leoLM-13b-sft-lora-oa / trainer_state.json
florentgbelidji's picture
Model save
8abf5a8
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4793315743183817,
"eval_steps": 500,
"global_step": 545,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.999996182768104e-05,
"loss": 1.6889,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 1.9999045706597178e-05,
"loss": 1.5727,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 1.999618300852388e-05,
"loss": 1.5409,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 1.999141245215089e-05,
"loss": 1.4338,
"step": 15
},
{
"epoch": 0.02,
"learning_rate": 1.99847349479803e-05,
"loss": 1.5329,
"step": 20
},
{
"epoch": 0.02,
"learning_rate": 1.9976151770471746e-05,
"loss": 1.4317,
"step": 25
},
{
"epoch": 0.03,
"learning_rate": 1.9965664557799164e-05,
"loss": 1.518,
"step": 30
},
{
"epoch": 0.03,
"learning_rate": 1.9953275311538124e-05,
"loss": 1.4521,
"step": 35
},
{
"epoch": 0.04,
"learning_rate": 1.993898639628382e-05,
"loss": 1.46,
"step": 40
},
{
"epoch": 0.04,
"learning_rate": 1.992280053919977e-05,
"loss": 1.3848,
"step": 45
},
{
"epoch": 0.04,
"learning_rate": 1.99047208294973e-05,
"loss": 1.3426,
"step": 50
},
{
"epoch": 0.05,
"learning_rate": 1.9884750717845945e-05,
"loss": 1.3429,
"step": 55
},
{
"epoch": 0.05,
"learning_rate": 1.9862894015714866e-05,
"loss": 1.3541,
"step": 60
},
{
"epoch": 0.06,
"learning_rate": 1.9839154894645393e-05,
"loss": 1.309,
"step": 65
},
{
"epoch": 0.06,
"learning_rate": 1.9813537885454854e-05,
"loss": 1.3096,
"step": 70
},
{
"epoch": 0.07,
"learning_rate": 1.9786047877371823e-05,
"loss": 1.2657,
"step": 75
},
{
"epoch": 0.07,
"learning_rate": 1.975669011710297e-05,
"loss": 1.3209,
"step": 80
},
{
"epoch": 0.07,
"learning_rate": 1.972547020783168e-05,
"loss": 1.2896,
"step": 85
},
{
"epoch": 0.08,
"learning_rate": 1.969239410814865e-05,
"loss": 1.2634,
"step": 90
},
{
"epoch": 0.08,
"learning_rate": 1.9657468130914626e-05,
"loss": 1.2756,
"step": 95
},
{
"epoch": 0.09,
"learning_rate": 1.962069894205553e-05,
"loss": 1.2675,
"step": 100
},
{
"epoch": 0.09,
"learning_rate": 1.9582093559290242e-05,
"loss": 1.2907,
"step": 105
},
{
"epoch": 0.1,
"learning_rate": 1.9541659350791174e-05,
"loss": 1.2533,
"step": 110
},
{
"epoch": 0.1,
"learning_rate": 1.9499404033778016e-05,
"loss": 1.2639,
"step": 115
},
{
"epoch": 0.11,
"learning_rate": 1.9455335673044817e-05,
"loss": 1.3001,
"step": 120
},
{
"epoch": 0.11,
"learning_rate": 1.9409462679420757e-05,
"loss": 1.2534,
"step": 125
},
{
"epoch": 0.11,
"learning_rate": 1.9361793808164878e-05,
"loss": 1.3083,
"step": 130
},
{
"epoch": 0.12,
"learning_rate": 1.931233815729505e-05,
"loss": 1.2891,
"step": 135
},
{
"epoch": 0.12,
"learning_rate": 1.9261105165851546e-05,
"loss": 1.2565,
"step": 140
},
{
"epoch": 0.13,
"learning_rate": 1.920810461209551e-05,
"loss": 1.265,
"step": 145
},
{
"epoch": 0.13,
"learning_rate": 1.9153346611642706e-05,
"loss": 1.253,
"step": 150
},
{
"epoch": 0.14,
"learning_rate": 1.909684161553285e-05,
"loss": 1.299,
"step": 155
},
{
"epoch": 0.14,
"learning_rate": 1.903860040823494e-05,
"loss": 1.2302,
"step": 160
},
{
"epoch": 0.15,
"learning_rate": 1.8978634105588963e-05,
"loss": 1.2977,
"step": 165
},
{
"epoch": 0.15,
"learning_rate": 1.8916954152684315e-05,
"loss": 1.2638,
"step": 170
},
{
"epoch": 0.15,
"learning_rate": 1.8853572321675428e-05,
"loss": 1.2513,
"step": 175
},
{
"epoch": 0.16,
"learning_rate": 1.8788500709534934e-05,
"loss": 1.2284,
"step": 180
},
{
"epoch": 0.16,
"learning_rate": 1.8721751735744873e-05,
"loss": 1.234,
"step": 185
},
{
"epoch": 0.17,
"learning_rate": 1.8653338139926313e-05,
"loss": 1.2946,
"step": 190
},
{
"epoch": 0.17,
"learning_rate": 1.8583272979407885e-05,
"loss": 1.2923,
"step": 195
},
{
"epoch": 0.18,
"learning_rate": 1.8511569626733673e-05,
"loss": 1.3634,
"step": 200
},
{
"epoch": 0.18,
"learning_rate": 1.8438241767110972e-05,
"loss": 1.3127,
"step": 205
},
{
"epoch": 0.18,
"learning_rate": 1.8363303395798305e-05,
"loss": 1.2428,
"step": 210
},
{
"epoch": 0.19,
"learning_rate": 1.828676881543435e-05,
"loss": 1.2556,
"step": 215
},
{
"epoch": 0.19,
"learning_rate": 1.8208652633308136e-05,
"loss": 1.2945,
"step": 220
},
{
"epoch": 0.2,
"learning_rate": 1.812896975857111e-05,
"loss": 1.2345,
"step": 225
},
{
"epoch": 0.2,
"learning_rate": 1.804773539939161e-05,
"loss": 1.2595,
"step": 230
},
{
"epoch": 0.21,
"learning_rate": 1.7964965060052243e-05,
"loss": 1.3029,
"step": 235
},
{
"epoch": 0.21,
"learning_rate": 1.788067453799077e-05,
"loss": 1.2565,
"step": 240
},
{
"epoch": 0.22,
"learning_rate": 1.7794879920785015e-05,
"loss": 1.2432,
"step": 245
},
{
"epoch": 0.22,
"learning_rate": 1.770759758308241e-05,
"loss": 1.2086,
"step": 250
},
{
"epoch": 0.22,
"learning_rate": 1.7618844183474775e-05,
"loss": 1.2351,
"step": 255
},
{
"epoch": 0.23,
"learning_rate": 1.752863666131885e-05,
"loss": 1.2635,
"step": 260
},
{
"epoch": 0.23,
"learning_rate": 1.7436992233503288e-05,
"loss": 1.1707,
"step": 265
},
{
"epoch": 0.24,
"learning_rate": 1.7343928391162673e-05,
"loss": 1.2742,
"step": 270
},
{
"epoch": 0.24,
"learning_rate": 1.7249462896339153e-05,
"loss": 1.2808,
"step": 275
},
{
"epoch": 0.25,
"learning_rate": 1.7153613778592435e-05,
"loss": 1.2544,
"step": 280
},
{
"epoch": 0.25,
"learning_rate": 1.705639933155866e-05,
"loss": 1.2833,
"step": 285
},
{
"epoch": 0.26,
"learning_rate": 1.695783810945892e-05,
"loss": 1.2681,
"step": 290
},
{
"epoch": 0.26,
"learning_rate": 1.685794892355803e-05,
"loss": 1.2044,
"step": 295
},
{
"epoch": 0.26,
"learning_rate": 1.6756750838574197e-05,
"loss": 1.225,
"step": 300
},
{
"epoch": 0.27,
"learning_rate": 1.6654263169040413e-05,
"loss": 1.272,
"step": 305
},
{
"epoch": 0.27,
"learning_rate": 1.6550505475618054e-05,
"loss": 1.1861,
"step": 310
},
{
"epoch": 0.28,
"learning_rate": 1.644549756136358e-05,
"loss": 1.2229,
"step": 315
},
{
"epoch": 0.28,
"learning_rate": 1.6339259467948965e-05,
"loss": 1.2541,
"step": 320
},
{
"epoch": 0.29,
"learning_rate": 1.6231811471836535e-05,
"loss": 1.2025,
"step": 325
},
{
"epoch": 0.29,
"learning_rate": 1.6123174080409055e-05,
"loss": 1.2166,
"step": 330
},
{
"epoch": 0.29,
"learning_rate": 1.6013368028055724e-05,
"loss": 1.249,
"step": 335
},
{
"epoch": 0.3,
"learning_rate": 1.5902414272214804e-05,
"loss": 1.2098,
"step": 340
},
{
"epoch": 0.3,
"learning_rate": 1.5790333989373738e-05,
"loss": 1.2635,
"step": 345
},
{
"epoch": 0.31,
"learning_rate": 1.567714857102743e-05,
"loss": 1.2451,
"step": 350
},
{
"epoch": 0.31,
"learning_rate": 1.5562879619595486e-05,
"loss": 1.2561,
"step": 355
},
{
"epoch": 0.32,
"learning_rate": 1.5447548944299203e-05,
"loss": 1.2773,
"step": 360
},
{
"epoch": 0.32,
"learning_rate": 1.5331178556999094e-05,
"loss": 1.2215,
"step": 365
},
{
"epoch": 0.33,
"learning_rate": 1.5213790667993742e-05,
"loss": 1.1732,
"step": 370
},
{
"epoch": 0.33,
"learning_rate": 1.5095407681780753e-05,
"loss": 1.2401,
"step": 375
},
{
"epoch": 0.33,
"learning_rate": 1.497605219278068e-05,
"loss": 1.2051,
"step": 380
},
{
"epoch": 0.34,
"learning_rate": 1.4855746981024667e-05,
"loss": 1.2675,
"step": 385
},
{
"epoch": 0.34,
"learning_rate": 1.4734515007806698e-05,
"loss": 1.2195,
"step": 390
},
{
"epoch": 0.35,
"learning_rate": 1.4612379411301225e-05,
"loss": 1.2506,
"step": 395
},
{
"epoch": 0.35,
"learning_rate": 1.4489363502147045e-05,
"loss": 1.2441,
"step": 400
},
{
"epoch": 0.36,
"learning_rate": 1.4365490758998268e-05,
"loss": 1.2906,
"step": 405
},
{
"epoch": 0.36,
"learning_rate": 1.424078482404321e-05,
"loss": 1.186,
"step": 410
},
{
"epoch": 0.36,
"learning_rate": 1.4115269498492075e-05,
"loss": 1.2911,
"step": 415
},
{
"epoch": 0.37,
"learning_rate": 1.3988968738034285e-05,
"loss": 1.2135,
"step": 420
},
{
"epoch": 0.37,
"learning_rate": 1.3861906648266339e-05,
"loss": 1.2365,
"step": 425
},
{
"epoch": 0.38,
"learning_rate": 1.3734107480091041e-05,
"loss": 1.2167,
"step": 430
},
{
"epoch": 0.38,
"learning_rate": 1.3605595625089006e-05,
"loss": 1.228,
"step": 435
},
{
"epoch": 0.39,
"learning_rate": 1.3476395610863314e-05,
"loss": 1.2487,
"step": 440
},
{
"epoch": 0.39,
"learning_rate": 1.3346532096358206e-05,
"loss": 1.2096,
"step": 445
},
{
"epoch": 0.4,
"learning_rate": 1.3216029867152724e-05,
"loss": 1.2412,
"step": 450
},
{
"epoch": 0.4,
"learning_rate": 1.308491383073014e-05,
"loss": 1.2388,
"step": 455
},
{
"epoch": 0.4,
"learning_rate": 1.2953209011724159e-05,
"loss": 1.1793,
"step": 460
},
{
"epoch": 0.41,
"learning_rate": 1.2820940547142773e-05,
"loss": 1.241,
"step": 465
},
{
"epoch": 0.41,
"learning_rate": 1.2688133681570604e-05,
"loss": 1.1933,
"step": 470
},
{
"epoch": 0.42,
"learning_rate": 1.255481376235079e-05,
"loss": 1.2435,
"step": 475
},
{
"epoch": 0.42,
"learning_rate": 1.2421006234747202e-05,
"loss": 1.2284,
"step": 480
},
{
"epoch": 0.43,
"learning_rate": 1.2286736637088012e-05,
"loss": 1.2461,
"step": 485
},
{
"epoch": 0.43,
"learning_rate": 1.2152030595891467e-05,
"loss": 1.2313,
"step": 490
},
{
"epoch": 0.44,
"learning_rate": 1.2016913820974855e-05,
"loss": 1.2388,
"step": 495
},
{
"epoch": 0.44,
"learning_rate": 1.1881412100547558e-05,
"loss": 1.2047,
"step": 500
},
{
"epoch": 0.44,
"learning_rate": 1.1745551296289151e-05,
"loss": 1.2011,
"step": 505
},
{
"epoch": 0.45,
"learning_rate": 1.1609357338413476e-05,
"loss": 1.1965,
"step": 510
},
{
"epoch": 0.45,
"learning_rate": 1.147285622071963e-05,
"loss": 1.259,
"step": 515
},
{
"epoch": 0.46,
"learning_rate": 1.133607399563084e-05,
"loss": 1.2615,
"step": 520
},
{
"epoch": 0.46,
"learning_rate": 1.1199036769222104e-05,
"loss": 1.201,
"step": 525
},
{
"epoch": 0.47,
"learning_rate": 1.1061770696237648e-05,
"loss": 1.2445,
"step": 530
},
{
"epoch": 0.47,
"learning_rate": 1.0924301975099043e-05,
"loss": 1.2002,
"step": 535
},
{
"epoch": 0.47,
"learning_rate": 1.0786656842905028e-05,
"loss": 1.2267,
"step": 540
},
{
"epoch": 0.48,
"learning_rate": 1.0648861570423919e-05,
"loss": 1.2257,
"step": 545
},
{
"epoch": 0.48,
"step": 545,
"total_flos": 6.944922061693256e+17,
"train_loss": 1.2715962869858524,
"train_runtime": 6114.717,
"train_samples_per_second": 2.975,
"train_steps_per_second": 0.186
}
],
"logging_steps": 5,
"max_steps": 1137,
"num_train_epochs": 1,
"save_steps": 200,
"total_flos": 6.944922061693256e+17,
"trial_name": null,
"trial_params": null
}