| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 100, | |
| "global_step": 370, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 31.31255863547505, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 1.5205, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 10.29562052531629, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 1.4439, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.285096223172462, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 1.3513, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.503641754079376, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 1.2532, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.068221865118029, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 1.2089, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 3.8587372955314843, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 1.1817, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 3.73116376117763, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 1.1559, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 3.832939537050073, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 1.1344, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 3.747693244279986, | |
| "learning_rate": 1.999599507118322e-05, | |
| "loss": 1.1199, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.68684932685565, | |
| "learning_rate": 1.9971532122280466e-05, | |
| "loss": 1.0902, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 3.694485690651995, | |
| "learning_rate": 1.992488554155135e-05, | |
| "loss": 1.0918, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.9192389906538514, | |
| "learning_rate": 1.9856159103477085e-05, | |
| "loss": 1.0677, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.395113740791023, | |
| "learning_rate": 1.9765505703518494e-05, | |
| "loss": 1.0449, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 55.64552490454145, | |
| "learning_rate": 1.9653127017970035e-05, | |
| "loss": 1.0412, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 5.907895975695113, | |
| "learning_rate": 1.9519273055291266e-05, | |
| "loss": 1.0291, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 16.980325045210197, | |
| "learning_rate": 1.9364241599913923e-05, | |
| "loss": 1.2227, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 26.879221211035098, | |
| "learning_rate": 1.9188377549761962e-05, | |
| "loss": 1.1577, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 17.465078721382948, | |
| "learning_rate": 1.8992072148958368e-05, | |
| "loss": 1.0105, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 4.258708985650342, | |
| "learning_rate": 1.8775762117425777e-05, | |
| "loss": 0.9719, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 3.9262468540271627, | |
| "learning_rate": 1.853992867931721e-05, | |
| "loss": 0.9714, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 4.127884777228108, | |
| "learning_rate": 1.8285096492438424e-05, | |
| "loss": 0.9227, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "eval_loss": 1.3365702629089355, | |
| "eval_runtime": 17.7256, | |
| "eval_samples_per_second": 33.454, | |
| "eval_steps_per_second": 1.072, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.6876289826671718, | |
| "learning_rate": 1.8011832481043577e-05, | |
| "loss": 0.9179, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.5580612433526255, | |
| "learning_rate": 1.7720744574600865e-05, | |
| "loss": 0.8885, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.840600055255248, | |
| "learning_rate": 1.7412480355334006e-05, | |
| "loss": 0.8921, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.014232990384134, | |
| "learning_rate": 1.7087725617548385e-05, | |
| "loss": 0.8761, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.362002410991992, | |
| "learning_rate": 1.6747202841946928e-05, | |
| "loss": 0.8497, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.270370211049122, | |
| "learning_rate": 1.639166958832985e-05, | |
| "loss": 0.8659, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.2957573116367196, | |
| "learning_rate": 1.6021916810254096e-05, | |
| "loss": 0.8237, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.968603893718009, | |
| "learning_rate": 1.5638767095401778e-05, | |
| "loss": 0.8345, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 3.458171302980965, | |
| "learning_rate": 1.5243072835572319e-05, | |
| "loss": 0.8189, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.695152856632861, | |
| "learning_rate": 1.4835714330369445e-05, | |
| "loss": 0.8028, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.148345741206558, | |
| "learning_rate": 1.4417597828801833e-05, | |
| "loss": 0.812, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 3.2188121700972543, | |
| "learning_rate": 1.3989653513154165e-05, | |
| "loss": 0.7882, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 3.4338055838751536, | |
| "learning_rate": 1.3552833429613939e-05, | |
| "loss": 0.7801, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.6193578206738666, | |
| "learning_rate": 1.3108109370257714e-05, | |
| "loss": 0.7733, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.313724488690459, | |
| "learning_rate": 1.2656470711108763e-05, | |
| "loss": 0.7493, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 3.1599413989481864, | |
| "learning_rate": 1.2198922211075779e-05, | |
| "loss": 0.7459, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.2871458343993627, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.712, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 3.4396952311119473, | |
| "learning_rate": 1.1270178197468788e-05, | |
| "loss": 0.3904, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 3.8937788749830875, | |
| "learning_rate": 1.080104885737807e-05, | |
| "loss": 0.3743, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 3.6367405057812885, | |
| "learning_rate": 1.0330137426761136e-05, | |
| "loss": 0.3659, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "eval_loss": 1.4554423093795776, | |
| "eval_runtime": 16.8492, | |
| "eval_samples_per_second": 35.195, | |
| "eval_steps_per_second": 1.128, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 3.6071182321442823, | |
| "learning_rate": 9.858491540592383e-06, | |
| "loss": 0.3717, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 3.0993281926679, | |
| "learning_rate": 9.38716046778684e-06, | |
| "loss": 0.3657, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 4.1194373147335, | |
| "learning_rate": 8.917192776895382e-06, | |
| "loss": 0.368, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 3.0585597878049553, | |
| "learning_rate": 8.449634003358022e-06, | |
| "loss": 0.3613, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 3.0753210269434197, | |
| "learning_rate": 7.985524323504948e-06, | |
| "loss": 0.3539, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 2.922314637903982, | |
| "learning_rate": 7.525896240479977e-06, | |
| "loss": 0.3594, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 3.2882731601562942, | |
| "learning_rate": 7.071772287234497e-06, | |
| "loss": 0.3469, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.9886234420402147, | |
| "learning_rate": 6.624162751702077e-06, | |
| "loss": 0.3513, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 3.0577844685937214, | |
| "learning_rate": 6.184063429214515e-06, | |
| "loss": 0.3432, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 17.4039721868179, | |
| "learning_rate": 5.752453407159521e-06, | |
| "loss": 0.3573, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 3.2984407603690156, | |
| "learning_rate": 5.33029288680852e-06, | |
| "loss": 0.366, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 3.1677473809022634, | |
| "learning_rate": 4.918521047160309e-06, | |
| "loss": 0.3602, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 3.0805955031351195, | |
| "learning_rate": 4.518053955552903e-06, | |
| "loss": 0.3487, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.9525772320976666, | |
| "learning_rate": 4.1297825296918145e-06, | |
| "loss": 0.3533, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 2.719592667191477, | |
| "learning_rate": 3.754570555628613e-06, | |
| "loss": 0.319, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 2.862304557924882, | |
| "learning_rate": 3.3932527660991877e-06, | |
| "loss": 0.3238, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 2.876259708390993, | |
| "learning_rate": 3.0466329834968234e-06, | |
| "loss": 0.337, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 2.833727652532009, | |
| "learning_rate": 2.715482331611393e-06, | |
| "loss": 0.3315, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 3.160644601967737, | |
| "learning_rate": 2.4005375201130275e-06, | |
| "loss": 0.3334, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 2.908187326503437, | |
| "learning_rate": 2.102499205596743e-06, | |
| "loss": 0.3187, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "eval_loss": 1.4746283292770386, | |
| "eval_runtime": 16.7664, | |
| "eval_samples_per_second": 35.368, | |
| "eval_steps_per_second": 1.133, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.8504402811606853, | |
| "learning_rate": 1.8220304328342253e-06, | |
| "loss": 0.32, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.7717294675757183, | |
| "learning_rate": 1.5597551597004968e-06, | |
| "loss": 0.3187, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.778864670191994, | |
| "learning_rate": 1.3162568690570743e-06, | |
| "loss": 0.3187, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 3.88092742516773, | |
| "learning_rate": 1.0920772706797166e-06, | |
| "loss": 0.3131, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.825202349538424, | |
| "learning_rate": 8.87715096118642e-07, | |
| "loss": 0.3096, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 2.8002712016640277, | |
| "learning_rate": 7.03624989172228e-07, | |
| "loss": 0.3058, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 2.9740828096451057, | |
| "learning_rate": 5.402164944425758e-07, | |
| "loss": 0.3041, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.780072576121478, | |
| "learning_rate": 3.97853146223105e-07, | |
| "loss": 0.3133, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 3.0969397831376018, | |
| "learning_rate": 2.7685165974510987e-07, | |
| "loss": 0.3179, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 2.7014720778740218, | |
| "learning_rate": 1.7748122658251877e-07, | |
| "loss": 0.3043, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.837315264450012, | |
| "learning_rate": 9.996291578236228e-08, | |
| "loss": 0.3143, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 2.8950081674485992, | |
| "learning_rate": 4.44691820532539e-08, | |
| "loss": 0.31, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.570055503485132, | |
| "learning_rate": 1.1123482106021322e-08, | |
| "loss": 0.3036, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.616680349359466, | |
| "learning_rate": 0.0, | |
| "loss": 0.3027, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 370, | |
| "total_flos": 19367618150400.0, | |
| "train_loss": 0.660463113075978, | |
| "train_runtime": 1759.0522, | |
| "train_samples_per_second": 6.713, | |
| "train_steps_per_second": 0.21 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 370, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "total_flos": 19367618150400.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |