Qwen-Code-7B_2 / checkpoint-375 /trainer_state.json
JerryFan011018's picture
Upload folder using huggingface_hub
0f60832 verified
{
"best_global_step": 225,
"best_metric": 0.36194586753845215,
"best_model_checkpoint": "/content/output/checkpoint-225",
"epoch": 5.0,
"eval_steps": 500,
"global_step": 375,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.13333333333333333,
"grad_norm": 1.7940553426742554,
"learning_rate": 5.917159763313609e-06,
"loss": 1.3972,
"mean_token_accuracy": 0.7618847399950027,
"num_tokens": 31152.0,
"step": 10
},
{
"epoch": 0.26666666666666666,
"grad_norm": 1.0511088371276855,
"learning_rate": 1.1834319526627219e-05,
"loss": 1.2288,
"mean_token_accuracy": 0.7673537939786911,
"num_tokens": 61773.0,
"step": 20
},
{
"epoch": 0.4,
"grad_norm": 0.8096975088119507,
"learning_rate": 1.7751479289940828e-05,
"loss": 0.9255,
"mean_token_accuracy": 0.7948055118322372,
"num_tokens": 94758.0,
"step": 30
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.6213255524635315,
"learning_rate": 2.3668639053254438e-05,
"loss": 0.7366,
"mean_token_accuracy": 0.8212289035320282,
"num_tokens": 125915.0,
"step": 40
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.42887863516807556,
"learning_rate": 2.958579881656805e-05,
"loss": 0.5556,
"mean_token_accuracy": 0.8463607966899872,
"num_tokens": 159364.0,
"step": 50
},
{
"epoch": 0.8,
"grad_norm": 0.5108550786972046,
"learning_rate": 3.5502958579881656e-05,
"loss": 0.4676,
"mean_token_accuracy": 0.8614094287157059,
"num_tokens": 190374.0,
"step": 60
},
{
"epoch": 0.9333333333333333,
"grad_norm": 0.4310983121395111,
"learning_rate": 4.142011834319527e-05,
"loss": 0.4214,
"mean_token_accuracy": 0.8710317760705948,
"num_tokens": 222156.0,
"step": 70
},
{
"epoch": 1.0,
"eval_loss": 0.4043419659137726,
"eval_mean_token_accuracy": 0.8734077858924866,
"eval_num_tokens": 238011.0,
"eval_runtime": 50.8925,
"eval_samples_per_second": 1.965,
"eval_steps_per_second": 0.491,
"step": 75
},
{
"epoch": 1.0666666666666667,
"grad_norm": 0.35767048597335815,
"learning_rate": 4.7337278106508875e-05,
"loss": 0.3901,
"mean_token_accuracy": 0.8787548273801804,
"num_tokens": 255825.0,
"step": 80
},
{
"epoch": 1.2,
"grad_norm": 0.3684692680835724,
"learning_rate": 5.3254437869822495e-05,
"loss": 0.3956,
"mean_token_accuracy": 0.8787799149751663,
"num_tokens": 287370.0,
"step": 90
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.35148611664772034,
"learning_rate": 5.91715976331361e-05,
"loss": 0.3822,
"mean_token_accuracy": 0.8814697057008744,
"num_tokens": 315883.0,
"step": 100
},
{
"epoch": 1.4666666666666668,
"grad_norm": 0.39074671268463135,
"learning_rate": 6.50887573964497e-05,
"loss": 0.3703,
"mean_token_accuracy": 0.8836676925420761,
"num_tokens": 347324.0,
"step": 110
},
{
"epoch": 1.6,
"grad_norm": 0.38130292296409607,
"learning_rate": 7.100591715976331e-05,
"loss": 0.3762,
"mean_token_accuracy": 0.8851484537124634,
"num_tokens": 379374.0,
"step": 120
},
{
"epoch": 1.7333333333333334,
"grad_norm": 0.4091486632823944,
"learning_rate": 7.692307692307693e-05,
"loss": 0.3446,
"mean_token_accuracy": 0.8894414573907852,
"num_tokens": 411843.0,
"step": 130
},
{
"epoch": 1.8666666666666667,
"grad_norm": 0.361017644405365,
"learning_rate": 8.284023668639054e-05,
"loss": 0.3494,
"mean_token_accuracy": 0.8915452778339386,
"num_tokens": 444017.0,
"step": 140
},
{
"epoch": 2.0,
"grad_norm": 0.3487047851085663,
"learning_rate": 8.875739644970414e-05,
"loss": 0.3674,
"mean_token_accuracy": 0.8846240967512131,
"num_tokens": 476022.0,
"step": 150
},
{
"epoch": 2.0,
"eval_loss": 0.3620273470878601,
"eval_mean_token_accuracy": 0.8875497984886169,
"eval_num_tokens": 476022.0,
"eval_runtime": 50.8961,
"eval_samples_per_second": 1.965,
"eval_steps_per_second": 0.491,
"step": 150
},
{
"epoch": 2.1333333333333333,
"grad_norm": 0.480814665555954,
"learning_rate": 9.467455621301775e-05,
"loss": 0.2974,
"mean_token_accuracy": 0.9050114572048187,
"num_tokens": 508445.0,
"step": 160
},
{
"epoch": 2.2666666666666666,
"grad_norm": 0.40775951743125916,
"learning_rate": 9.989539748953976e-05,
"loss": 0.2896,
"mean_token_accuracy": 0.9084265738725662,
"num_tokens": 541530.0,
"step": 170
},
{
"epoch": 2.4,
"grad_norm": 0.5174453854560852,
"learning_rate": 9.884937238493725e-05,
"loss": 0.29,
"mean_token_accuracy": 0.9073476493358612,
"num_tokens": 572212.0,
"step": 180
},
{
"epoch": 2.533333333333333,
"grad_norm": 0.5889921188354492,
"learning_rate": 9.780334728033474e-05,
"loss": 0.2828,
"mean_token_accuracy": 0.909446981549263,
"num_tokens": 603269.0,
"step": 190
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.5635319352149963,
"learning_rate": 9.675732217573223e-05,
"loss": 0.2744,
"mean_token_accuracy": 0.9101487964391708,
"num_tokens": 635792.0,
"step": 200
},
{
"epoch": 2.8,
"grad_norm": 0.609230637550354,
"learning_rate": 9.571129707112972e-05,
"loss": 0.2863,
"mean_token_accuracy": 0.9104649156332016,
"num_tokens": 666470.0,
"step": 210
},
{
"epoch": 2.9333333333333336,
"grad_norm": 0.5209991931915283,
"learning_rate": 9.46652719665272e-05,
"loss": 0.3027,
"mean_token_accuracy": 0.9050683736801147,
"num_tokens": 699119.0,
"step": 220
},
{
"epoch": 3.0,
"eval_loss": 0.36194586753845215,
"eval_mean_token_accuracy": 0.8886684679985046,
"eval_num_tokens": 714033.0,
"eval_runtime": 50.8962,
"eval_samples_per_second": 1.965,
"eval_steps_per_second": 0.491,
"step": 225
},
{
"epoch": 3.066666666666667,
"grad_norm": 0.5214265584945679,
"learning_rate": 9.361924686192469e-05,
"loss": 0.2371,
"mean_token_accuracy": 0.9248212277889252,
"num_tokens": 729520.0,
"step": 230
},
{
"epoch": 3.2,
"grad_norm": 0.7459234595298767,
"learning_rate": 9.257322175732218e-05,
"loss": 0.1874,
"mean_token_accuracy": 0.939895498752594,
"num_tokens": 759510.0,
"step": 240
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.614969789981842,
"learning_rate": 9.152719665271967e-05,
"loss": 0.1754,
"mean_token_accuracy": 0.943790751695633,
"num_tokens": 791875.0,
"step": 250
},
{
"epoch": 3.466666666666667,
"grad_norm": 0.8585807085037231,
"learning_rate": 9.048117154811716e-05,
"loss": 0.1794,
"mean_token_accuracy": 0.9401535421609879,
"num_tokens": 821467.0,
"step": 260
},
{
"epoch": 3.6,
"grad_norm": 0.6798614263534546,
"learning_rate": 8.943514644351465e-05,
"loss": 0.1754,
"mean_token_accuracy": 0.9439484149217605,
"num_tokens": 854000.0,
"step": 270
},
{
"epoch": 3.7333333333333334,
"grad_norm": 0.6594717502593994,
"learning_rate": 8.838912133891214e-05,
"loss": 0.179,
"mean_token_accuracy": 0.9433883309364319,
"num_tokens": 885296.0,
"step": 280
},
{
"epoch": 3.8666666666666667,
"grad_norm": 0.6909582614898682,
"learning_rate": 8.734309623430963e-05,
"loss": 0.1855,
"mean_token_accuracy": 0.9395292073488235,
"num_tokens": 919387.0,
"step": 290
},
{
"epoch": 4.0,
"grad_norm": 0.7755655646324158,
"learning_rate": 8.629707112970712e-05,
"loss": 0.1851,
"mean_token_accuracy": 0.9409447878599166,
"num_tokens": 952044.0,
"step": 300
},
{
"epoch": 4.0,
"eval_loss": 0.39551886916160583,
"eval_mean_token_accuracy": 0.887341833114624,
"eval_num_tokens": 952044.0,
"eval_runtime": 50.8937,
"eval_samples_per_second": 1.965,
"eval_steps_per_second": 0.491,
"step": 300
},
{
"epoch": 4.133333333333334,
"grad_norm": 0.6951614022254944,
"learning_rate": 8.525104602510461e-05,
"loss": 0.0942,
"mean_token_accuracy": 0.9724623709917068,
"num_tokens": 984335.0,
"step": 310
},
{
"epoch": 4.266666666666667,
"grad_norm": 0.6803505420684814,
"learning_rate": 8.42050209205021e-05,
"loss": 0.1073,
"mean_token_accuracy": 0.9653927534818649,
"num_tokens": 1014854.0,
"step": 320
},
{
"epoch": 4.4,
"grad_norm": 0.7284913063049316,
"learning_rate": 8.315899581589958e-05,
"loss": 0.1054,
"mean_token_accuracy": 0.9662943929433823,
"num_tokens": 1047986.0,
"step": 330
},
{
"epoch": 4.533333333333333,
"grad_norm": 0.6803147792816162,
"learning_rate": 8.211297071129707e-05,
"loss": 0.1065,
"mean_token_accuracy": 0.9682819366455078,
"num_tokens": 1079663.0,
"step": 340
},
{
"epoch": 4.666666666666667,
"grad_norm": 0.6864665150642395,
"learning_rate": 8.106694560669456e-05,
"loss": 0.1085,
"mean_token_accuracy": 0.9646553307771683,
"num_tokens": 1109495.0,
"step": 350
},
{
"epoch": 4.8,
"grad_norm": 0.7594568729400635,
"learning_rate": 8.002092050209205e-05,
"loss": 0.1079,
"mean_token_accuracy": 0.9674631506204605,
"num_tokens": 1141598.0,
"step": 360
},
{
"epoch": 4.933333333333334,
"grad_norm": 0.6360507607460022,
"learning_rate": 7.897489539748954e-05,
"loss": 0.118,
"mean_token_accuracy": 0.9628556370735168,
"num_tokens": 1172834.0,
"step": 370
},
{
"epoch": 5.0,
"eval_loss": 0.45760437846183777,
"eval_mean_token_accuracy": 0.8874933505058289,
"eval_num_tokens": 1190055.0,
"eval_runtime": 50.9015,
"eval_samples_per_second": 1.965,
"eval_steps_per_second": 0.491,
"step": 375
}
],
"logging_steps": 10,
"max_steps": 1125,
"num_input_tokens_seen": 0,
"num_train_epochs": 15,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.10032851682345e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}