checkpoint-800/trainer_state.json

{
 "best_metric": NaN,
 "best_model_checkpoint": "japanese-lora-alpaca/checkpoint-200",
 "epoch": 2.051195904327654,
 "global_step": 800,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.05,
 "learning_rate": 0.0,
 "loss": 1.9688795594149578e+30,
 "step": 20
 },
 {
 "epoch": 0.1,
 "learning_rate": 0.0,
 "loss": 1.317319656026326e+31,
 "step": 40
 },
 {
 "epoch": 0.15,
 "learning_rate": 0.0,
 "loss": 1.431492931568305e+30,
 "step": 60
 },
 {
 "epoch": 0.2,
 "learning_rate": 0.0,
 "loss": 1.2596809078781474e+29,
 "step": 80
 },
 {
 "epoch": 0.26,
 "learning_rate": 0.0,
 "loss": 7.78457860626805e+29,
 "step": 100
 },
 {
 "epoch": 0.31,
 "learning_rate": 0.0,
 "loss": 4.3205824045723626e+29,
 "step": 120
 },
 {
 "epoch": 0.36,
 "learning_rate": 0.0,
 "loss": 4.2652842037741146e+26,
 "step": 140
 },
 {
 "epoch": 0.41,
 "learning_rate": 0.0,
 "loss": 1.2042161347636073e+31,
 "step": 160
 },
 {
 "epoch": 0.46,
 "learning_rate": 4.2e-05,
 "loss": 5.4732492073607035e+32,
 "step": 180
 },
 {
 "epoch": 0.51,
 "learning_rate": 0.000102,
 "loss": 5.026698446384883e+28,
 "step": 200
 },
 {
 "epoch": 0.51,
 "eval_loss": NaN,
 "eval_runtime": 104.6208,
 "eval_samples_per_second": 19.117,
 "eval_steps_per_second": 2.39,
 "step": 200
 },
 {
 "epoch": 0.56,
 "learning_rate": 0.000162,
 "loss": 3.527059212612975e+30,
 "step": 220
 },
 {
 "epoch": 0.61,
 "learning_rate": 0.00022199999999999998,
 "loss": 3.552828977615335e+29,
 "step": 240
 },
 {
 "epoch": 0.67,
 "learning_rate": 0.00028199999999999997,
 "loss": 1.9742319651445058e+28,
 "step": 260
 },
 {
 "epoch": 0.72,
 "learning_rate": 0.00029607476635514017,
 "loss": 2.2754181268717054e+31,
 "step": 280
 },
 {
 "epoch": 0.77,
 "learning_rate": 0.00029046728971962615,
 "loss": 2.8977228958522534e+30,
 "step": 300
 },
 {
 "epoch": 0.82,
 "learning_rate": 0.00028485981308411213,
 "loss": 9.199314224149786e+28,
 "step": 320
 },
 {
 "epoch": 0.87,
 "learning_rate": 0.0002792523364485981,
 "loss": 4.467849818908903e+28,
 "step": 340
 },
 {
 "epoch": 0.92,
 "learning_rate": 0.0002736448598130841,
 "loss": 7.683142355554883e+30,
 "step": 360
 },
 {
 "epoch": 0.97,
 "learning_rate": 0.00026803738317757007,
 "loss": 8.47695762399229e+28,
 "step": 380
 },
 {
 "epoch": 1.03,
 "learning_rate": 0.00026242990654205605,
 "loss": 5.750158191352777e+28,
 "step": 400
 },
 {
 "epoch": 1.03,
 "eval_loss": NaN,
 "eval_runtime": 104.8725,
 "eval_samples_per_second": 19.071,
 "eval_steps_per_second": 2.384,
 "step": 400
 },
 {
 "epoch": 1.08,
 "learning_rate": 0.00025682242990654204,
 "loss": 1.511076131417274e+31,
 "step": 420
 },
 {
 "epoch": 1.13,
 "learning_rate": 0.000251214953271028,
 "loss": 8.163499842142527e+30,
 "step": 440
 },
 {
 "epoch": 1.18,
 "learning_rate": 0.000245607476635514,
 "loss": 8.08162943624534e+27,
 "step": 460
 },
 {
 "epoch": 1.23,
 "learning_rate": 0.00023999999999999998,
 "loss": 7.625678726356304e+30,
 "step": 480
 },
 {
 "epoch": 1.28,
 "learning_rate": 0.00023439252336448596,
 "loss": 1.6030343069905966e+30,
 "step": 500
 },
 {
 "epoch": 1.33,
 "learning_rate": 0.00022878504672897194,
 "loss": 1.5262935093501894e+30,
 "step": 520
 },
 {
 "epoch": 1.38,
 "learning_rate": 0.00022317757009345792,
 "loss": 7.634831987306935e+28,
 "step": 540
 },
 {
 "epoch": 1.44,
 "learning_rate": 0.0002175700934579439,
 "loss": 3.5534372486414743e+31,
 "step": 560
 },
 {
 "epoch": 1.49,
 "learning_rate": 0.00021196261682242989,
 "loss": 2.4269485037904036e+29,
 "step": 580
 },
 {
 "epoch": 1.54,
 "learning_rate": 0.00020635514018691587,
 "loss": 5.4389742841217565e+31,
 "step": 600
 },
 {
 "epoch": 1.54,
 "eval_loss": NaN,
 "eval_runtime": 104.2474,
 "eval_samples_per_second": 19.185,
 "eval_steps_per_second": 2.398,
 "step": 600
 },
 {
 "epoch": 1.59,
 "learning_rate": 0.00020074766355140185,
 "loss": 1.1437688736199055e+28,
 "step": 620
 },
 {
 "epoch": 1.64,
 "learning_rate": 0.00019514018691588783,
 "loss": 3.8293792591118507e+27,
 "step": 640
 },
 {
 "epoch": 1.69,
 "learning_rate": 0.0001895327102803738,
 "loss": 6.2081077458410555e+32,
 "step": 660
 },
 {
 "epoch": 1.74,
 "learning_rate": 0.0001839252336448598,
 "loss": 8.158099328721144e+30,
 "step": 680
 },
 {
 "epoch": 1.79,
 "learning_rate": 0.00017831775700934577,
 "loss": 1.3757688257315705e+30,
 "step": 700
 },
 {
 "epoch": 1.84,
 "learning_rate": 0.00017271028037383175,
 "loss": 1.0185220430876456e+29,
 "step": 720
 },
 {
 "epoch": 1.9,
 "learning_rate": 0.00016710280373831774,
 "loss": 1.7034122942644185e+29,
 "step": 740
 },
 {
 "epoch": 1.95,
 "learning_rate": 0.00016149532710280372,
 "loss": 2.744296367142522e+28,
 "step": 760
 },
 {
 "epoch": 2.0,
 "learning_rate": 0.0001558878504672897,
 "loss": 2.237018774278122e+31,
 "step": 780
 },
 {
 "epoch": 2.05,
 "learning_rate": 0.00015028037383177568,
 "loss": 6.440347599779466e+34,
 "step": 800
 },
 {
 "epoch": 2.05,
 "eval_loss": NaN,
 "eval_runtime": 104.6953,
 "eval_samples_per_second": 19.103,
 "eval_steps_per_second": 2.388,
 "step": 800
 }
 ],
 "max_steps": 1170,
 "num_train_epochs": 3,
 "total_flos": 1.0415704707484877e+18,
 "trial_name": null,
 "trial_params": null
}