chinese_extract_pert / trainer_state.json
frett's picture
End of training
e1fdb4d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 34595,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07226477814713109,
"grad_norm": 22.313980102539062,
"learning_rate": 2.9566411331117216e-05,
"loss": 3.3429,
"step": 500
},
{
"epoch": 0.14452955629426217,
"grad_norm": 8.817935943603516,
"learning_rate": 2.9132822662234428e-05,
"loss": 1.7422,
"step": 1000
},
{
"epoch": 0.21679433444139326,
"grad_norm": 96.14582061767578,
"learning_rate": 2.8699233993351643e-05,
"loss": 1.1619,
"step": 1500
},
{
"epoch": 0.28905911258852435,
"grad_norm": 4.378906726837158,
"learning_rate": 2.8265645324468854e-05,
"loss": 1.0505,
"step": 2000
},
{
"epoch": 0.36132389073565546,
"grad_norm": 4.367362022399902,
"learning_rate": 2.783205665558607e-05,
"loss": 0.9841,
"step": 2500
},
{
"epoch": 0.4335886688827865,
"grad_norm": 342.0606689453125,
"learning_rate": 2.739846798670328e-05,
"loss": 0.861,
"step": 3000
},
{
"epoch": 0.5058534470299176,
"grad_norm": 23.573060989379883,
"learning_rate": 2.6964879317820496e-05,
"loss": 0.8402,
"step": 3500
},
{
"epoch": 0.5781182251770487,
"grad_norm": 42.652748107910156,
"learning_rate": 2.6531290648937708e-05,
"loss": 0.82,
"step": 4000
},
{
"epoch": 0.6503830033241798,
"grad_norm": 46.95698928833008,
"learning_rate": 2.6097701980054923e-05,
"loss": 0.77,
"step": 4500
},
{
"epoch": 0.7226477814713109,
"grad_norm": 44.56264877319336,
"learning_rate": 2.5664113311172135e-05,
"loss": 0.8083,
"step": 5000
},
{
"epoch": 0.794912559618442,
"grad_norm": 80.95307159423828,
"learning_rate": 2.523052464228935e-05,
"loss": 0.8054,
"step": 5500
},
{
"epoch": 0.867177337765573,
"grad_norm": 57.8803825378418,
"learning_rate": 2.4796935973406565e-05,
"loss": 0.7206,
"step": 6000
},
{
"epoch": 0.9394421159127041,
"grad_norm": 17.753629684448242,
"learning_rate": 2.4363347304523777e-05,
"loss": 0.7892,
"step": 6500
},
{
"epoch": 1.0117068940598353,
"grad_norm": 33.004276275634766,
"learning_rate": 2.392975863564099e-05,
"loss": 0.7122,
"step": 7000
},
{
"epoch": 1.0839716722069663,
"grad_norm": 0.034875206649303436,
"learning_rate": 2.34961699667582e-05,
"loss": 0.4892,
"step": 7500
},
{
"epoch": 1.1562364503540974,
"grad_norm": 26.65053367614746,
"learning_rate": 2.3062581297875416e-05,
"loss": 0.4959,
"step": 8000
},
{
"epoch": 1.2285012285012284,
"grad_norm": 0.3080042898654938,
"learning_rate": 2.2628992628992627e-05,
"loss": 0.5166,
"step": 8500
},
{
"epoch": 1.3007660066483595,
"grad_norm": 47.525901794433594,
"learning_rate": 2.2195403960109842e-05,
"loss": 0.4875,
"step": 9000
},
{
"epoch": 1.3730307847954908,
"grad_norm": 2.4876151084899902,
"learning_rate": 2.1761815291227054e-05,
"loss": 0.4864,
"step": 9500
},
{
"epoch": 1.4452955629426218,
"grad_norm": 0.8099455833435059,
"learning_rate": 2.132822662234427e-05,
"loss": 0.4836,
"step": 10000
},
{
"epoch": 1.517560341089753,
"grad_norm": 9.619250297546387,
"learning_rate": 2.089463795346148e-05,
"loss": 0.5483,
"step": 10500
},
{
"epoch": 1.589825119236884,
"grad_norm": 0.05848146229982376,
"learning_rate": 2.0461049284578696e-05,
"loss": 0.546,
"step": 11000
},
{
"epoch": 1.662089897384015,
"grad_norm": 42.597862243652344,
"learning_rate": 2.002746061569591e-05,
"loss": 0.5164,
"step": 11500
},
{
"epoch": 1.734354675531146,
"grad_norm": 47.1865119934082,
"learning_rate": 1.9593871946813123e-05,
"loss": 0.5734,
"step": 12000
},
{
"epoch": 1.8066194536782771,
"grad_norm": 23.797060012817383,
"learning_rate": 1.9160283277930338e-05,
"loss": 0.5089,
"step": 12500
},
{
"epoch": 1.8788842318254084,
"grad_norm": 0.4582177996635437,
"learning_rate": 1.872669460904755e-05,
"loss": 0.5102,
"step": 13000
},
{
"epoch": 1.9511490099725393,
"grad_norm": 75.59239196777344,
"learning_rate": 1.8293105940164765e-05,
"loss": 0.5414,
"step": 13500
},
{
"epoch": 2.0234137881196705,
"grad_norm": 27.790529251098633,
"learning_rate": 1.7859517271281977e-05,
"loss": 0.436,
"step": 14000
},
{
"epoch": 2.0956785662668014,
"grad_norm": 0.07247856259346008,
"learning_rate": 1.7425928602399192e-05,
"loss": 0.3202,
"step": 14500
},
{
"epoch": 2.1679433444139327,
"grad_norm": 6.203830242156982,
"learning_rate": 1.6992339933516404e-05,
"loss": 0.2893,
"step": 15000
},
{
"epoch": 2.240208122561064,
"grad_norm": 0.011503485031425953,
"learning_rate": 1.655875126463362e-05,
"loss": 0.321,
"step": 15500
},
{
"epoch": 2.312472900708195,
"grad_norm": 0.3793994188308716,
"learning_rate": 1.6125162595750834e-05,
"loss": 0.3078,
"step": 16000
},
{
"epoch": 2.384737678855326,
"grad_norm": 42.47032165527344,
"learning_rate": 1.5691573926868046e-05,
"loss": 0.2855,
"step": 16500
},
{
"epoch": 2.457002457002457,
"grad_norm": 73.11065673828125,
"learning_rate": 1.5257985257985259e-05,
"loss": 0.3357,
"step": 17000
},
{
"epoch": 2.529267235149588,
"grad_norm": 0.03242184966802597,
"learning_rate": 1.482439658910247e-05,
"loss": 0.3016,
"step": 17500
},
{
"epoch": 2.601532013296719,
"grad_norm": 0.020083703100681305,
"learning_rate": 1.4390807920219684e-05,
"loss": 0.3017,
"step": 18000
},
{
"epoch": 2.6737967914438503,
"grad_norm": 12.890129089355469,
"learning_rate": 1.39572192513369e-05,
"loss": 0.3065,
"step": 18500
},
{
"epoch": 2.7460615695909816,
"grad_norm": 0.07345220446586609,
"learning_rate": 1.3523630582454113e-05,
"loss": 0.338,
"step": 19000
},
{
"epoch": 2.8183263477381124,
"grad_norm": 37.88114547729492,
"learning_rate": 1.3090041913571326e-05,
"loss": 0.313,
"step": 19500
},
{
"epoch": 2.8905911258852437,
"grad_norm": 0.04430060461163521,
"learning_rate": 1.265645324468854e-05,
"loss": 0.2948,
"step": 20000
},
{
"epoch": 2.9628559040323745,
"grad_norm": 0.06189202517271042,
"learning_rate": 1.2222864575805753e-05,
"loss": 0.2838,
"step": 20500
},
{
"epoch": 3.035120682179506,
"grad_norm": 0.015777448192238808,
"learning_rate": 1.1789275906922966e-05,
"loss": 0.2385,
"step": 21000
},
{
"epoch": 3.1073854603266367,
"grad_norm": 0.02253924310207367,
"learning_rate": 1.135568723804018e-05,
"loss": 0.1713,
"step": 21500
},
{
"epoch": 3.179650238473768,
"grad_norm": 0.014288770034909248,
"learning_rate": 1.0922098569157393e-05,
"loss": 0.1905,
"step": 22000
},
{
"epoch": 3.2519150166208988,
"grad_norm": 0.1594344526529312,
"learning_rate": 1.0488509900274607e-05,
"loss": 0.1578,
"step": 22500
},
{
"epoch": 3.32417979476803,
"grad_norm": 0.011662287637591362,
"learning_rate": 1.0054921231391818e-05,
"loss": 0.1585,
"step": 23000
},
{
"epoch": 3.3964445729151613,
"grad_norm": 0.02400069124996662,
"learning_rate": 9.621332562509032e-06,
"loss": 0.1732,
"step": 23500
},
{
"epoch": 3.468709351062292,
"grad_norm": 0.07058549672365189,
"learning_rate": 9.187743893626247e-06,
"loss": 0.1597,
"step": 24000
},
{
"epoch": 3.5409741292094234,
"grad_norm": 0.04298946261405945,
"learning_rate": 8.75415522474346e-06,
"loss": 0.1542,
"step": 24500
},
{
"epoch": 3.6132389073565543,
"grad_norm": 0.005008801817893982,
"learning_rate": 8.320566555860674e-06,
"loss": 0.1514,
"step": 25000
},
{
"epoch": 3.6855036855036856,
"grad_norm": 0.09852935373783112,
"learning_rate": 7.886977886977887e-06,
"loss": 0.1706,
"step": 25500
},
{
"epoch": 3.757768463650817,
"grad_norm": 140.0009765625,
"learning_rate": 7.453389218095101e-06,
"loss": 0.1442,
"step": 26000
},
{
"epoch": 3.8300332417979477,
"grad_norm": 0.0774250477552414,
"learning_rate": 7.019800549212314e-06,
"loss": 0.1743,
"step": 26500
},
{
"epoch": 3.9022980199450785,
"grad_norm": 0.045473262667655945,
"learning_rate": 6.586211880329528e-06,
"loss": 0.1855,
"step": 27000
},
{
"epoch": 3.97456279809221,
"grad_norm": 0.022451166063547134,
"learning_rate": 6.152623211446741e-06,
"loss": 0.1622,
"step": 27500
},
{
"epoch": 4.046827576239341,
"grad_norm": 0.014400485903024673,
"learning_rate": 5.7190345425639545e-06,
"loss": 0.0775,
"step": 28000
},
{
"epoch": 4.119092354386472,
"grad_norm": 0.025828225538134575,
"learning_rate": 5.285445873681168e-06,
"loss": 0.0835,
"step": 28500
},
{
"epoch": 4.191357132533603,
"grad_norm": 0.048525627702474594,
"learning_rate": 4.851857204798381e-06,
"loss": 0.073,
"step": 29000
},
{
"epoch": 4.263621910680734,
"grad_norm": 0.01667250506579876,
"learning_rate": 4.418268535915595e-06,
"loss": 0.074,
"step": 29500
},
{
"epoch": 4.335886688827865,
"grad_norm": 0.01051491778343916,
"learning_rate": 3.984679867032808e-06,
"loss": 0.0839,
"step": 30000
},
{
"epoch": 4.408151466974997,
"grad_norm": 0.02080466039478779,
"learning_rate": 3.5510911981500216e-06,
"loss": 0.1082,
"step": 30500
},
{
"epoch": 4.480416245122128,
"grad_norm": 0.0018432741053402424,
"learning_rate": 3.117502529267235e-06,
"loss": 0.0994,
"step": 31000
},
{
"epoch": 4.552681023269258,
"grad_norm": 0.2674684524536133,
"learning_rate": 2.683913860384449e-06,
"loss": 0.121,
"step": 31500
},
{
"epoch": 4.62494580141639,
"grad_norm": 0.0021196361631155014,
"learning_rate": 2.250325191501662e-06,
"loss": 0.0784,
"step": 32000
},
{
"epoch": 4.697210579563521,
"grad_norm": 0.010477827861905098,
"learning_rate": 1.8167365226188756e-06,
"loss": 0.0836,
"step": 32500
},
{
"epoch": 4.769475357710652,
"grad_norm": 0.002533436520025134,
"learning_rate": 1.383147853736089e-06,
"loss": 0.0633,
"step": 33000
},
{
"epoch": 4.841740135857783,
"grad_norm": 0.0024999654851853848,
"learning_rate": 9.495591848533026e-07,
"loss": 0.0588,
"step": 33500
},
{
"epoch": 4.914004914004914,
"grad_norm": 0.0020577749237418175,
"learning_rate": 5.15970515970516e-07,
"loss": 0.0794,
"step": 34000
},
{
"epoch": 4.986269692152045,
"grad_norm": 0.19813172519207,
"learning_rate": 8.238184708772944e-08,
"loss": 0.0608,
"step": 34500
},
{
"epoch": 5.0,
"step": 34595,
"total_flos": 3.6156938713344e+16,
"train_loss": 0.43488813713979024,
"train_runtime": 4328.4475,
"train_samples_per_second": 31.969,
"train_steps_per_second": 7.992
}
],
"logging_steps": 500,
"max_steps": 34595,
"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": 3.6156938713344e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}