},
|
|
{
|
|
"epoch": 4.1472,
|
|
"grad_norm": 0.4008959323794245,
|
|
"learning_rate": 3.3892783350397675e-06,
|
|
"loss": 0.7016,
|
|
"step": 324
|
|
},
|
|
{
|
|
"epoch": 4.16,
|
|
"grad_norm": 0.44940259724687015,
|
|
"learning_rate": 3.290243771741275e-06,
|
|
"loss": 0.6472,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 4.1728,
|
|
"grad_norm": 0.43604266733159375,
|
|
"learning_rate": 3.1925478143197418e-06,
|
|
"loss": 0.7008,
|
|
"step": 326
|
|
},
|
|
{
|
|
"epoch": 4.1856,
|
|
"grad_norm": 0.45059373555578275,
|
|
"learning_rate": 3.0961982891241083e-06,
|
|
"loss": 0.693,
|
|
"step": 327
|
|
},
|
|
{
|
|
"epoch": 4.1984,
|
|
"grad_norm": 0.42572730736655867,
|
|
"learning_rate": 3.001202914641628e-06,
|
|
"loss": 0.6633,
|
|
"step": 328
|
|
},
|
|
{
|
|
"epoch": 4.2112,
|
|
"grad_norm": 0.4174156855040001,
|
|
"learning_rate": 2.907569300879596e-06,
|
|
"loss": 0.6922,
|
|
"step": 329
|
|
},
|
|
{
|
|
"epoch": 4.224,
|
|
"grad_norm": 0.3932112878326503,
|
|
"learning_rate": 2.815304948755664e-06,
|
|
"loss": 0.6372,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 4.2368,
|
|
"grad_norm": 0.4021422678728423,
|
|
"learning_rate": 2.7244172494969978e-06,
|
|
"loss": 0.7109,
|
|
"step": 331
|
|
},
|
|
{
|
|
"epoch": 4.2496,
|
|
"grad_norm": 0.41603513057451186,
|
|
"learning_rate": 2.6349134840481294e-06,
|
|
"loss": 0.677,
|
|
"step": 332
|
|
},
|
|
{
|
|
"epoch": 4.2624,
|
|
"grad_norm": 0.40701323138346235,
|
|
"learning_rate": 2.546800822487714e-06,
|
|
"loss": 0.677,
|
|
"step": 333
|
|
},
|
|
{
|
|
"epoch": 4.2752,
|
|
"grad_norm": 0.3618875995027296,
|
|
"learning_rate": 2.4600863234541338e-06,
|
|
"loss": 0.67,
|
|
"step": 334
|
|
},
|
|
{
|
|
"epoch": 4.288,
|
|
"grad_norm": 0.36204162502778214,
|
|
"learning_rate": 2.374776933580025e-06,
|
|
"loss": 0.6799,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 4.3008,
|
|
"grad_norm": 0.3354404902702719,
|
|
"learning_rate": 2.2908794869358044e-06,
|
|
"loss": 0.6442,
|
|
"step": 336
|
|
},
|
|
{
|
|
"epoch": 4.3136,
|
|
"grad_norm": 0.3505177803543528,
|
|
"learning_rate": 2.2084007044821764e-06,
|
|
"loss": 0.6758,
|
|
"step": 337
|
|
},
|
|
{
|
|
"epoch": 4.3264,
|
|
"grad_norm": 0.37232280470334717,
|
|
"learning_rate": 2.127347193531757e-06,
|
|
"loss": 0.6838,
|
|
"step": 338
|
|
},
|
|
{
|
|
"epoch": 4.3392,
|
|
"grad_norm": 0.38496967345842165,
|
|
"learning_rate": 2.0477254472197237e-06,
|
|
"loss": 0.6978,
|
|
"step": 339
|
|
},
|
|
{
|
|
"epoch": 4.352,
|
|
"grad_norm": 0.3887365973453678,
|
|
"learning_rate": 1.96954184398368e-06,
|
|
"loss": 0.6465,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 4.3648,
|
|
"grad_norm": 0.3560874514147073,
|
|
"learning_rate": 1.8928026470526917e-06,
|
|
"loss": 0.6143,
|
|
"step": 341
|
|
},
|
|
{
|
|
"epoch": 4.3776,
|
|
"grad_norm": 0.3234250557774697,
|
|
"learning_rate": 1.817514003945524e-06,
|
|
"loss": 0.6868,
|
|
"step": 342
|
|
},
|
|
{
|
|
"epoch": 4.3904,
|
|
"grad_norm": 0.30275731309740356,
|
|
"learning_rate": 1.743681945978184e-06,
|
|
"loss": 0.6901,
|
|
"step": 343
|
|
},
|
|
{
|
|
"epoch": 4.4032,
|
|
"grad_norm": 0.30620924708135805,
|
|
"learning_rate": 1.6713123877807413e-06,
|
|
"loss": 0.6339,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 4.416,
|
|
"grad_norm": 0.31118959285928466,
|
|
"learning_rate": 1.6004111268235156e-06,
|
|
"loss": 0.682,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 4.4288,
|
|
"grad_norm": 0.35141999300884796,
|
|
"learning_rate": 1.5309838429526714e-06,
|
|
"loss": 0.7029,
|
|
"step": 346
|
|
},
|
|
{
|
|
"epoch": 4.4416,
|
|
"grad_norm": 0.3606055859454003,
|
|
"learning_rate": 1.4630360979351644e-06,
|
|
"loss": 0.6667,
|
|
"step": 347
|
|
},
|
|
{
|
|
"epoch": 4.4544,
|
|
"grad_norm": 0.3345093450040408,
|
|
"learning_rate": 1.396573335013236e-06,
|
|
"loss": 0.7034,
|
|
"step": 348
|
|
},
|
|
{
|
|
"epoch": 4.4672,
|
|
"grad_norm": 0.31136247386805826,
|
|
"learning_rate": 1.3316008784683265e-06,
|
|
"loss": 0.6817,
|
|
"step": 349
|
|
},
|
|
{
|
|
"epoch": 4.48,
|
|
"grad_norm": 0.2947359919955738,
|
|
"learning_rate": 1.2681239331945695e-06,
|
|
"loss": 0.6848,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 4.4928,
|
|
"grad_norm": 0.30007880824834926,
|
|
"learning_rate": 1.2061475842818337e-06,
|
|
"loss": 0.7368,
|
|
"step": 351
|
|
},
|
|
{
|
|
"epoch": 4.5056,
|
|
"grad_norm": 0.3050371284095851,
|
|
"learning_rate": 1.1456767966083393e-06,
|
|
"loss": 0.6892,
|
|
"step": 352
|
|
},
|
|
{
|
|
"epoch": 4.5184,
|
|
"grad_norm": 0.3291555653795982,
|
|
"learning_rate": 1.086716414442952e-06,
|
|
"loss": 0.684,
|
|
"step": 353
|
|
},
|
|
{
|
|
"epoch": 4.5312,
|
|
"grad_norm": 0.3310226992443401,
|
|
"learning_rate": 1.0292711610570904e-06,
|
|
"loss": 0.7055,
|
|
"step": 354
|
|
},
|
|
{
|
|
"epoch": 4.5440000000000005,
|
|
"grad_norm": 0.3124014951458642,
|
|
"learning_rate": 9.733456383463658e-07,
|
|
"loss": 0.5981,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 4.5568,
|
|
"grad_norm": 0.311371463834397,
|
|
"learning_rate": 9.189443264619102e-07,
|
|
"loss": 0.7119,
|
|
"step": 356
|
|
},
|
|
{
|
|
"epoch": 4.5696,
|
|
"grad_norm": 0.30060369011333293,
|
|
"learning_rate": 8.660715834514977e-07,
|
|
"loss": 0.6717,
|
|
"step": 357
|
|
},
|
|
{
|
|
"epoch": 4.5824,
|
|
"grad_norm": 0.3116295751001881,
|
|
"learning_rate": 8.147316449103959e-07,
|
|
"loss": 0.646,
|
|
"step": 358
|
|
},
|
|
{
|
|
"epoch": 4.5952,
|
|
"grad_norm": 0.31372899414770516,
|
|
"learning_rate": 7.649286236420806e-07,
|
|
"loss": 0.6589,
|
|
"step": 359
|
|
},
|
|
{
|
|
"epoch": 4.608,
|
|
"grad_norm": 0.2973766852081861,
|
|
"learning_rate": 7.166665093287539e-07,
|
|
"loss": 0.7039,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 4.6208,
|
|
"grad_norm": 0.2956464896205796,
|
|
"learning_rate": 6.69949168211721e-07,
|
|
"loss": 0.7086,
|
|
"step": 361
|
|
},
|
|
{
|
|
"epoch": 4.6336,
|
|
"grad_norm": 0.3050968161758319,
|
|
"learning_rate": 6.247803427816945e-07,
|
|
"loss": 0.6299,
|
|
"step": 362
|
|
},
|
|
{
|
|
"epoch": 4.6464,
|
|
"grad_norm": 0.2895560012638669,
|
|
"learning_rate": 5.811636514789598e-07,
|
|
"loss": 0.7068,
|
|
"step": 363
|
|
},
|
|
{
|
|
"epoch": 4.6592,
|
|
"grad_norm": 0.2860382042764144,
|
|
"learning_rate": 5.391025884035239e-07,
|
|
"loss": 0.674,
|
|
"step": 364
|
|
},
|
|
{
|
|
"epoch": 4.672,
|
|
"grad_norm": 0.31440519446759935,
|
|
"learning_rate": 4.986005230351954e-07,
|
|
"loss": 0.7265,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 4.6848,
|
|
"grad_norm": 0.2863757299429382,
|
|
"learning_rate": 4.5966069996365993e-07,
|
|
"loss": 0.7612,
|
|
"step": 366
|
|
},
|
|
{
|
|
"epoch": 4.6975999999999996,
|
|
"grad_norm": 0.30459290151108215,
|
|
"learning_rate": 4.22286238628562e-07,
|
|
"loss": 0.6638,
|
|
"step": 367
|
|
},
|
|
{
|
|
"epoch": 4.7104,
|
|
"grad_norm": 0.30622116120951404,
|
|
"learning_rate": 3.8648013306960664e-07,
|
|
"loss": 0.6798,
|
|
"step": 368
|
|
},
|
|
{
|
|
"epoch": 4.7232,
|
|
"grad_norm": 0.3135265781849215,
|
|
"learning_rate": 3.522452516867048e-07,
|
|
"loss": 0.6936,
|
|
"step": 369
|
|
},
|
|
{
|
|
"epoch": 4.736,
|
|
"grad_norm": 0.2974286222437718,
|
|
"learning_rate": 3.1958433701019697e-07,
|
|
"loss": 0.6592,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 4.7488,
|
|
"grad_norm": 0.29169010439567084,
|
|
"learning_rate": 2.8850000548115155e-07,
|
|
"loss": 0.6883,
|
|
"step": 371
|
|
},
|
|
{
|
|
"epoch": 4.7616,
|
|
"grad_norm": 0.2954857385163254,
|
|
"learning_rate": 2.5899474724174313e-07,
|
|
"loss": 0.6762,
|
|
"step": 372
|
|
},
|
|
{
|
|
"epoch": 4.7744,
|
|
"grad_norm": 0.30554350788092144,
|
|
"learning_rate": 2.3107092593579905e-07,
|
|
"loss": 0.6383,
|
|
"step": 373
|
|
},
|
|
{
|
|
"epoch": 4.7872,
|
|
"grad_norm": 0.2871638236130085,
|
|
"learning_rate": 2.0473077851942858e-07,
|
|
"loss": 0.6458,
|
|
"step": 374
|
|
},
|
|
{
|
|
"epoch": 4.8,
|
|
"grad_norm": 0.3045991261393286,
|
|
"learning_rate": 1.799764150818306e-07,
|
|
"loss": 0.6113,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 4.8128,
|
|
"grad_norm": 0.303297019465311,
|
|
"learning_rate": 1.5680981867625566e-07,
|
|
"loss": 0.6758,
|
|
"step": 376
|
|
},
|
|
{
|
|
"epoch": 4.8256,
|
|
"grad_norm": 0.29854505865022923,
|
|
"learning_rate": 1.3523284516113955e-07,
|
|
"loss": 0.6966,
|
|
"step": 377
|
|
},
|
|
{
|
|
"epoch": 4.8384,
|
|
"grad_norm": 0.29404174749341805,
|
|
"learning_rate": 1.1524722305144231e-07,
|
|
"loss": 0.7033,
|
|
"step": 378
|
|
},
|
|
{
|
|
"epoch": 4.8512,
|
|
"grad_norm": 0.3008331194546275,
|
|
"learning_rate": 9.685455338016347e-08,
|
|
"loss": 0.6582,
|
|
"step": 379
|
|
},
|
|
{
|
|
"epoch": 4.864,
|
|
"grad_norm": 0.3004375174980317,
|
|
"learning_rate": 8.005630957010014e-08,
|
|
"loss": 0.6699,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 4.8768,
|
|
"grad_norm": 0.3188125927248378,
|
|
"learning_rate": 6.485383731580142e-08,
|
|
"loss": 0.6217,
|
|
"step": 381
|
|
},
|
|
{
|
|
"epoch": 4.8896,
|
|
"grad_norm": 0.2949431769950282,
|
|
"learning_rate": 5.1248354475768034e-08,
|
|
"loss": 0.6765,
|
|
"step": 382
|
|
},
|
|
{
|
|
"epoch": 4.9024,
|
|
"grad_norm": 0.28985927803177125,
|
|
"learning_rate": 3.924095097489922e-08,
|
|
"loss": 0.7327,
|
|
"step": 383
|
|
},
|
|
{
|
|
"epoch": 4.9152000000000005,
|
|
"grad_norm": 0.30470347407831944,
|
|
"learning_rate": 2.8832588717164766e-08,
|
|
"loss": 0.6233,
|
|
"step": 384
|
|
},
|
|
{
|
|
"epoch": 4.928,
|
|
"grad_norm": 0.3015858128464978,
|
|
"learning_rate": 2.0024101508555604e-08,
|
|
"loss": 0.6435,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 4.9408,
|
|
"grad_norm": 0.30520060622615225,
|
|
"learning_rate": 1.281619499029274e-08,
|
|
"loss": 0.7127,
|
|
"step": 386
|
|
},
|
|
{
|
|
"epoch": 4.9536,
|
|
"grad_norm": 0.287862724065402,
|
|
"learning_rate": 7.209446582292501e-09,
|
|
"loss": 0.708,
|
|
"step": 387
|
|
},
|
|
{
|
|
"epoch": 4.9664,
|
|
"grad_norm": 0.28644752871236845,
|
|
"learning_rate": 3.2043054369057523e-09,
|
|
"loss": 0.7203,
|
|
"step": 388
|
|
},
|
|
{
|
|
"epoch": 4.9792,
|
|
"grad_norm": 0.29956360136524457,
|
|
"learning_rate": 8.010924029533406e-10,
|
|
"loss": 0.6833,
|
|
"step": 389
|
|
},
|
|
{
|
|
"epoch": 4.992,
|
|
"grad_norm": 0.2815319626646595,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.7082,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 4.992,
|
|
"step": 390,
|
|
"total_flos": 1.3856636017416602e+18,
|
|
"train_loss": 0.9147117905127696,
|
|
"train_runtime": 48380.7223,
|
|
"train_samples_per_second": 1.033,
|
|
"train_steps_per_second": 0.008
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 390,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 5,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.3856636017416602e+18,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|