sentiment-lora-r8a1d0.15-1 / trainer_state.json
apwic's picture
End of training
f4297b3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 2440,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"grad_norm": 5.11376428604126,
"learning_rate": 4.75e-05,
"loss": 0.5647,
"step": 122
},
{
"epoch": 1.0,
"eval_accuracy": 0.706766917293233,
"eval_f1": 0.6297262783854312,
"eval_loss": 0.5166164636611938,
"eval_precision": 0.6379598662207357,
"eval_recall": 0.6250227314057101,
"eval_runtime": 5.1151,
"eval_samples_per_second": 78.005,
"eval_steps_per_second": 9.775,
"step": 122
},
{
"epoch": 2.0,
"grad_norm": 4.067341327667236,
"learning_rate": 4.5e-05,
"loss": 0.5067,
"step": 244
},
{
"epoch": 2.0,
"eval_accuracy": 0.7343358395989975,
"eval_f1": 0.6926129426129426,
"eval_loss": 0.4954279363155365,
"eval_precision": 0.6870370370370371,
"eval_recall": 0.7020367339516276,
"eval_runtime": 5.0465,
"eval_samples_per_second": 79.065,
"eval_steps_per_second": 9.908,
"step": 244
},
{
"epoch": 3.0,
"grad_norm": 4.497439861297607,
"learning_rate": 4.25e-05,
"loss": 0.4617,
"step": 366
},
{
"epoch": 3.0,
"eval_accuracy": 0.7919799498746867,
"eval_f1": 0.7496919995465023,
"eval_loss": 0.4390866756439209,
"eval_precision": 0.7490801616502805,
"eval_recall": 0.7503182396799418,
"eval_runtime": 5.0495,
"eval_samples_per_second": 79.017,
"eval_steps_per_second": 9.902,
"step": 366
},
{
"epoch": 4.0,
"grad_norm": 2.5452804565429688,
"learning_rate": 4e-05,
"loss": 0.4044,
"step": 488
},
{
"epoch": 4.0,
"eval_accuracy": 0.8145363408521303,
"eval_f1": 0.7773840400506664,
"eval_loss": 0.39114564657211304,
"eval_precision": 0.7760504201680672,
"eval_recall": 0.7787779596290234,
"eval_runtime": 5.0465,
"eval_samples_per_second": 79.065,
"eval_steps_per_second": 9.908,
"step": 488
},
{
"epoch": 5.0,
"grad_norm": 5.4987263679504395,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.382,
"step": 610
},
{
"epoch": 5.0,
"eval_accuracy": 0.8195488721804511,
"eval_f1": 0.7962206332992849,
"eval_loss": 0.38273975253105164,
"eval_precision": 0.7848639455782312,
"eval_recall": 0.8198308783415167,
"eval_runtime": 5.0801,
"eval_samples_per_second": 78.542,
"eval_steps_per_second": 9.842,
"step": 610
},
{
"epoch": 6.0,
"grad_norm": 6.1503801345825195,
"learning_rate": 3.5e-05,
"loss": 0.3494,
"step": 732
},
{
"epoch": 6.0,
"eval_accuracy": 0.8421052631578947,
"eval_f1": 0.8100071803786705,
"eval_loss": 0.3528314530849457,
"eval_precision": 0.8092466373122624,
"eval_recall": 0.8107837788688852,
"eval_runtime": 5.0684,
"eval_samples_per_second": 78.723,
"eval_steps_per_second": 9.865,
"step": 732
},
{
"epoch": 7.0,
"grad_norm": 1.3256051540374756,
"learning_rate": 3.2500000000000004e-05,
"loss": 0.3423,
"step": 854
},
{
"epoch": 7.0,
"eval_accuracy": 0.8546365914786967,
"eval_f1": 0.8255172205802521,
"eval_loss": 0.3441900908946991,
"eval_precision": 0.8239495798319327,
"eval_recall": 0.8271503909801782,
"eval_runtime": 5.0527,
"eval_samples_per_second": 78.968,
"eval_steps_per_second": 9.896,
"step": 854
},
{
"epoch": 8.0,
"grad_norm": 3.73527455329895,
"learning_rate": 3e-05,
"loss": 0.33,
"step": 976
},
{
"epoch": 8.0,
"eval_accuracy": 0.8671679197994987,
"eval_f1": 0.8341632880321839,
"eval_loss": 0.3399864733219147,
"eval_precision": 0.8479139504563233,
"eval_recall": 0.8235133660665576,
"eval_runtime": 5.0846,
"eval_samples_per_second": 78.473,
"eval_steps_per_second": 9.834,
"step": 976
},
{
"epoch": 9.0,
"grad_norm": 7.524757385253906,
"learning_rate": 2.7500000000000004e-05,
"loss": 0.3296,
"step": 1098
},
{
"epoch": 9.0,
"eval_accuracy": 0.849624060150376,
"eval_f1": 0.8127815315315315,
"eval_loss": 0.3349246084690094,
"eval_precision": 0.8244897959183674,
"eval_recall": 0.8036006546644845,
"eval_runtime": 5.0487,
"eval_samples_per_second": 79.03,
"eval_steps_per_second": 9.904,
"step": 1098
},
{
"epoch": 10.0,
"grad_norm": 6.317535877227783,
"learning_rate": 2.5e-05,
"loss": 0.3074,
"step": 1220
},
{
"epoch": 10.0,
"eval_accuracy": 0.8621553884711779,
"eval_f1": 0.8249232119350592,
"eval_loss": 0.33487534523010254,
"eval_precision": 0.8467014712861889,
"eval_recall": 0.8099654482633206,
"eval_runtime": 5.0732,
"eval_samples_per_second": 78.649,
"eval_steps_per_second": 9.856,
"step": 1220
},
{
"epoch": 11.0,
"grad_norm": 3.027658224105835,
"learning_rate": 2.25e-05,
"loss": 0.2911,
"step": 1342
},
{
"epoch": 11.0,
"eval_accuracy": 0.8696741854636592,
"eval_f1": 0.8377439939939939,
"eval_loss": 0.32399529218673706,
"eval_precision": 0.8503401360544218,
"eval_recall": 0.8277868703400618,
"eval_runtime": 5.0441,
"eval_samples_per_second": 79.102,
"eval_steps_per_second": 9.913,
"step": 1342
},
{
"epoch": 12.0,
"grad_norm": 9.344508171081543,
"learning_rate": 2e-05,
"loss": 0.2855,
"step": 1464
},
{
"epoch": 12.0,
"eval_accuracy": 0.8721804511278195,
"eval_f1": 0.8454251965513313,
"eval_loss": 0.3273135721683502,
"eval_precision": 0.8463049835506276,
"eval_recall": 0.8445626477541371,
"eval_runtime": 5.049,
"eval_samples_per_second": 79.026,
"eval_steps_per_second": 9.903,
"step": 1464
},
{
"epoch": 13.0,
"grad_norm": 4.517082214355469,
"learning_rate": 1.75e-05,
"loss": 0.2903,
"step": 1586
},
{
"epoch": 13.0,
"eval_accuracy": 0.8646616541353384,
"eval_f1": 0.8295950648528947,
"eval_loss": 0.32846182584762573,
"eval_precision": 0.8472157618446409,
"eval_recall": 0.816739407164939,
"eval_runtime": 5.0968,
"eval_samples_per_second": 78.285,
"eval_steps_per_second": 9.81,
"step": 1586
},
{
"epoch": 14.0,
"grad_norm": 8.310128211975098,
"learning_rate": 1.5e-05,
"loss": 0.2896,
"step": 1708
},
{
"epoch": 14.0,
"eval_accuracy": 0.8671679197994987,
"eval_f1": 0.8341632880321839,
"eval_loss": 0.3254058063030243,
"eval_precision": 0.8479139504563233,
"eval_recall": 0.8235133660665576,
"eval_runtime": 5.0513,
"eval_samples_per_second": 78.99,
"eval_steps_per_second": 9.898,
"step": 1708
},
{
"epoch": 15.0,
"grad_norm": 1.1135729551315308,
"learning_rate": 1.25e-05,
"loss": 0.2744,
"step": 1830
},
{
"epoch": 15.0,
"eval_accuracy": 0.8646616541353384,
"eval_f1": 0.8359175094431583,
"eval_loss": 0.3240545392036438,
"eval_precision": 0.8376607470912432,
"eval_recall": 0.8342425895617385,
"eval_runtime": 5.0649,
"eval_samples_per_second": 78.778,
"eval_steps_per_second": 9.872,
"step": 1830
},
{
"epoch": 16.0,
"grad_norm": 6.9109649658203125,
"learning_rate": 1e-05,
"loss": 0.2691,
"step": 1952
},
{
"epoch": 16.0,
"eval_accuracy": 0.8571428571428571,
"eval_f1": 0.8263588263588264,
"eval_loss": 0.3209517002105713,
"eval_precision": 0.8289473684210527,
"eval_recall": 0.8239225313693399,
"eval_runtime": 5.0874,
"eval_samples_per_second": 78.429,
"eval_steps_per_second": 9.828,
"step": 1952
},
{
"epoch": 17.0,
"grad_norm": 6.5280585289001465,
"learning_rate": 7.5e-06,
"loss": 0.2671,
"step": 2074
},
{
"epoch": 17.0,
"eval_accuracy": 0.8696741854636592,
"eval_f1": 0.8377439939939939,
"eval_loss": 0.3208070397377014,
"eval_precision": 0.8503401360544218,
"eval_recall": 0.8277868703400618,
"eval_runtime": 5.0622,
"eval_samples_per_second": 78.82,
"eval_steps_per_second": 9.877,
"step": 2074
},
{
"epoch": 18.0,
"grad_norm": 3.9085135459899902,
"learning_rate": 5e-06,
"loss": 0.2736,
"step": 2196
},
{
"epoch": 18.0,
"eval_accuracy": 0.8721804511278195,
"eval_f1": 0.8421640488656195,
"eval_loss": 0.31788739562034607,
"eval_precision": 0.8512313860252005,
"eval_recall": 0.8345608292416803,
"eval_runtime": 5.0487,
"eval_samples_per_second": 79.031,
"eval_steps_per_second": 9.904,
"step": 2196
},
{
"epoch": 19.0,
"grad_norm": 3.626889705657959,
"learning_rate": 2.5e-06,
"loss": 0.2662,
"step": 2318
},
{
"epoch": 19.0,
"eval_accuracy": 0.8721804511278195,
"eval_f1": 0.8404212771630449,
"eval_loss": 0.3179715573787689,
"eval_precision": 0.854416558018253,
"eval_recall": 0.8295599199854519,
"eval_runtime": 5.0563,
"eval_samples_per_second": 78.912,
"eval_steps_per_second": 9.889,
"step": 2318
},
{
"epoch": 20.0,
"grad_norm": 9.089592933654785,
"learning_rate": 0.0,
"loss": 0.2664,
"step": 2440
},
{
"epoch": 20.0,
"eval_accuracy": 0.8721804511278195,
"eval_f1": 0.8413023981282901,
"eval_loss": 0.3167513608932495,
"eval_precision": 0.8527593534677056,
"eval_recall": 0.8320603746135662,
"eval_runtime": 5.0535,
"eval_samples_per_second": 78.955,
"eval_steps_per_second": 9.894,
"step": 2440
},
{
"epoch": 20.0,
"step": 2440,
"total_flos": 7609911792720000.0,
"train_loss": 0.33757916747546585,
"train_runtime": 1952.408,
"train_samples_per_second": 37.267,
"train_steps_per_second": 1.25
}
],
"logging_steps": 500,
"max_steps": 2440,
"num_input_tokens_seen": 0,
"num_train_epochs": 20,
"save_steps": 500,
"total_flos": 7609911792720000.0,
"train_batch_size": 30,
"trial_name": null,
"trial_params": null
}