| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3105, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004830917874396135, | |
| "grad_norm": 0.4590633809566498, | |
| "learning_rate": 1.3986464711569449e-05, | |
| "loss": 3.9553, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00966183574879227, | |
| "grad_norm": 0.562667965888977, | |
| "learning_rate": 1.396390589751853e-05, | |
| "loss": 3.8462, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014492753623188406, | |
| "grad_norm": 0.5178580284118652, | |
| "learning_rate": 1.3941347083467613e-05, | |
| "loss": 3.7563, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01932367149758454, | |
| "grad_norm": 0.549095869064331, | |
| "learning_rate": 1.3918788269416693e-05, | |
| "loss": 3.5661, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.024154589371980676, | |
| "grad_norm": 0.5863974094390869, | |
| "learning_rate": 1.3896229455365775e-05, | |
| "loss": 3.3998, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.028985507246376812, | |
| "grad_norm": 0.6717987060546875, | |
| "learning_rate": 1.3873670641314857e-05, | |
| "loss": 3.2327, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.033816425120772944, | |
| "grad_norm": 0.6742504835128784, | |
| "learning_rate": 1.3851111827263939e-05, | |
| "loss": 3.0419, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03864734299516908, | |
| "grad_norm": 0.7488217949867249, | |
| "learning_rate": 1.382855301321302e-05, | |
| "loss": 2.8092, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 0.7981260418891907, | |
| "learning_rate": 1.3805994199162101e-05, | |
| "loss": 2.5637, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 0.7837016582489014, | |
| "learning_rate": 1.3783435385111182e-05, | |
| "loss": 2.311, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05314009661835749, | |
| "grad_norm": 0.8491142988204956, | |
| "learning_rate": 1.3760876571060263e-05, | |
| "loss": 2.0814, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.057971014492753624, | |
| "grad_norm": 0.7609491348266602, | |
| "learning_rate": 1.3738317757009345e-05, | |
| "loss": 1.7811, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06280193236714976, | |
| "grad_norm": 0.7346836924552917, | |
| "learning_rate": 1.3715758942958427e-05, | |
| "loss": 1.5668, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06763285024154589, | |
| "grad_norm": 0.7201610803604126, | |
| "learning_rate": 1.369320012890751e-05, | |
| "loss": 1.3152, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07246376811594203, | |
| "grad_norm": 0.6400141716003418, | |
| "learning_rate": 1.3670641314856591e-05, | |
| "loss": 1.0742, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07729468599033816, | |
| "grad_norm": 0.38841813802719116, | |
| "learning_rate": 1.3648082500805672e-05, | |
| "loss": 0.9316, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0821256038647343, | |
| "grad_norm": 0.4104098677635193, | |
| "learning_rate": 1.3625523686754754e-05, | |
| "loss": 0.8594, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 0.5033922791481018, | |
| "learning_rate": 1.3602964872703834e-05, | |
| "loss": 0.7873, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09178743961352658, | |
| "grad_norm": 0.3223589360713959, | |
| "learning_rate": 1.3580406058652916e-05, | |
| "loss": 0.7265, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 0.31837838888168335, | |
| "learning_rate": 1.3557847244601998e-05, | |
| "loss": 0.7056, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10144927536231885, | |
| "grad_norm": 0.35547807812690735, | |
| "learning_rate": 1.353528843055108e-05, | |
| "loss": 0.6684, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10628019323671498, | |
| "grad_norm": 0.3624265789985657, | |
| "learning_rate": 1.351272961650016e-05, | |
| "loss": 0.6424, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.37934672832489014, | |
| "learning_rate": 1.3490170802449242e-05, | |
| "loss": 0.6473, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11594202898550725, | |
| "grad_norm": 0.43373095989227295, | |
| "learning_rate": 1.3467611988398324e-05, | |
| "loss": 0.6108, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12077294685990338, | |
| "grad_norm": 0.46837344765663147, | |
| "learning_rate": 1.3445053174347406e-05, | |
| "loss": 0.5718, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12560386473429952, | |
| "grad_norm": 0.4655485153198242, | |
| "learning_rate": 1.3422494360296488e-05, | |
| "loss": 0.5618, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 0.5438677072525024, | |
| "learning_rate": 1.3399935546245569e-05, | |
| "loss": 0.5834, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.13526570048309178, | |
| "grad_norm": 0.5986974239349365, | |
| "learning_rate": 1.3377376732194649e-05, | |
| "loss": 0.51, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14009661835748793, | |
| "grad_norm": 0.7286536693572998, | |
| "learning_rate": 1.3354817918143731e-05, | |
| "loss": 0.4704, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 0.9337557554244995, | |
| "learning_rate": 1.3332259104092813e-05, | |
| "loss": 0.4379, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1497584541062802, | |
| "grad_norm": 1.0971410274505615, | |
| "learning_rate": 1.3309700290041895e-05, | |
| "loss": 0.3994, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.15458937198067632, | |
| "grad_norm": 1.5142974853515625, | |
| "learning_rate": 1.3287141475990977e-05, | |
| "loss": 0.3397, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.15942028985507245, | |
| "grad_norm": 0.5735320448875427, | |
| "learning_rate": 1.3264582661940057e-05, | |
| "loss": 0.3047, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1642512077294686, | |
| "grad_norm": 0.31310656666755676, | |
| "learning_rate": 1.324202384788914e-05, | |
| "loss": 0.2761, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.16908212560386474, | |
| "grad_norm": 0.3159743845462799, | |
| "learning_rate": 1.3219465033838221e-05, | |
| "loss": 0.2584, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.2747821509838104, | |
| "learning_rate": 1.3196906219787303e-05, | |
| "loss": 0.2696, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.178743961352657, | |
| "grad_norm": 0.26819175481796265, | |
| "learning_rate": 1.3174347405736383e-05, | |
| "loss": 0.2633, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.18357487922705315, | |
| "grad_norm": 0.27827367186546326, | |
| "learning_rate": 1.3151788591685465e-05, | |
| "loss": 0.2587, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.18840579710144928, | |
| "grad_norm": 0.30673256516456604, | |
| "learning_rate": 1.3129229777634546e-05, | |
| "loss": 0.2749, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 0.28767552971839905, | |
| "learning_rate": 1.3106670963583628e-05, | |
| "loss": 0.2527, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19806763285024154, | |
| "grad_norm": 0.2788391709327698, | |
| "learning_rate": 1.308411214953271e-05, | |
| "loss": 0.2548, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2028985507246377, | |
| "grad_norm": 0.26774516701698303, | |
| "learning_rate": 1.3061553335481792e-05, | |
| "loss": 0.2426, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.20772946859903382, | |
| "grad_norm": 0.3280729651451111, | |
| "learning_rate": 1.3038994521430874e-05, | |
| "loss": 0.2343, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.21256038647342995, | |
| "grad_norm": 0.28210124373435974, | |
| "learning_rate": 1.3016435707379956e-05, | |
| "loss": 0.2385, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 0.2706020176410675, | |
| "learning_rate": 1.2993876893329036e-05, | |
| "loss": 0.2418, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.2814071476459503, | |
| "learning_rate": 1.2971318079278118e-05, | |
| "loss": 0.2309, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.22705314009661837, | |
| "grad_norm": 0.311310738325119, | |
| "learning_rate": 1.2948759265227198e-05, | |
| "loss": 0.239, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2318840579710145, | |
| "grad_norm": 0.31521573662757874, | |
| "learning_rate": 1.292620045117628e-05, | |
| "loss": 0.2271, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.23671497584541062, | |
| "grad_norm": 0.3365338146686554, | |
| "learning_rate": 1.2903641637125362e-05, | |
| "loss": 0.2418, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 0.32416385412216187, | |
| "learning_rate": 1.2881082823074444e-05, | |
| "loss": 0.2282, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2463768115942029, | |
| "grad_norm": 0.3272862434387207, | |
| "learning_rate": 1.2858524009023525e-05, | |
| "loss": 0.2319, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.25120772946859904, | |
| "grad_norm": 0.34287795424461365, | |
| "learning_rate": 1.2835965194972607e-05, | |
| "loss": 0.2529, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2560386473429952, | |
| "grad_norm": 0.338498055934906, | |
| "learning_rate": 1.2813406380921689e-05, | |
| "loss": 0.2216, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 0.34228625893592834, | |
| "learning_rate": 1.279084756687077e-05, | |
| "loss": 0.2306, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.26570048309178745, | |
| "grad_norm": 0.38022157549858093, | |
| "learning_rate": 1.2768288752819853e-05, | |
| "loss": 0.2091, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.27053140096618356, | |
| "grad_norm": 0.35013625025749207, | |
| "learning_rate": 1.2745729938768935e-05, | |
| "loss": 0.2219, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2753623188405797, | |
| "grad_norm": 0.3915255665779114, | |
| "learning_rate": 1.2723171124718013e-05, | |
| "loss": 0.202, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.28019323671497587, | |
| "grad_norm": 0.4278201758861542, | |
| "learning_rate": 1.2700612310667095e-05, | |
| "loss": 0.2223, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.28502415458937197, | |
| "grad_norm": 0.4377511441707611, | |
| "learning_rate": 1.2678053496616177e-05, | |
| "loss": 0.2001, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 0.44731107354164124, | |
| "learning_rate": 1.265549468256526e-05, | |
| "loss": 0.1884, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2946859903381642, | |
| "grad_norm": 0.4644255042076111, | |
| "learning_rate": 1.2632935868514341e-05, | |
| "loss": 0.195, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2995169082125604, | |
| "grad_norm": 0.46685394644737244, | |
| "learning_rate": 1.2610377054463423e-05, | |
| "loss": 0.1867, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 0.484323650598526, | |
| "learning_rate": 1.2587818240412503e-05, | |
| "loss": 0.1855, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.30917874396135264, | |
| "grad_norm": 0.4667232632637024, | |
| "learning_rate": 1.2565259426361585e-05, | |
| "loss": 0.1823, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3140096618357488, | |
| "grad_norm": 0.5028926134109497, | |
| "learning_rate": 1.2542700612310667e-05, | |
| "loss": 0.1726, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3188405797101449, | |
| "grad_norm": 0.5125951766967773, | |
| "learning_rate": 1.252014179825975e-05, | |
| "loss": 0.1709, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.32367149758454106, | |
| "grad_norm": 0.4960808753967285, | |
| "learning_rate": 1.249758298420883e-05, | |
| "loss": 0.1539, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3285024154589372, | |
| "grad_norm": 0.42437031865119934, | |
| "learning_rate": 1.2475024170157912e-05, | |
| "loss": 0.1484, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.32370465993881226, | |
| "learning_rate": 1.2452465356106992e-05, | |
| "loss": 0.128, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 0.2940502166748047, | |
| "learning_rate": 1.2429906542056074e-05, | |
| "loss": 0.1152, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.34299516908212563, | |
| "grad_norm": 0.3140239715576172, | |
| "learning_rate": 1.2407347728005156e-05, | |
| "loss": 0.1466, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.26790526509284973, | |
| "learning_rate": 1.2384788913954238e-05, | |
| "loss": 0.1148, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3526570048309179, | |
| "grad_norm": 0.276149183511734, | |
| "learning_rate": 1.236223009990332e-05, | |
| "loss": 0.0996, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.357487922705314, | |
| "grad_norm": 0.24558521807193756, | |
| "learning_rate": 1.2339671285852402e-05, | |
| "loss": 0.105, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.36231884057971014, | |
| "grad_norm": 0.31819969415664673, | |
| "learning_rate": 1.2317112471801482e-05, | |
| "loss": 0.0968, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3671497584541063, | |
| "grad_norm": 0.28777143359184265, | |
| "learning_rate": 1.2294553657750564e-05, | |
| "loss": 0.1029, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3719806763285024, | |
| "grad_norm": 0.26321807503700256, | |
| "learning_rate": 1.2271994843699645e-05, | |
| "loss": 0.1018, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.37681159420289856, | |
| "grad_norm": 0.23004086315631866, | |
| "learning_rate": 1.2249436029648727e-05, | |
| "loss": 0.0912, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.38164251207729466, | |
| "grad_norm": 0.21847793459892273, | |
| "learning_rate": 1.2226877215597809e-05, | |
| "loss": 0.0913, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 0.19117720425128937, | |
| "learning_rate": 1.220431840154689e-05, | |
| "loss": 0.0895, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 0.2535386085510254, | |
| "learning_rate": 1.2181759587495971e-05, | |
| "loss": 0.1103, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3961352657004831, | |
| "grad_norm": 0.26542574167251587, | |
| "learning_rate": 1.2159200773445053e-05, | |
| "loss": 0.0923, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.40096618357487923, | |
| "grad_norm": 0.20197734236717224, | |
| "learning_rate": 1.2136641959394135e-05, | |
| "loss": 0.0931, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4057971014492754, | |
| "grad_norm": 0.20673911273479462, | |
| "learning_rate": 1.2114083145343217e-05, | |
| "loss": 0.1138, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4106280193236715, | |
| "grad_norm": 0.24391110241413116, | |
| "learning_rate": 1.2091524331292299e-05, | |
| "loss": 0.094, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.41545893719806765, | |
| "grad_norm": 0.2456451952457428, | |
| "learning_rate": 1.2068965517241379e-05, | |
| "loss": 0.1078, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.42028985507246375, | |
| "grad_norm": 0.29903218150138855, | |
| "learning_rate": 1.204640670319046e-05, | |
| "loss": 0.0999, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4251207729468599, | |
| "grad_norm": 0.17596346139907837, | |
| "learning_rate": 1.2023847889139541e-05, | |
| "loss": 0.0995, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.42995169082125606, | |
| "grad_norm": 0.14841659367084503, | |
| "learning_rate": 1.2001289075088623e-05, | |
| "loss": 0.0934, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 0.18399696052074432, | |
| "learning_rate": 1.1978730261037705e-05, | |
| "loss": 0.0967, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4396135265700483, | |
| "grad_norm": 0.1746302992105484, | |
| "learning_rate": 1.1956171446986787e-05, | |
| "loss": 0.0947, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.2423829287290573, | |
| "learning_rate": 1.1933612632935868e-05, | |
| "loss": 0.0936, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4492753623188406, | |
| "grad_norm": 0.15260176360607147, | |
| "learning_rate": 1.191105381888495e-05, | |
| "loss": 0.1034, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.45410628019323673, | |
| "grad_norm": 0.2334187626838684, | |
| "learning_rate": 1.1888495004834032e-05, | |
| "loss": 0.1053, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.45893719806763283, | |
| "grad_norm": 0.19356365501880646, | |
| "learning_rate": 1.1865936190783114e-05, | |
| "loss": 0.0842, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.463768115942029, | |
| "grad_norm": 0.20395216345787048, | |
| "learning_rate": 1.1843377376732194e-05, | |
| "loss": 0.0792, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.46859903381642515, | |
| "grad_norm": 0.1807161122560501, | |
| "learning_rate": 1.1820818562681276e-05, | |
| "loss": 0.0882, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.47342995169082125, | |
| "grad_norm": 0.16710110008716583, | |
| "learning_rate": 1.1798259748630358e-05, | |
| "loss": 0.0822, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 0.1776697188615799, | |
| "learning_rate": 1.1775700934579438e-05, | |
| "loss": 0.0922, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 0.21817447245121002, | |
| "learning_rate": 1.175314212052852e-05, | |
| "loss": 0.0949, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.48792270531400966, | |
| "grad_norm": 0.20692448318004608, | |
| "learning_rate": 1.1730583306477602e-05, | |
| "loss": 0.0907, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4927536231884058, | |
| "grad_norm": 0.1886768341064453, | |
| "learning_rate": 1.1708024492426684e-05, | |
| "loss": 0.0858, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4975845410628019, | |
| "grad_norm": 0.19374988973140717, | |
| "learning_rate": 1.1685465678375766e-05, | |
| "loss": 0.084, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5024154589371981, | |
| "grad_norm": 0.1982010304927826, | |
| "learning_rate": 1.1662906864324847e-05, | |
| "loss": 0.087, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5072463768115942, | |
| "grad_norm": 0.292267769575119, | |
| "learning_rate": 1.1640348050273929e-05, | |
| "loss": 0.0918, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5120772946859904, | |
| "grad_norm": 0.19581086933612823, | |
| "learning_rate": 1.1617789236223009e-05, | |
| "loss": 0.1012, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5169082125603864, | |
| "grad_norm": 0.1730077862739563, | |
| "learning_rate": 1.159523042217209e-05, | |
| "loss": 0.0853, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.20485533773899078, | |
| "learning_rate": 1.1572671608121173e-05, | |
| "loss": 0.093, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5265700483091788, | |
| "grad_norm": 0.2086704820394516, | |
| "learning_rate": 1.1550112794070255e-05, | |
| "loss": 0.0945, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 0.15911467373371124, | |
| "learning_rate": 1.1527553980019335e-05, | |
| "loss": 0.1034, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5362318840579711, | |
| "grad_norm": 0.2168796807527542, | |
| "learning_rate": 1.1504995165968417e-05, | |
| "loss": 0.0945, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5410628019323671, | |
| "grad_norm": 0.20228448510169983, | |
| "learning_rate": 1.1482436351917499e-05, | |
| "loss": 0.1029, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5458937198067633, | |
| "grad_norm": 0.2441129982471466, | |
| "learning_rate": 1.1459877537866581e-05, | |
| "loss": 0.0906, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5507246376811594, | |
| "grad_norm": 0.22443729639053345, | |
| "learning_rate": 1.1437318723815663e-05, | |
| "loss": 0.0994, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.18132899701595306, | |
| "learning_rate": 1.1414759909764745e-05, | |
| "loss": 0.0938, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5603864734299517, | |
| "grad_norm": 0.19448505342006683, | |
| "learning_rate": 1.1392201095713824e-05, | |
| "loss": 0.0835, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 0.23075686395168304, | |
| "learning_rate": 1.1369642281662906e-05, | |
| "loss": 0.0983, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5700483091787439, | |
| "grad_norm": 0.22883069515228271, | |
| "learning_rate": 1.1347083467611988e-05, | |
| "loss": 0.0787, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5748792270531401, | |
| "grad_norm": 0.23262719810009003, | |
| "learning_rate": 1.132452465356107e-05, | |
| "loss": 0.0939, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 0.20541128516197205, | |
| "learning_rate": 1.1301965839510152e-05, | |
| "loss": 0.0776, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5845410628019324, | |
| "grad_norm": 0.21663478016853333, | |
| "learning_rate": 1.1279407025459234e-05, | |
| "loss": 0.0918, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.5893719806763285, | |
| "grad_norm": 0.22586220502853394, | |
| "learning_rate": 1.1256848211408314e-05, | |
| "loss": 0.0824, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5942028985507246, | |
| "grad_norm": 0.1860446035861969, | |
| "learning_rate": 1.1234289397357396e-05, | |
| "loss": 0.0853, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5990338164251208, | |
| "grad_norm": 0.195932075381279, | |
| "learning_rate": 1.1211730583306478e-05, | |
| "loss": 0.0818, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6038647342995169, | |
| "grad_norm": 0.19570867717266083, | |
| "learning_rate": 1.118917176925556e-05, | |
| "loss": 0.0859, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 0.16349905729293823, | |
| "learning_rate": 1.116661295520464e-05, | |
| "loss": 0.0938, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6135265700483091, | |
| "grad_norm": 0.1926320493221283, | |
| "learning_rate": 1.1144054141153722e-05, | |
| "loss": 0.0846, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6183574879227053, | |
| "grad_norm": 0.19020161032676697, | |
| "learning_rate": 1.1121495327102803e-05, | |
| "loss": 0.086, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6231884057971014, | |
| "grad_norm": 0.20265896618366241, | |
| "learning_rate": 1.1098936513051885e-05, | |
| "loss": 0.0793, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 0.17398878931999207, | |
| "learning_rate": 1.1076377699000967e-05, | |
| "loss": 0.09, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6328502415458938, | |
| "grad_norm": 0.19005955755710602, | |
| "learning_rate": 1.1053818884950049e-05, | |
| "loss": 0.0792, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6376811594202898, | |
| "grad_norm": 0.18029935657978058, | |
| "learning_rate": 1.103126007089913e-05, | |
| "loss": 0.0923, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.642512077294686, | |
| "grad_norm": 0.1881086826324463, | |
| "learning_rate": 1.1008701256848212e-05, | |
| "loss": 0.0936, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6473429951690821, | |
| "grad_norm": 0.269255667924881, | |
| "learning_rate": 1.0986142442797293e-05, | |
| "loss": 0.0916, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 0.18038909137248993, | |
| "learning_rate": 1.0963583628746373e-05, | |
| "loss": 0.0855, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6570048309178744, | |
| "grad_norm": 0.17990528047084808, | |
| "learning_rate": 1.0941024814695455e-05, | |
| "loss": 0.0926, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6618357487922706, | |
| "grad_norm": 0.2431405931711197, | |
| "learning_rate": 1.0918466000644537e-05, | |
| "loss": 0.0917, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.23243603110313416, | |
| "learning_rate": 1.0895907186593619e-05, | |
| "loss": 0.1046, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6714975845410628, | |
| "grad_norm": 0.20667722821235657, | |
| "learning_rate": 1.0873348372542701e-05, | |
| "loss": 0.0952, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 0.20045587420463562, | |
| "learning_rate": 1.0850789558491781e-05, | |
| "loss": 0.0839, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6811594202898551, | |
| "grad_norm": 0.15829257667064667, | |
| "learning_rate": 1.0828230744440863e-05, | |
| "loss": 0.0931, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6859903381642513, | |
| "grad_norm": 0.18778935074806213, | |
| "learning_rate": 1.0805671930389945e-05, | |
| "loss": 0.083, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6908212560386473, | |
| "grad_norm": 0.1949867457151413, | |
| "learning_rate": 1.0783113116339027e-05, | |
| "loss": 0.0797, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.3177832365036011, | |
| "learning_rate": 1.076055430228811e-05, | |
| "loss": 0.081, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7004830917874396, | |
| "grad_norm": 0.1714804619550705, | |
| "learning_rate": 1.073799548823719e-05, | |
| "loss": 0.0908, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7053140096618358, | |
| "grad_norm": 0.25471800565719604, | |
| "learning_rate": 1.071543667418627e-05, | |
| "loss": 0.0988, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7101449275362319, | |
| "grad_norm": 0.21141599118709564, | |
| "learning_rate": 1.0692877860135352e-05, | |
| "loss": 0.0944, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.714975845410628, | |
| "grad_norm": 0.17371544241905212, | |
| "learning_rate": 1.0670319046084434e-05, | |
| "loss": 0.0836, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7198067632850241, | |
| "grad_norm": 0.19493460655212402, | |
| "learning_rate": 1.0647760232033516e-05, | |
| "loss": 0.0855, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 0.25241127610206604, | |
| "learning_rate": 1.0625201417982598e-05, | |
| "loss": 0.0844, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7294685990338164, | |
| "grad_norm": 0.2515096664428711, | |
| "learning_rate": 1.060264260393168e-05, | |
| "loss": 0.0883, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7342995169082126, | |
| "grad_norm": 0.15292327105998993, | |
| "learning_rate": 1.058008378988076e-05, | |
| "loss": 0.0792, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 0.20495273172855377, | |
| "learning_rate": 1.0557524975829842e-05, | |
| "loss": 0.0789, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7439613526570048, | |
| "grad_norm": 0.261168897151947, | |
| "learning_rate": 1.0534966161778924e-05, | |
| "loss": 0.0832, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.748792270531401, | |
| "grad_norm": 0.28218600153923035, | |
| "learning_rate": 1.0512407347728004e-05, | |
| "loss": 0.1046, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7536231884057971, | |
| "grad_norm": 0.1737246960401535, | |
| "learning_rate": 1.0489848533677086e-05, | |
| "loss": 0.0722, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7584541062801933, | |
| "grad_norm": 0.24183641374111176, | |
| "learning_rate": 1.0467289719626168e-05, | |
| "loss": 0.0752, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.7632850241545893, | |
| "grad_norm": 0.23685990273952484, | |
| "learning_rate": 1.0444730905575249e-05, | |
| "loss": 0.1037, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7681159420289855, | |
| "grad_norm": 0.22956091165542603, | |
| "learning_rate": 1.042217209152433e-05, | |
| "loss": 0.0761, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 0.18922095000743866, | |
| "learning_rate": 1.0399613277473413e-05, | |
| "loss": 0.0885, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.18391458690166473, | |
| "learning_rate": 1.0377054463422495e-05, | |
| "loss": 0.0859, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 0.27890563011169434, | |
| "learning_rate": 1.0354495649371577e-05, | |
| "loss": 0.0924, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7874396135265701, | |
| "grad_norm": 0.22491532564163208, | |
| "learning_rate": 1.0331936835320657e-05, | |
| "loss": 0.0811, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7922705314009661, | |
| "grad_norm": 0.21809989213943481, | |
| "learning_rate": 1.0309378021269739e-05, | |
| "loss": 0.0784, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7971014492753623, | |
| "grad_norm": 0.27180778980255127, | |
| "learning_rate": 1.028681920721882e-05, | |
| "loss": 0.088, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8019323671497585, | |
| "grad_norm": 0.22717216610908508, | |
| "learning_rate": 1.0264260393167901e-05, | |
| "loss": 0.0755, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8067632850241546, | |
| "grad_norm": 0.19013768434524536, | |
| "learning_rate": 1.0241701579116983e-05, | |
| "loss": 0.0782, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8115942028985508, | |
| "grad_norm": 0.2028125375509262, | |
| "learning_rate": 1.0219142765066065e-05, | |
| "loss": 0.1034, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8164251207729468, | |
| "grad_norm": 0.24243703484535217, | |
| "learning_rate": 1.0196583951015146e-05, | |
| "loss": 0.0899, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 0.21742011606693268, | |
| "learning_rate": 1.0174025136964228e-05, | |
| "loss": 0.0898, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8260869565217391, | |
| "grad_norm": 0.2000913769006729, | |
| "learning_rate": 1.015146632291331e-05, | |
| "loss": 0.0828, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8309178743961353, | |
| "grad_norm": 0.1902933269739151, | |
| "learning_rate": 1.0128907508862392e-05, | |
| "loss": 0.0934, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8357487922705314, | |
| "grad_norm": 0.20363092422485352, | |
| "learning_rate": 1.0106348694811474e-05, | |
| "loss": 0.081, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.8405797101449275, | |
| "grad_norm": 0.2238474041223526, | |
| "learning_rate": 1.0083789880760556e-05, | |
| "loss": 0.0963, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8454106280193237, | |
| "grad_norm": 0.19188345968723297, | |
| "learning_rate": 1.0061231066709634e-05, | |
| "loss": 0.0809, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.8502415458937198, | |
| "grad_norm": 0.18286921083927155, | |
| "learning_rate": 1.0038672252658716e-05, | |
| "loss": 0.0891, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.855072463768116, | |
| "grad_norm": 0.19798459112644196, | |
| "learning_rate": 1.0016113438607798e-05, | |
| "loss": 0.0789, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8599033816425121, | |
| "grad_norm": 0.1937275230884552, | |
| "learning_rate": 9.99355462455688e-06, | |
| "loss": 0.0748, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8647342995169082, | |
| "grad_norm": 0.2399519830942154, | |
| "learning_rate": 9.970995810505962e-06, | |
| "loss": 0.0941, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.2435486763715744, | |
| "learning_rate": 9.948436996455044e-06, | |
| "loss": 0.078, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8743961352657005, | |
| "grad_norm": 0.22818566858768463, | |
| "learning_rate": 9.925878182404124e-06, | |
| "loss": 0.0813, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.8792270531400966, | |
| "grad_norm": 0.19992083311080933, | |
| "learning_rate": 9.903319368353206e-06, | |
| "loss": 0.0757, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8840579710144928, | |
| "grad_norm": 0.24121499061584473, | |
| "learning_rate": 9.880760554302288e-06, | |
| "loss": 0.0878, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.2414121776819229, | |
| "learning_rate": 9.85820174025137e-06, | |
| "loss": 0.076, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.893719806763285, | |
| "grad_norm": 0.17777179181575775, | |
| "learning_rate": 9.83564292620045e-06, | |
| "loss": 0.0903, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8985507246376812, | |
| "grad_norm": 0.23024319112300873, | |
| "learning_rate": 9.813084112149533e-06, | |
| "loss": 0.0965, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9033816425120773, | |
| "grad_norm": 0.20664696395397186, | |
| "learning_rate": 9.790525298098613e-06, | |
| "loss": 0.0707, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9082125603864735, | |
| "grad_norm": 0.1725015491247177, | |
| "learning_rate": 9.767966484047695e-06, | |
| "loss": 0.0821, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9130434782608695, | |
| "grad_norm": 0.2138936072587967, | |
| "learning_rate": 9.745407669996777e-06, | |
| "loss": 0.087, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 0.24879959225654602, | |
| "learning_rate": 9.722848855945859e-06, | |
| "loss": 0.0782, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9227053140096618, | |
| "grad_norm": 0.24507424235343933, | |
| "learning_rate": 9.700290041894941e-06, | |
| "loss": 0.0976, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.927536231884058, | |
| "grad_norm": 0.21825656294822693, | |
| "learning_rate": 9.677731227844023e-06, | |
| "loss": 0.0846, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9323671497584541, | |
| "grad_norm": 0.22634956240653992, | |
| "learning_rate": 9.655172413793103e-06, | |
| "loss": 0.0833, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.9371980676328503, | |
| "grad_norm": 0.20103132724761963, | |
| "learning_rate": 9.632613599742184e-06, | |
| "loss": 0.0739, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9420289855072463, | |
| "grad_norm": 0.19459068775177002, | |
| "learning_rate": 9.610054785691266e-06, | |
| "loss": 0.0841, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.9468599033816425, | |
| "grad_norm": 0.18598385155200958, | |
| "learning_rate": 9.587495971640348e-06, | |
| "loss": 0.0734, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9516908212560387, | |
| "grad_norm": 0.24302643537521362, | |
| "learning_rate": 9.56493715758943e-06, | |
| "loss": 0.0893, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 0.23758938908576965, | |
| "learning_rate": 9.542378343538512e-06, | |
| "loss": 0.0763, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.961352657004831, | |
| "grad_norm": 0.2180752158164978, | |
| "learning_rate": 9.519819529487592e-06, | |
| "loss": 0.0792, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 0.22509507834911346, | |
| "learning_rate": 9.497260715436674e-06, | |
| "loss": 0.0682, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9710144927536232, | |
| "grad_norm": 0.197494238615036, | |
| "learning_rate": 9.474701901385756e-06, | |
| "loss": 0.0903, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.9758454106280193, | |
| "grad_norm": 0.1817607879638672, | |
| "learning_rate": 9.452143087334838e-06, | |
| "loss": 0.0846, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9806763285024155, | |
| "grad_norm": 0.19075438380241394, | |
| "learning_rate": 9.42958427328392e-06, | |
| "loss": 0.0846, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.9855072463768116, | |
| "grad_norm": 0.15087321400642395, | |
| "learning_rate": 9.407025459233e-06, | |
| "loss": 0.0753, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9903381642512077, | |
| "grad_norm": 0.2226846069097519, | |
| "learning_rate": 9.38446664518208e-06, | |
| "loss": 0.0666, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.9951690821256038, | |
| "grad_norm": 0.30765634775161743, | |
| "learning_rate": 9.361907831131162e-06, | |
| "loss": 0.0779, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4031631052494049, | |
| "learning_rate": 9.339349017080244e-06, | |
| "loss": 0.095, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_runtime": 338.4602, | |
| "eval_samples_per_second": 3.055, | |
| "eval_steps_per_second": 0.384, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.0048309178743962, | |
| "grad_norm": 0.2173592448234558, | |
| "learning_rate": 9.316790203029326e-06, | |
| "loss": 0.0803, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.0096618357487923, | |
| "grad_norm": 0.22241808474063873, | |
| "learning_rate": 9.294231388978408e-06, | |
| "loss": 0.0909, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.0144927536231885, | |
| "grad_norm": 0.2699296474456787, | |
| "learning_rate": 9.27167257492749e-06, | |
| "loss": 0.0842, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.0193236714975846, | |
| "grad_norm": 0.27080684900283813, | |
| "learning_rate": 9.24911376087657e-06, | |
| "loss": 0.0837, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.0241545893719808, | |
| "grad_norm": 0.1808546930551529, | |
| "learning_rate": 9.226554946825653e-06, | |
| "loss": 0.0816, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.0289855072463767, | |
| "grad_norm": 0.19763918220996857, | |
| "learning_rate": 9.203996132774735e-06, | |
| "loss": 0.0705, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.0338164251207729, | |
| "grad_norm": 0.21294108033180237, | |
| "learning_rate": 9.181437318723815e-06, | |
| "loss": 0.0726, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.038647342995169, | |
| "grad_norm": 0.19769993424415588, | |
| "learning_rate": 9.158878504672897e-06, | |
| "loss": 0.0743, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.23708152770996094, | |
| "learning_rate": 9.136319690621979e-06, | |
| "loss": 0.0785, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.0483091787439613, | |
| "grad_norm": 0.232899010181427, | |
| "learning_rate": 9.11376087657106e-06, | |
| "loss": 0.093, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.0531400966183575, | |
| "grad_norm": 0.267478883266449, | |
| "learning_rate": 9.091202062520141e-06, | |
| "loss": 0.0901, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.0579710144927537, | |
| "grad_norm": 0.23761190474033356, | |
| "learning_rate": 9.068643248469223e-06, | |
| "loss": 0.0898, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.0628019323671498, | |
| "grad_norm": 0.19679813086986542, | |
| "learning_rate": 9.046084434418305e-06, | |
| "loss": 0.0877, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.067632850241546, | |
| "grad_norm": 0.20915761590003967, | |
| "learning_rate": 9.023525620367387e-06, | |
| "loss": 0.0731, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.0724637681159421, | |
| "grad_norm": 0.18718890845775604, | |
| "learning_rate": 9.000966806316468e-06, | |
| "loss": 0.0885, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.077294685990338, | |
| "grad_norm": 0.29885435104370117, | |
| "learning_rate": 8.97840799226555e-06, | |
| "loss": 0.0861, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.0821256038647342, | |
| "grad_norm": 0.16953594982624054, | |
| "learning_rate": 8.95584917821463e-06, | |
| "loss": 0.0862, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 0.21629682183265686, | |
| "learning_rate": 8.933290364163712e-06, | |
| "loss": 0.1022, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.0917874396135265, | |
| "grad_norm": 0.26614615321159363, | |
| "learning_rate": 8.910731550112794e-06, | |
| "loss": 0.0825, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.0966183574879227, | |
| "grad_norm": 0.2642468810081482, | |
| "learning_rate": 8.888172736061876e-06, | |
| "loss": 0.0796, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.1014492753623188, | |
| "grad_norm": 0.16877882182598114, | |
| "learning_rate": 8.865613922010956e-06, | |
| "loss": 0.0726, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.106280193236715, | |
| "grad_norm": 0.2619246542453766, | |
| "learning_rate": 8.843055107960038e-06, | |
| "loss": 0.0835, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.2424723505973816, | |
| "learning_rate": 8.82049629390912e-06, | |
| "loss": 0.0896, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.1159420289855073, | |
| "grad_norm": 0.20973582565784454, | |
| "learning_rate": 8.797937479858202e-06, | |
| "loss": 0.0751, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.1207729468599035, | |
| "grad_norm": 0.23418009281158447, | |
| "learning_rate": 8.775378665807284e-06, | |
| "loss": 0.077, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.1256038647342996, | |
| "grad_norm": 0.3117668032646179, | |
| "learning_rate": 8.752819851756366e-06, | |
| "loss": 0.0769, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 0.25092753767967224, | |
| "learning_rate": 8.730261037705446e-06, | |
| "loss": 0.0729, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.1352657004830917, | |
| "grad_norm": 0.1926090270280838, | |
| "learning_rate": 8.707702223654527e-06, | |
| "loss": 0.0772, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.1400966183574879, | |
| "grad_norm": 0.27212995290756226, | |
| "learning_rate": 8.685143409603609e-06, | |
| "loss": 0.0712, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.144927536231884, | |
| "grad_norm": 0.2097581923007965, | |
| "learning_rate": 8.66258459555269e-06, | |
| "loss": 0.0767, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.1497584541062802, | |
| "grad_norm": 0.2765638828277588, | |
| "learning_rate": 8.640025781501773e-06, | |
| "loss": 0.0885, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.1545893719806763, | |
| "grad_norm": 0.28414320945739746, | |
| "learning_rate": 8.617466967450855e-06, | |
| "loss": 0.0631, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.1594202898550725, | |
| "grad_norm": 0.21230548620224, | |
| "learning_rate": 8.594908153399935e-06, | |
| "loss": 0.0848, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.1642512077294687, | |
| "grad_norm": 0.1870320439338684, | |
| "learning_rate": 8.572349339349017e-06, | |
| "loss": 0.0724, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.1690821256038648, | |
| "grad_norm": 0.23322801291942596, | |
| "learning_rate": 8.549790525298099e-06, | |
| "loss": 0.0846, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 0.22248071432113647, | |
| "learning_rate": 8.52723171124718e-06, | |
| "loss": 0.0624, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.178743961352657, | |
| "grad_norm": 0.196117103099823, | |
| "learning_rate": 8.504672897196261e-06, | |
| "loss": 0.093, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.183574879227053, | |
| "grad_norm": 0.2212802767753601, | |
| "learning_rate": 8.482114083145343e-06, | |
| "loss": 0.0853, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.1884057971014492, | |
| "grad_norm": 0.17421841621398926, | |
| "learning_rate": 8.459555269094424e-06, | |
| "loss": 0.0747, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.1932367149758454, | |
| "grad_norm": 0.2645537853240967, | |
| "learning_rate": 8.436996455043506e-06, | |
| "loss": 0.0859, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.1980676328502415, | |
| "grad_norm": 0.27182498574256897, | |
| "learning_rate": 8.414437640992588e-06, | |
| "loss": 0.0943, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.2028985507246377, | |
| "grad_norm": 0.20389291644096375, | |
| "learning_rate": 8.39187882694167e-06, | |
| "loss": 0.0783, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.2077294685990339, | |
| "grad_norm": 0.3193868398666382, | |
| "learning_rate": 8.369320012890752e-06, | |
| "loss": 0.0917, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.21256038647343, | |
| "grad_norm": 0.2852030098438263, | |
| "learning_rate": 8.346761198839834e-06, | |
| "loss": 0.077, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 0.256452739238739, | |
| "learning_rate": 8.324202384788914e-06, | |
| "loss": 0.0804, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.209047332406044, | |
| "learning_rate": 8.301643570737994e-06, | |
| "loss": 0.0928, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.2270531400966185, | |
| "grad_norm": 0.21215900778770447, | |
| "learning_rate": 8.279084756687076e-06, | |
| "loss": 0.0788, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.2318840579710144, | |
| "grad_norm": 0.15550634264945984, | |
| "learning_rate": 8.256525942636158e-06, | |
| "loss": 0.0804, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.2367149758454106, | |
| "grad_norm": 0.16960662603378296, | |
| "learning_rate": 8.23396712858524e-06, | |
| "loss": 0.0737, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.2415458937198067, | |
| "grad_norm": 0.20484741032123566, | |
| "learning_rate": 8.211408314534322e-06, | |
| "loss": 0.0794, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.2463768115942029, | |
| "grad_norm": 0.24889996647834778, | |
| "learning_rate": 8.188849500483402e-06, | |
| "loss": 0.0903, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.251207729468599, | |
| "grad_norm": 0.23695576190948486, | |
| "learning_rate": 8.166290686432484e-06, | |
| "loss": 0.0826, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 1.2560386473429952, | |
| "grad_norm": 0.23449349403381348, | |
| "learning_rate": 8.143731872381566e-06, | |
| "loss": 0.0922, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 0.2362452745437622, | |
| "learning_rate": 8.121173058330648e-06, | |
| "loss": 0.0716, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 1.2657004830917875, | |
| "grad_norm": 0.33280622959136963, | |
| "learning_rate": 8.09861424427973e-06, | |
| "loss": 0.0909, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.2705314009661834, | |
| "grad_norm": 0.22267523407936096, | |
| "learning_rate": 8.07605543022881e-06, | |
| "loss": 0.0816, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 1.2753623188405796, | |
| "grad_norm": 0.23176385462284088, | |
| "learning_rate": 8.053496616177891e-06, | |
| "loss": 0.091, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.2801932367149758, | |
| "grad_norm": 0.21951176226139069, | |
| "learning_rate": 8.030937802126973e-06, | |
| "loss": 0.0752, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.285024154589372, | |
| "grad_norm": 0.19361701607704163, | |
| "learning_rate": 8.008378988076055e-06, | |
| "loss": 0.0731, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.289855072463768, | |
| "grad_norm": 0.2284880429506302, | |
| "learning_rate": 7.985820174025137e-06, | |
| "loss": 0.0821, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 1.2946859903381642, | |
| "grad_norm": 0.28775539994239807, | |
| "learning_rate": 7.963261359974219e-06, | |
| "loss": 0.0865, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.2995169082125604, | |
| "grad_norm": 0.22133222222328186, | |
| "learning_rate": 7.940702545923301e-06, | |
| "loss": 0.0722, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 0.2120644450187683, | |
| "learning_rate": 7.918143731872381e-06, | |
| "loss": 0.0642, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.3091787439613527, | |
| "grad_norm": 0.2922479212284088, | |
| "learning_rate": 7.895584917821463e-06, | |
| "loss": 0.0804, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 1.3140096618357489, | |
| "grad_norm": 0.2302795797586441, | |
| "learning_rate": 7.873026103770545e-06, | |
| "loss": 0.0726, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.318840579710145, | |
| "grad_norm": 0.28763264417648315, | |
| "learning_rate": 7.850467289719626e-06, | |
| "loss": 0.0976, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 1.3236714975845412, | |
| "grad_norm": 0.2106347233057022, | |
| "learning_rate": 7.827908475668708e-06, | |
| "loss": 0.0744, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.3285024154589373, | |
| "grad_norm": 0.23215855658054352, | |
| "learning_rate": 7.80534966161779e-06, | |
| "loss": 0.0916, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.20885543525218964, | |
| "learning_rate": 7.78279084756687e-06, | |
| "loss": 0.0774, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.3381642512077294, | |
| "grad_norm": 0.20533576607704163, | |
| "learning_rate": 7.760232033515952e-06, | |
| "loss": 0.0751, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 1.3429951690821256, | |
| "grad_norm": 0.20719490945339203, | |
| "learning_rate": 7.737673219465034e-06, | |
| "loss": 0.0816, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.3478260869565217, | |
| "grad_norm": 0.19761165976524353, | |
| "learning_rate": 7.715114405414116e-06, | |
| "loss": 0.0804, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 1.3526570048309179, | |
| "grad_norm": 0.20369771122932434, | |
| "learning_rate": 7.692555591363198e-06, | |
| "loss": 0.0845, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.357487922705314, | |
| "grad_norm": 0.20887012779712677, | |
| "learning_rate": 7.669996777312278e-06, | |
| "loss": 0.0704, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 1.3623188405797102, | |
| "grad_norm": 0.29784587025642395, | |
| "learning_rate": 7.64743796326136e-06, | |
| "loss": 0.0866, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.3671497584541064, | |
| "grad_norm": 0.31036221981048584, | |
| "learning_rate": 7.62487914921044e-06, | |
| "loss": 0.0862, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 1.3719806763285023, | |
| "grad_norm": 0.25198647379875183, | |
| "learning_rate": 7.602320335159522e-06, | |
| "loss": 0.0825, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.3768115942028984, | |
| "grad_norm": 0.24515630304813385, | |
| "learning_rate": 7.579761521108604e-06, | |
| "loss": 0.0787, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.3816425120772946, | |
| "grad_norm": 0.22536733746528625, | |
| "learning_rate": 7.5572027070576855e-06, | |
| "loss": 0.0928, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.3864734299516908, | |
| "grad_norm": 0.23405781388282776, | |
| "learning_rate": 7.5346438930067675e-06, | |
| "loss": 0.0917, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.24243396520614624, | |
| "learning_rate": 7.5120850789558495e-06, | |
| "loss": 0.0675, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.396135265700483, | |
| "grad_norm": 0.2637854814529419, | |
| "learning_rate": 7.489526264904931e-06, | |
| "loss": 0.0863, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 1.4009661835748792, | |
| "grad_norm": 0.2491244375705719, | |
| "learning_rate": 7.466967450854013e-06, | |
| "loss": 0.0808, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.4057971014492754, | |
| "grad_norm": 0.23132705688476562, | |
| "learning_rate": 7.444408636803095e-06, | |
| "loss": 0.0797, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 1.4106280193236715, | |
| "grad_norm": 0.2987098693847656, | |
| "learning_rate": 7.421849822752176e-06, | |
| "loss": 0.0766, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.4154589371980677, | |
| "grad_norm": 0.23995457589626312, | |
| "learning_rate": 7.399291008701256e-06, | |
| "loss": 0.0764, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 1.4202898550724639, | |
| "grad_norm": 0.21818973124027252, | |
| "learning_rate": 7.376732194650338e-06, | |
| "loss": 0.09, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.42512077294686, | |
| "grad_norm": 0.19304029643535614, | |
| "learning_rate": 7.354173380599419e-06, | |
| "loss": 0.0759, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.4299516908212562, | |
| "grad_norm": 0.26081785559654236, | |
| "learning_rate": 7.331614566548501e-06, | |
| "loss": 0.0781, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.434782608695652, | |
| "grad_norm": 0.23940761387348175, | |
| "learning_rate": 7.309055752497583e-06, | |
| "loss": 0.085, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 1.4396135265700483, | |
| "grad_norm": 0.21909761428833008, | |
| "learning_rate": 7.286496938446664e-06, | |
| "loss": 0.0815, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.16527162492275238, | |
| "learning_rate": 7.263938124395746e-06, | |
| "loss": 0.0698, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 1.4492753623188406, | |
| "grad_norm": 0.21258555352687836, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.0806, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4541062801932367, | |
| "grad_norm": 0.18572719395160675, | |
| "learning_rate": 7.2188204962939095e-06, | |
| "loss": 0.0757, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 1.458937198067633, | |
| "grad_norm": 0.16916704177856445, | |
| "learning_rate": 7.19626168224299e-06, | |
| "loss": 0.07, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.463768115942029, | |
| "grad_norm": 0.289044588804245, | |
| "learning_rate": 7.173702868192072e-06, | |
| "loss": 0.0656, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 1.4685990338164252, | |
| "grad_norm": 0.27173757553100586, | |
| "learning_rate": 7.151144054141153e-06, | |
| "loss": 0.0704, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.4734299516908211, | |
| "grad_norm": 0.2929324209690094, | |
| "learning_rate": 7.128585240090235e-06, | |
| "loss": 0.0833, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 0.2387627214193344, | |
| "learning_rate": 7.106026426039317e-06, | |
| "loss": 0.075, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.4830917874396135, | |
| "grad_norm": 0.3277483582496643, | |
| "learning_rate": 7.083467611988398e-06, | |
| "loss": 0.074, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 1.4879227053140096, | |
| "grad_norm": 0.23673392832279205, | |
| "learning_rate": 7.06090879793748e-06, | |
| "loss": 0.0697, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.4927536231884058, | |
| "grad_norm": 0.19109922647476196, | |
| "learning_rate": 7.038349983886562e-06, | |
| "loss": 0.0775, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.497584541062802, | |
| "grad_norm": 0.2344091832637787, | |
| "learning_rate": 7.015791169835643e-06, | |
| "loss": 0.0644, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.502415458937198, | |
| "grad_norm": 0.28420698642730713, | |
| "learning_rate": 6.993232355784724e-06, | |
| "loss": 0.0935, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 1.5072463768115942, | |
| "grad_norm": 0.2632888853549957, | |
| "learning_rate": 6.970673541733806e-06, | |
| "loss": 0.083, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.5120772946859904, | |
| "grad_norm": 0.2461112141609192, | |
| "learning_rate": 6.9481147276828875e-06, | |
| "loss": 0.0729, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 1.5169082125603865, | |
| "grad_norm": 0.2015853226184845, | |
| "learning_rate": 6.9255559136319695e-06, | |
| "loss": 0.0836, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 0.2409069985151291, | |
| "learning_rate": 6.902997099581051e-06, | |
| "loss": 0.0797, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.5265700483091789, | |
| "grad_norm": 0.2014143019914627, | |
| "learning_rate": 6.880438285530132e-06, | |
| "loss": 0.09, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.531400966183575, | |
| "grad_norm": 0.2173725664615631, | |
| "learning_rate": 6.857879471479214e-06, | |
| "loss": 0.0648, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 1.5362318840579712, | |
| "grad_norm": 0.20185904204845428, | |
| "learning_rate": 6.835320657428296e-06, | |
| "loss": 0.0924, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.541062801932367, | |
| "grad_norm": 0.29456228017807007, | |
| "learning_rate": 6.812761843377377e-06, | |
| "loss": 0.0764, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.5458937198067633, | |
| "grad_norm": 0.22320301830768585, | |
| "learning_rate": 6.790203029326458e-06, | |
| "loss": 0.0754, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.5507246376811594, | |
| "grad_norm": 0.2032977044582367, | |
| "learning_rate": 6.76764421527554e-06, | |
| "loss": 0.089, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.24341309070587158, | |
| "learning_rate": 6.745085401224621e-06, | |
| "loss": 0.0767, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.5603864734299517, | |
| "grad_norm": 0.22675780951976776, | |
| "learning_rate": 6.722526587173703e-06, | |
| "loss": 0.0811, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 0.2980429232120514, | |
| "learning_rate": 6.699967773122784e-06, | |
| "loss": 0.0714, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.5700483091787438, | |
| "grad_norm": 0.2221527248620987, | |
| "learning_rate": 6.6774089590718655e-06, | |
| "loss": 0.0811, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.57487922705314, | |
| "grad_norm": 0.29102587699890137, | |
| "learning_rate": 6.6548501450209474e-06, | |
| "loss": 0.0717, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.5797101449275361, | |
| "grad_norm": 0.24565882980823517, | |
| "learning_rate": 6.632291330970029e-06, | |
| "loss": 0.0688, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.5845410628019323, | |
| "grad_norm": 0.2056146264076233, | |
| "learning_rate": 6.609732516919111e-06, | |
| "loss": 0.0739, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.5893719806763285, | |
| "grad_norm": 0.25777336955070496, | |
| "learning_rate": 6.587173702868192e-06, | |
| "loss": 0.0746, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 1.5942028985507246, | |
| "grad_norm": 0.20640453696250916, | |
| "learning_rate": 6.564614888817273e-06, | |
| "loss": 0.0757, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.5990338164251208, | |
| "grad_norm": 0.16480913758277893, | |
| "learning_rate": 6.542056074766355e-06, | |
| "loss": 0.0752, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 1.603864734299517, | |
| "grad_norm": 0.23693595826625824, | |
| "learning_rate": 6.519497260715437e-06, | |
| "loss": 0.0813, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.608695652173913, | |
| "grad_norm": 0.24152866005897522, | |
| "learning_rate": 6.496938446664518e-06, | |
| "loss": 0.0784, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.6135265700483092, | |
| "grad_norm": 0.23890602588653564, | |
| "learning_rate": 6.474379632613599e-06, | |
| "loss": 0.0813, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.6183574879227054, | |
| "grad_norm": 0.2686842679977417, | |
| "learning_rate": 6.451820818562681e-06, | |
| "loss": 0.0833, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.6231884057971016, | |
| "grad_norm": 0.2103358954191208, | |
| "learning_rate": 6.429262004511762e-06, | |
| "loss": 0.0906, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.6280193236714977, | |
| "grad_norm": 0.23938271403312683, | |
| "learning_rate": 6.406703190460844e-06, | |
| "loss": 0.0721, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 1.6328502415458939, | |
| "grad_norm": 0.1797400861978531, | |
| "learning_rate": 6.384144376409926e-06, | |
| "loss": 0.0678, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.6376811594202898, | |
| "grad_norm": 0.23905880749225616, | |
| "learning_rate": 6.361585562359007e-06, | |
| "loss": 0.0886, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.642512077294686, | |
| "grad_norm": 0.19138076901435852, | |
| "learning_rate": 6.339026748308089e-06, | |
| "loss": 0.0705, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.6473429951690821, | |
| "grad_norm": 0.19759757816791534, | |
| "learning_rate": 6.3164679342571706e-06, | |
| "loss": 0.0772, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 0.22951267659664154, | |
| "learning_rate": 6.293909120206252e-06, | |
| "loss": 0.0701, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.6570048309178744, | |
| "grad_norm": 0.3317079246044159, | |
| "learning_rate": 6.271350306155334e-06, | |
| "loss": 0.0838, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 1.6618357487922706, | |
| "grad_norm": 0.2875089645385742, | |
| "learning_rate": 6.248791492104415e-06, | |
| "loss": 0.0711, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.22365209460258484, | |
| "learning_rate": 6.226232678053496e-06, | |
| "loss": 0.0913, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.6714975845410627, | |
| "grad_norm": 0.26004156470298767, | |
| "learning_rate": 6.203673864002578e-06, | |
| "loss": 0.0749, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.6763285024154588, | |
| "grad_norm": 0.24029529094696045, | |
| "learning_rate": 6.18111504995166e-06, | |
| "loss": 0.0872, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 1.681159420289855, | |
| "grad_norm": 0.2503759562969208, | |
| "learning_rate": 6.158556235900741e-06, | |
| "loss": 0.0662, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.6859903381642511, | |
| "grad_norm": 0.24961721897125244, | |
| "learning_rate": 6.135997421849822e-06, | |
| "loss": 0.0805, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 1.6908212560386473, | |
| "grad_norm": 0.20291025936603546, | |
| "learning_rate": 6.113438607798904e-06, | |
| "loss": 0.0723, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.6956521739130435, | |
| "grad_norm": 0.24923092126846313, | |
| "learning_rate": 6.0908797937479854e-06, | |
| "loss": 0.0766, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.7004830917874396, | |
| "grad_norm": 0.3006664514541626, | |
| "learning_rate": 6.068320979697067e-06, | |
| "loss": 0.0767, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.7053140096618358, | |
| "grad_norm": 0.22034914791584015, | |
| "learning_rate": 6.045762165646149e-06, | |
| "loss": 0.0716, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 1.710144927536232, | |
| "grad_norm": 0.22951188683509827, | |
| "learning_rate": 6.02320335159523e-06, | |
| "loss": 0.0713, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.714975845410628, | |
| "grad_norm": 0.22270874679088593, | |
| "learning_rate": 6.000644537544312e-06, | |
| "loss": 0.0671, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.7198067632850242, | |
| "grad_norm": 0.23195502161979675, | |
| "learning_rate": 5.978085723493394e-06, | |
| "loss": 0.0864, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.7246376811594204, | |
| "grad_norm": 0.2421010136604309, | |
| "learning_rate": 5.955526909442475e-06, | |
| "loss": 0.0886, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.7294685990338166, | |
| "grad_norm": 0.20693883299827576, | |
| "learning_rate": 5.932968095391557e-06, | |
| "loss": 0.0715, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.7342995169082127, | |
| "grad_norm": 0.32137101888656616, | |
| "learning_rate": 5.910409281340638e-06, | |
| "loss": 0.0639, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.21108365058898926, | |
| "learning_rate": 5.887850467289719e-06, | |
| "loss": 0.0786, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.7439613526570048, | |
| "grad_norm": 0.2952270805835724, | |
| "learning_rate": 5.865291653238801e-06, | |
| "loss": 0.0641, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 1.748792270531401, | |
| "grad_norm": 0.26709944009780884, | |
| "learning_rate": 5.842732839187883e-06, | |
| "loss": 0.0698, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.7536231884057971, | |
| "grad_norm": 0.30126988887786865, | |
| "learning_rate": 5.820174025136964e-06, | |
| "loss": 0.0773, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.7584541062801933, | |
| "grad_norm": 0.2402152717113495, | |
| "learning_rate": 5.797615211086045e-06, | |
| "loss": 0.0778, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.7632850241545892, | |
| "grad_norm": 0.19652244448661804, | |
| "learning_rate": 5.775056397035127e-06, | |
| "loss": 0.082, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.7681159420289854, | |
| "grad_norm": 0.21389204263687134, | |
| "learning_rate": 5.7524975829842086e-06, | |
| "loss": 0.0727, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.7729468599033815, | |
| "grad_norm": 0.2189796268939972, | |
| "learning_rate": 5.7299387689332905e-06, | |
| "loss": 0.0757, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.28000935912132263, | |
| "learning_rate": 5.7073799548823725e-06, | |
| "loss": 0.0803, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 0.24566881358623505, | |
| "learning_rate": 5.684821140831453e-06, | |
| "loss": 0.0815, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.78743961352657, | |
| "grad_norm": 0.22037634253501892, | |
| "learning_rate": 5.662262326780535e-06, | |
| "loss": 0.0871, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.7922705314009661, | |
| "grad_norm": 0.1990278661251068, | |
| "learning_rate": 5.639703512729617e-06, | |
| "loss": 0.0756, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.7971014492753623, | |
| "grad_norm": 0.3180176615715027, | |
| "learning_rate": 5.617144698678698e-06, | |
| "loss": 0.0735, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.8019323671497585, | |
| "grad_norm": 0.2075718492269516, | |
| "learning_rate": 5.59458588462778e-06, | |
| "loss": 0.0665, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 1.8067632850241546, | |
| "grad_norm": 0.2611768841743469, | |
| "learning_rate": 5.572027070576861e-06, | |
| "loss": 0.0873, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.8115942028985508, | |
| "grad_norm": 0.22146160900592804, | |
| "learning_rate": 5.549468256525942e-06, | |
| "loss": 0.0638, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.816425120772947, | |
| "grad_norm": 0.29287296533584595, | |
| "learning_rate": 5.526909442475024e-06, | |
| "loss": 0.0812, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.821256038647343, | |
| "grad_norm": 0.2280767410993576, | |
| "learning_rate": 5.504350628424106e-06, | |
| "loss": 0.0766, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 0.20453138649463654, | |
| "learning_rate": 5.4817918143731865e-06, | |
| "loss": 0.0748, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.8309178743961354, | |
| "grad_norm": 0.2855188250541687, | |
| "learning_rate": 5.4592330003222685e-06, | |
| "loss": 0.0901, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 1.8357487922705316, | |
| "grad_norm": 0.21556098759174347, | |
| "learning_rate": 5.4366741862713505e-06, | |
| "loss": 0.0735, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.8405797101449275, | |
| "grad_norm": 0.3091937303543091, | |
| "learning_rate": 5.414115372220432e-06, | |
| "loss": 0.0636, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.8454106280193237, | |
| "grad_norm": 0.2939262390136719, | |
| "learning_rate": 5.391556558169514e-06, | |
| "loss": 0.0753, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.8502415458937198, | |
| "grad_norm": 0.2101174294948578, | |
| "learning_rate": 5.368997744118595e-06, | |
| "loss": 0.0714, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 1.855072463768116, | |
| "grad_norm": 0.2570497691631317, | |
| "learning_rate": 5.346438930067676e-06, | |
| "loss": 0.0877, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.8599033816425121, | |
| "grad_norm": 0.2754373848438263, | |
| "learning_rate": 5.323880116016758e-06, | |
| "loss": 0.0729, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.864734299516908, | |
| "grad_norm": 0.2952544391155243, | |
| "learning_rate": 5.30132130196584e-06, | |
| "loss": 0.0714, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.8695652173913042, | |
| "grad_norm": 0.2360425889492035, | |
| "learning_rate": 5.278762487914921e-06, | |
| "loss": 0.0711, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.8743961352657004, | |
| "grad_norm": 0.22847935557365417, | |
| "learning_rate": 5.256203673864002e-06, | |
| "loss": 0.07, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.8792270531400965, | |
| "grad_norm": 0.26060476899147034, | |
| "learning_rate": 5.233644859813084e-06, | |
| "loss": 0.086, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 1.8840579710144927, | |
| "grad_norm": 0.28593048453330994, | |
| "learning_rate": 5.211086045762165e-06, | |
| "loss": 0.0782, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.2553214430809021, | |
| "learning_rate": 5.188527231711247e-06, | |
| "loss": 0.0689, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 1.893719806763285, | |
| "grad_norm": 0.38168102502822876, | |
| "learning_rate": 5.1659684176603285e-06, | |
| "loss": 0.0917, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.8985507246376812, | |
| "grad_norm": 0.22879190742969513, | |
| "learning_rate": 5.14340960360941e-06, | |
| "loss": 0.0833, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.9033816425120773, | |
| "grad_norm": 0.19676880538463593, | |
| "learning_rate": 5.120850789558492e-06, | |
| "loss": 0.0594, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.9082125603864735, | |
| "grad_norm": 0.36660292744636536, | |
| "learning_rate": 5.098291975507573e-06, | |
| "loss": 0.0932, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 0.23486468195915222, | |
| "learning_rate": 5.075733161456655e-06, | |
| "loss": 0.0941, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.9178743961352658, | |
| "grad_norm": 0.2950279414653778, | |
| "learning_rate": 5.053174347405737e-06, | |
| "loss": 0.0796, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 1.922705314009662, | |
| "grad_norm": 0.1995108425617218, | |
| "learning_rate": 5.030615533354817e-06, | |
| "loss": 0.0766, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.927536231884058, | |
| "grad_norm": 0.3509507179260254, | |
| "learning_rate": 5.008056719303899e-06, | |
| "loss": 0.0718, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.9323671497584543, | |
| "grad_norm": 0.22868584096431732, | |
| "learning_rate": 4.985497905252981e-06, | |
| "loss": 0.0724, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.9371980676328504, | |
| "grad_norm": 0.270059734582901, | |
| "learning_rate": 4.962939091202062e-06, | |
| "loss": 0.0761, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 1.9420289855072463, | |
| "grad_norm": 0.24437829852104187, | |
| "learning_rate": 4.940380277151144e-06, | |
| "loss": 0.0729, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.9468599033816425, | |
| "grad_norm": 0.24446424841880798, | |
| "learning_rate": 4.917821463100225e-06, | |
| "loss": 0.0648, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 1.9516908212560387, | |
| "grad_norm": 0.21626543998718262, | |
| "learning_rate": 4.8952626490493065e-06, | |
| "loss": 0.0739, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.9565217391304348, | |
| "grad_norm": 0.20689117908477783, | |
| "learning_rate": 4.8727038349983885e-06, | |
| "loss": 0.0701, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.961352657004831, | |
| "grad_norm": 0.2660706043243408, | |
| "learning_rate": 4.8501450209474705e-06, | |
| "loss": 0.0571, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.966183574879227, | |
| "grad_norm": 0.24084658920764923, | |
| "learning_rate": 4.827586206896552e-06, | |
| "loss": 0.0764, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 1.971014492753623, | |
| "grad_norm": 0.2771299481391907, | |
| "learning_rate": 4.805027392845633e-06, | |
| "loss": 0.0738, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.9758454106280192, | |
| "grad_norm": 0.2248222976922989, | |
| "learning_rate": 4.782468578794715e-06, | |
| "loss": 0.0774, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 1.9806763285024154, | |
| "grad_norm": 0.22526535391807556, | |
| "learning_rate": 4.759909764743796e-06, | |
| "loss": 0.0678, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.9855072463768115, | |
| "grad_norm": 0.21107898652553558, | |
| "learning_rate": 4.737350950692878e-06, | |
| "loss": 0.1011, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.9903381642512077, | |
| "grad_norm": 0.22934384644031525, | |
| "learning_rate": 4.71479213664196e-06, | |
| "loss": 0.0715, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.9951690821256038, | |
| "grad_norm": 0.2517627775669098, | |
| "learning_rate": 4.69223332259104e-06, | |
| "loss": 0.0796, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.40475329756736755, | |
| "learning_rate": 4.669674508540122e-06, | |
| "loss": 0.0919, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_runtime": 339.1035, | |
| "eval_samples_per_second": 3.049, | |
| "eval_steps_per_second": 0.383, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.004830917874396, | |
| "grad_norm": 0.23014891147613525, | |
| "learning_rate": 4.647115694489204e-06, | |
| "loss": 0.0721, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.0096618357487923, | |
| "grad_norm": 0.292595773935318, | |
| "learning_rate": 4.624556880438285e-06, | |
| "loss": 0.0797, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.0144927536231885, | |
| "grad_norm": 0.2784234583377838, | |
| "learning_rate": 4.601998066387367e-06, | |
| "loss": 0.0783, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 2.0193236714975846, | |
| "grad_norm": 0.21615320444107056, | |
| "learning_rate": 4.5794392523364485e-06, | |
| "loss": 0.0794, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 2.024154589371981, | |
| "grad_norm": 0.30054816603660583, | |
| "learning_rate": 4.55688043828553e-06, | |
| "loss": 0.078, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 2.028985507246377, | |
| "grad_norm": 0.21918036043643951, | |
| "learning_rate": 4.534321624234612e-06, | |
| "loss": 0.0706, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.033816425120773, | |
| "grad_norm": 0.22675025463104248, | |
| "learning_rate": 4.511762810183694e-06, | |
| "loss": 0.0578, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 2.0386473429951693, | |
| "grad_norm": 0.3500133454799652, | |
| "learning_rate": 4.489203996132775e-06, | |
| "loss": 0.077, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 2.0434782608695654, | |
| "grad_norm": 0.2782948315143585, | |
| "learning_rate": 4.466645182081856e-06, | |
| "loss": 0.0747, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 2.0483091787439616, | |
| "grad_norm": 0.3685343265533447, | |
| "learning_rate": 4.444086368030938e-06, | |
| "loss": 0.0775, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 2.0531400966183573, | |
| "grad_norm": 0.26994946599006653, | |
| "learning_rate": 4.421527553980019e-06, | |
| "loss": 0.076, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.0579710144927534, | |
| "grad_norm": 0.2926693856716156, | |
| "learning_rate": 4.398968739929101e-06, | |
| "loss": 0.0797, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 2.0628019323671496, | |
| "grad_norm": 0.26841118931770325, | |
| "learning_rate": 4.376409925878183e-06, | |
| "loss": 0.0733, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 2.0676328502415457, | |
| "grad_norm": 0.25837743282318115, | |
| "learning_rate": 4.353851111827263e-06, | |
| "loss": 0.0572, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 2.072463768115942, | |
| "grad_norm": 0.23347356915473938, | |
| "learning_rate": 4.331292297776345e-06, | |
| "loss": 0.0824, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 2.077294685990338, | |
| "grad_norm": 0.31139683723449707, | |
| "learning_rate": 4.308733483725427e-06, | |
| "loss": 0.0801, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.082125603864734, | |
| "grad_norm": 0.33561673760414124, | |
| "learning_rate": 4.2861746696745085e-06, | |
| "loss": 0.0816, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.2744121551513672, | |
| "learning_rate": 4.26361585562359e-06, | |
| "loss": 0.0709, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.0917874396135265, | |
| "grad_norm": 0.29332056641578674, | |
| "learning_rate": 4.241057041572672e-06, | |
| "loss": 0.0768, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 2.0966183574879227, | |
| "grad_norm": 0.26820820569992065, | |
| "learning_rate": 4.218498227521753e-06, | |
| "loss": 0.0854, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 2.101449275362319, | |
| "grad_norm": 0.3563501536846161, | |
| "learning_rate": 4.195939413470835e-06, | |
| "loss": 0.0829, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.106280193236715, | |
| "grad_norm": 0.35537421703338623, | |
| "learning_rate": 4.173380599419917e-06, | |
| "loss": 0.0763, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 0.2760440707206726, | |
| "learning_rate": 4.150821785368997e-06, | |
| "loss": 0.092, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 2.1159420289855073, | |
| "grad_norm": 0.21750731766223907, | |
| "learning_rate": 4.128262971318079e-06, | |
| "loss": 0.0756, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 2.1207729468599035, | |
| "grad_norm": 0.2815890610218048, | |
| "learning_rate": 4.105704157267161e-06, | |
| "loss": 0.0844, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 2.1256038647342996, | |
| "grad_norm": 0.20408152043819427, | |
| "learning_rate": 4.083145343216242e-06, | |
| "loss": 0.0603, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.130434782608696, | |
| "grad_norm": 0.2452622503042221, | |
| "learning_rate": 4.060586529165324e-06, | |
| "loss": 0.0767, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 2.135265700483092, | |
| "grad_norm": 0.3027113080024719, | |
| "learning_rate": 4.038027715114405e-06, | |
| "loss": 0.0716, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 2.140096618357488, | |
| "grad_norm": 0.23567864298820496, | |
| "learning_rate": 4.0154689010634865e-06, | |
| "loss": 0.0845, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 2.1449275362318843, | |
| "grad_norm": 0.28407129645347595, | |
| "learning_rate": 3.9929100870125685e-06, | |
| "loss": 0.0784, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.14975845410628, | |
| "grad_norm": 0.28088170289993286, | |
| "learning_rate": 3.9703512729616505e-06, | |
| "loss": 0.0771, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.154589371980676, | |
| "grad_norm": 0.3641108274459839, | |
| "learning_rate": 3.947792458910732e-06, | |
| "loss": 0.0791, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 2.1594202898550723, | |
| "grad_norm": 0.23423610627651215, | |
| "learning_rate": 3.925233644859813e-06, | |
| "loss": 0.0735, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 2.1642512077294684, | |
| "grad_norm": 0.21887804567813873, | |
| "learning_rate": 3.902674830808895e-06, | |
| "loss": 0.0795, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.1690821256038646, | |
| "grad_norm": 0.24810364842414856, | |
| "learning_rate": 3.880116016757976e-06, | |
| "loss": 0.076, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 2.1739130434782608, | |
| "grad_norm": 0.217853844165802, | |
| "learning_rate": 3.857557202707058e-06, | |
| "loss": 0.0794, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.178743961352657, | |
| "grad_norm": 0.28543898463249207, | |
| "learning_rate": 3.834998388656139e-06, | |
| "loss": 0.0707, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 2.183574879227053, | |
| "grad_norm": 0.2932458221912384, | |
| "learning_rate": 3.81243957460522e-06, | |
| "loss": 0.0715, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.1884057971014492, | |
| "grad_norm": 0.3077555000782013, | |
| "learning_rate": 3.789880760554302e-06, | |
| "loss": 0.0756, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 2.1932367149758454, | |
| "grad_norm": 0.295901358127594, | |
| "learning_rate": 3.7673219465033837e-06, | |
| "loss": 0.0785, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 2.1980676328502415, | |
| "grad_norm": 0.2174501270055771, | |
| "learning_rate": 3.7447631324524653e-06, | |
| "loss": 0.0578, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.2028985507246377, | |
| "grad_norm": 0.2652744948863983, | |
| "learning_rate": 3.7222043184015473e-06, | |
| "loss": 0.0579, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.207729468599034, | |
| "grad_norm": 0.34323185682296753, | |
| "learning_rate": 3.699645504350628e-06, | |
| "loss": 0.072, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 2.21256038647343, | |
| "grad_norm": 0.3072277903556824, | |
| "learning_rate": 3.6770866902997096e-06, | |
| "loss": 0.0676, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 2.217391304347826, | |
| "grad_norm": 0.27712109684944153, | |
| "learning_rate": 3.6545278762487916e-06, | |
| "loss": 0.0699, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.2862177789211273, | |
| "learning_rate": 3.631969062197873e-06, | |
| "loss": 0.0643, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.2270531400966185, | |
| "grad_norm": 0.2914809286594391, | |
| "learning_rate": 3.6094102481469547e-06, | |
| "loss": 0.0702, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 2.2318840579710146, | |
| "grad_norm": 0.19755889475345612, | |
| "learning_rate": 3.586851434096036e-06, | |
| "loss": 0.0817, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 2.236714975845411, | |
| "grad_norm": 0.25922340154647827, | |
| "learning_rate": 3.5642926200451175e-06, | |
| "loss": 0.0602, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 2.241545893719807, | |
| "grad_norm": 0.30358242988586426, | |
| "learning_rate": 3.541733805994199e-06, | |
| "loss": 0.0725, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.246376811594203, | |
| "grad_norm": 0.2505339980125427, | |
| "learning_rate": 3.519174991943281e-06, | |
| "loss": 0.079, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.2512077294685993, | |
| "grad_norm": 0.2911323308944702, | |
| "learning_rate": 3.496616177892362e-06, | |
| "loss": 0.0673, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 2.2560386473429954, | |
| "grad_norm": 0.3253360092639923, | |
| "learning_rate": 3.4740573638414437e-06, | |
| "loss": 0.0776, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 2.260869565217391, | |
| "grad_norm": 0.2546384036540985, | |
| "learning_rate": 3.4514985497905253e-06, | |
| "loss": 0.0689, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.2657004830917873, | |
| "grad_norm": 0.29095250368118286, | |
| "learning_rate": 3.428939735739607e-06, | |
| "loss": 0.0812, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 2.2705314009661834, | |
| "grad_norm": 0.29789912700653076, | |
| "learning_rate": 3.4063809216886884e-06, | |
| "loss": 0.087, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.2753623188405796, | |
| "grad_norm": 0.23721112310886383, | |
| "learning_rate": 3.38382210763777e-06, | |
| "loss": 0.075, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 2.2801932367149758, | |
| "grad_norm": 0.2618652284145355, | |
| "learning_rate": 3.3612632935868516e-06, | |
| "loss": 0.0781, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.285024154589372, | |
| "grad_norm": 0.3185523748397827, | |
| "learning_rate": 3.3387044795359327e-06, | |
| "loss": 0.0865, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 2.289855072463768, | |
| "grad_norm": 0.30211564898490906, | |
| "learning_rate": 3.3161456654850143e-06, | |
| "loss": 0.0755, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 2.2946859903381642, | |
| "grad_norm": 0.18218393623828888, | |
| "learning_rate": 3.293586851434096e-06, | |
| "loss": 0.0695, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.2995169082125604, | |
| "grad_norm": 0.20001597702503204, | |
| "learning_rate": 3.2710280373831774e-06, | |
| "loss": 0.0744, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.3043478260869565, | |
| "grad_norm": 0.37984150648117065, | |
| "learning_rate": 3.248469223332259e-06, | |
| "loss": 0.0585, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 2.3091787439613527, | |
| "grad_norm": 0.31228166818618774, | |
| "learning_rate": 3.2259104092813406e-06, | |
| "loss": 0.0731, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.314009661835749, | |
| "grad_norm": 0.27851906418800354, | |
| "learning_rate": 3.203351595230422e-06, | |
| "loss": 0.0767, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 2.318840579710145, | |
| "grad_norm": 0.22976937890052795, | |
| "learning_rate": 3.1807927811795033e-06, | |
| "loss": 0.0738, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.323671497584541, | |
| "grad_norm": 0.24843037128448486, | |
| "learning_rate": 3.1582339671285853e-06, | |
| "loss": 0.0792, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 2.3285024154589373, | |
| "grad_norm": 0.23123487830162048, | |
| "learning_rate": 3.135675153077667e-06, | |
| "loss": 0.0752, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.23363561928272247, | |
| "learning_rate": 3.113116339026748e-06, | |
| "loss": 0.0693, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 2.3381642512077296, | |
| "grad_norm": 0.2371598780155182, | |
| "learning_rate": 3.09055752497583e-06, | |
| "loss": 0.0781, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.342995169082126, | |
| "grad_norm": 0.320534884929657, | |
| "learning_rate": 3.067998710924911e-06, | |
| "loss": 0.0635, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.3478260869565215, | |
| "grad_norm": 0.2920200824737549, | |
| "learning_rate": 3.0454398968739927e-06, | |
| "loss": 0.0771, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.3526570048309177, | |
| "grad_norm": 0.32089921832084656, | |
| "learning_rate": 3.0228810828230747e-06, | |
| "loss": 0.0733, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 2.357487922705314, | |
| "grad_norm": 0.2733156979084015, | |
| "learning_rate": 3.000322268772156e-06, | |
| "loss": 0.0681, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.36231884057971, | |
| "grad_norm": 0.24564507603645325, | |
| "learning_rate": 2.9777634547212374e-06, | |
| "loss": 0.0771, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 2.367149758454106, | |
| "grad_norm": 0.24026136100292206, | |
| "learning_rate": 2.955204640670319e-06, | |
| "loss": 0.0748, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.3719806763285023, | |
| "grad_norm": 0.20703287422657013, | |
| "learning_rate": 2.9326458266194006e-06, | |
| "loss": 0.0688, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 2.3768115942028984, | |
| "grad_norm": 0.18269629776477814, | |
| "learning_rate": 2.910087012568482e-06, | |
| "loss": 0.0728, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.3816425120772946, | |
| "grad_norm": 0.3421408236026764, | |
| "learning_rate": 2.8875281985175637e-06, | |
| "loss": 0.0679, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 2.3864734299516908, | |
| "grad_norm": 0.4087986350059509, | |
| "learning_rate": 2.8649693844666453e-06, | |
| "loss": 0.0791, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.391304347826087, | |
| "grad_norm": 0.2629115879535675, | |
| "learning_rate": 2.8424105704157264e-06, | |
| "loss": 0.074, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.396135265700483, | |
| "grad_norm": 0.2295183390378952, | |
| "learning_rate": 2.8198517563648084e-06, | |
| "loss": 0.0739, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.4009661835748792, | |
| "grad_norm": 0.31765657663345337, | |
| "learning_rate": 2.79729294231389e-06, | |
| "loss": 0.0708, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 2.4057971014492754, | |
| "grad_norm": 0.31528520584106445, | |
| "learning_rate": 2.774734128262971e-06, | |
| "loss": 0.0673, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.4106280193236715, | |
| "grad_norm": 0.2358902543783188, | |
| "learning_rate": 2.752175314212053e-06, | |
| "loss": 0.0543, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 2.4154589371980677, | |
| "grad_norm": 0.2725466787815094, | |
| "learning_rate": 2.7296165001611343e-06, | |
| "loss": 0.0703, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.420289855072464, | |
| "grad_norm": 0.24531903862953186, | |
| "learning_rate": 2.707057686110216e-06, | |
| "loss": 0.0715, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 2.42512077294686, | |
| "grad_norm": 0.29307085275650024, | |
| "learning_rate": 2.6844988720592974e-06, | |
| "loss": 0.0752, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 2.429951690821256, | |
| "grad_norm": 0.2959176003932953, | |
| "learning_rate": 2.661940058008379e-06, | |
| "loss": 0.0685, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.2573854923248291, | |
| "learning_rate": 2.6393812439574605e-06, | |
| "loss": 0.0664, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.4396135265700485, | |
| "grad_norm": 0.3154689371585846, | |
| "learning_rate": 2.616822429906542e-06, | |
| "loss": 0.0615, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.21446138620376587, | |
| "learning_rate": 2.5942636158556237e-06, | |
| "loss": 0.0635, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 2.449275362318841, | |
| "grad_norm": 0.3040371537208557, | |
| "learning_rate": 2.571704801804705e-06, | |
| "loss": 0.0788, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 2.454106280193237, | |
| "grad_norm": 0.2636314034461975, | |
| "learning_rate": 2.5491459877537864e-06, | |
| "loss": 0.072, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.4589371980676327, | |
| "grad_norm": 0.26327863335609436, | |
| "learning_rate": 2.5265871737028684e-06, | |
| "loss": 0.0777, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 2.463768115942029, | |
| "grad_norm": 0.28980839252471924, | |
| "learning_rate": 2.5040283596519495e-06, | |
| "loss": 0.0694, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.468599033816425, | |
| "grad_norm": 0.2889906167984009, | |
| "learning_rate": 2.481469545601031e-06, | |
| "loss": 0.0703, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 2.473429951690821, | |
| "grad_norm": 0.2539612650871277, | |
| "learning_rate": 2.4589107315501127e-06, | |
| "loss": 0.0894, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.4782608695652173, | |
| "grad_norm": 0.25100603699684143, | |
| "learning_rate": 2.4363519174991943e-06, | |
| "loss": 0.0649, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 2.4830917874396135, | |
| "grad_norm": 0.24855615198612213, | |
| "learning_rate": 2.413793103448276e-06, | |
| "loss": 0.0687, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 2.4879227053140096, | |
| "grad_norm": 0.2766883671283722, | |
| "learning_rate": 2.3912342893973574e-06, | |
| "loss": 0.0712, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.4927536231884058, | |
| "grad_norm": 0.24230973422527313, | |
| "learning_rate": 2.368675475346439e-06, | |
| "loss": 0.0792, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.497584541062802, | |
| "grad_norm": 0.2981168031692505, | |
| "learning_rate": 2.34611666129552e-06, | |
| "loss": 0.0724, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 2.502415458937198, | |
| "grad_norm": 0.26249799132347107, | |
| "learning_rate": 2.323557847244602e-06, | |
| "loss": 0.0727, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 2.5072463768115942, | |
| "grad_norm": 0.23193541169166565, | |
| "learning_rate": 2.3009990331936837e-06, | |
| "loss": 0.0658, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 2.5120772946859904, | |
| "grad_norm": 0.3478648364543915, | |
| "learning_rate": 2.278440219142765e-06, | |
| "loss": 0.0766, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.5169082125603865, | |
| "grad_norm": 0.2009768933057785, | |
| "learning_rate": 2.255881405091847e-06, | |
| "loss": 0.0735, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 2.5217391304347827, | |
| "grad_norm": 0.2750122547149658, | |
| "learning_rate": 2.233322591040928e-06, | |
| "loss": 0.0778, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 2.526570048309179, | |
| "grad_norm": 0.22165286540985107, | |
| "learning_rate": 2.2107637769900095e-06, | |
| "loss": 0.0656, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 2.531400966183575, | |
| "grad_norm": 0.26584914326667786, | |
| "learning_rate": 2.1882049629390915e-06, | |
| "loss": 0.0723, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.536231884057971, | |
| "grad_norm": 0.30248183012008667, | |
| "learning_rate": 2.1656461488881727e-06, | |
| "loss": 0.0647, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.541062801932367, | |
| "grad_norm": 0.2667482793331146, | |
| "learning_rate": 2.1430873348372542e-06, | |
| "loss": 0.0694, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 2.545893719806763, | |
| "grad_norm": 0.2767150402069092, | |
| "learning_rate": 2.120528520786336e-06, | |
| "loss": 0.0818, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 2.550724637681159, | |
| "grad_norm": 0.30463531613349915, | |
| "learning_rate": 2.0979697067354174e-06, | |
| "loss": 0.0684, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 0.2667052447795868, | |
| "learning_rate": 2.0754108926844985e-06, | |
| "loss": 0.068, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 2.5603864734299515, | |
| "grad_norm": 0.37567076086997986, | |
| "learning_rate": 2.0528520786335805e-06, | |
| "loss": 0.0578, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.5652173913043477, | |
| "grad_norm": 0.24227222800254822, | |
| "learning_rate": 2.030293264582662e-06, | |
| "loss": 0.0748, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 2.570048309178744, | |
| "grad_norm": 0.3247409760951996, | |
| "learning_rate": 2.0077344505317432e-06, | |
| "loss": 0.073, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.57487922705314, | |
| "grad_norm": 0.30261141061782837, | |
| "learning_rate": 1.9851756364808252e-06, | |
| "loss": 0.0722, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 2.579710144927536, | |
| "grad_norm": 0.2872192859649658, | |
| "learning_rate": 1.9626168224299064e-06, | |
| "loss": 0.0728, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 2.5845410628019323, | |
| "grad_norm": 0.3606136441230774, | |
| "learning_rate": 1.940058008378988e-06, | |
| "loss": 0.0735, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.5893719806763285, | |
| "grad_norm": 0.21871723234653473, | |
| "learning_rate": 1.9174991943280695e-06, | |
| "loss": 0.0682, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.5942028985507246, | |
| "grad_norm": 0.2941882312297821, | |
| "learning_rate": 1.894940380277151e-06, | |
| "loss": 0.0722, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 2.5990338164251208, | |
| "grad_norm": 0.31706181168556213, | |
| "learning_rate": 1.8723815662262327e-06, | |
| "loss": 0.0698, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.603864734299517, | |
| "grad_norm": 0.25599217414855957, | |
| "learning_rate": 1.849822752175314e-06, | |
| "loss": 0.0691, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 0.2954462468624115, | |
| "learning_rate": 1.8272639381243958e-06, | |
| "loss": 0.0831, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.6135265700483092, | |
| "grad_norm": 0.31768399477005005, | |
| "learning_rate": 1.8047051240734774e-06, | |
| "loss": 0.0684, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 2.6183574879227054, | |
| "grad_norm": 0.2380971759557724, | |
| "learning_rate": 1.7821463100225587e-06, | |
| "loss": 0.0604, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.6231884057971016, | |
| "grad_norm": 0.2857172191143036, | |
| "learning_rate": 1.7595874959716405e-06, | |
| "loss": 0.0648, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 2.6280193236714977, | |
| "grad_norm": 0.2866944968700409, | |
| "learning_rate": 1.7370286819207219e-06, | |
| "loss": 0.067, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.632850241545894, | |
| "grad_norm": 0.3259107172489166, | |
| "learning_rate": 1.7144698678698034e-06, | |
| "loss": 0.0789, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.63768115942029, | |
| "grad_norm": 0.23563902080059052, | |
| "learning_rate": 1.691911053818885e-06, | |
| "loss": 0.0826, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.642512077294686, | |
| "grad_norm": 0.33754512667655945, | |
| "learning_rate": 1.6693522397679664e-06, | |
| "loss": 0.0756, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 2.6473429951690823, | |
| "grad_norm": 0.22349333763122559, | |
| "learning_rate": 1.646793425717048e-06, | |
| "loss": 0.0773, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.6521739130434785, | |
| "grad_norm": 0.42616990208625793, | |
| "learning_rate": 1.6242346116661295e-06, | |
| "loss": 0.0676, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 2.6570048309178746, | |
| "grad_norm": 0.27920448780059814, | |
| "learning_rate": 1.601675797615211e-06, | |
| "loss": 0.07, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.661835748792271, | |
| "grad_norm": 0.34114235639572144, | |
| "learning_rate": 1.5791169835642926e-06, | |
| "loss": 0.0807, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.2515537142753601, | |
| "learning_rate": 1.556558169513374e-06, | |
| "loss": 0.0739, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.6714975845410627, | |
| "grad_norm": 0.24267147481441498, | |
| "learning_rate": 1.5339993554624556e-06, | |
| "loss": 0.0727, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 2.676328502415459, | |
| "grad_norm": 0.290988564491272, | |
| "learning_rate": 1.5114405414115374e-06, | |
| "loss": 0.0739, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.681159420289855, | |
| "grad_norm": 0.3821360766887665, | |
| "learning_rate": 1.4888817273606187e-06, | |
| "loss": 0.0792, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.685990338164251, | |
| "grad_norm": 0.284109890460968, | |
| "learning_rate": 1.4663229133097003e-06, | |
| "loss": 0.0767, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.6908212560386473, | |
| "grad_norm": 0.303076833486557, | |
| "learning_rate": 1.4437640992587819e-06, | |
| "loss": 0.0714, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 2.6956521739130435, | |
| "grad_norm": 0.37678495049476624, | |
| "learning_rate": 1.4212052852078632e-06, | |
| "loss": 0.0555, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.7004830917874396, | |
| "grad_norm": 0.23108994960784912, | |
| "learning_rate": 1.398646471156945e-06, | |
| "loss": 0.0833, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 2.7053140096618358, | |
| "grad_norm": 0.3246385157108307, | |
| "learning_rate": 1.3760876571060266e-06, | |
| "loss": 0.076, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.710144927536232, | |
| "grad_norm": 0.2140025794506073, | |
| "learning_rate": 1.353528843055108e-06, | |
| "loss": 0.0791, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 2.714975845410628, | |
| "grad_norm": 0.2923656404018402, | |
| "learning_rate": 1.3309700290041895e-06, | |
| "loss": 0.0892, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.7198067632850242, | |
| "grad_norm": 0.2978055775165558, | |
| "learning_rate": 1.308411214953271e-06, | |
| "loss": 0.0647, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 2.7246376811594204, | |
| "grad_norm": 0.2982514500617981, | |
| "learning_rate": 1.2858524009023524e-06, | |
| "loss": 0.0677, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.7294685990338166, | |
| "grad_norm": 0.2721270024776459, | |
| "learning_rate": 1.2632935868514342e-06, | |
| "loss": 0.0633, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 2.7342995169082127, | |
| "grad_norm": 0.2582114636898041, | |
| "learning_rate": 1.2407347728005156e-06, | |
| "loss": 0.0721, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.7391304347826084, | |
| "grad_norm": 0.2242422103881836, | |
| "learning_rate": 1.2181759587495971e-06, | |
| "loss": 0.0694, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 2.7439613526570046, | |
| "grad_norm": 0.2729090750217438, | |
| "learning_rate": 1.1956171446986787e-06, | |
| "loss": 0.0726, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.7487922705314007, | |
| "grad_norm": 0.34203121066093445, | |
| "learning_rate": 1.17305833064776e-06, | |
| "loss": 0.0796, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 2.753623188405797, | |
| "grad_norm": 0.30749765038490295, | |
| "learning_rate": 1.1504995165968418e-06, | |
| "loss": 0.07, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.758454106280193, | |
| "grad_norm": 0.3750080168247223, | |
| "learning_rate": 1.1279407025459234e-06, | |
| "loss": 0.08, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 2.763285024154589, | |
| "grad_norm": 0.32321617007255554, | |
| "learning_rate": 1.1053818884950048e-06, | |
| "loss": 0.082, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.7681159420289854, | |
| "grad_norm": 0.25304415822029114, | |
| "learning_rate": 1.0828230744440863e-06, | |
| "loss": 0.076, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 2.7729468599033815, | |
| "grad_norm": 0.30696550011634827, | |
| "learning_rate": 1.060264260393168e-06, | |
| "loss": 0.0703, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 0.3218288719654083, | |
| "learning_rate": 1.0377054463422493e-06, | |
| "loss": 0.0696, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.2573774755001068, | |
| "learning_rate": 1.015146632291331e-06, | |
| "loss": 0.0711, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.78743961352657, | |
| "grad_norm": 0.3438413143157959, | |
| "learning_rate": 9.925878182404126e-07, | |
| "loss": 0.0805, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 2.792270531400966, | |
| "grad_norm": 0.3613496422767639, | |
| "learning_rate": 9.70029004189494e-07, | |
| "loss": 0.0742, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.7971014492753623, | |
| "grad_norm": 0.2860325276851654, | |
| "learning_rate": 9.474701901385755e-07, | |
| "loss": 0.0735, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 2.8019323671497585, | |
| "grad_norm": 0.240507572889328, | |
| "learning_rate": 9.24911376087657e-07, | |
| "loss": 0.0677, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.8067632850241546, | |
| "grad_norm": 0.28737547993659973, | |
| "learning_rate": 9.023525620367387e-07, | |
| "loss": 0.0666, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 2.8115942028985508, | |
| "grad_norm": 0.34197041392326355, | |
| "learning_rate": 8.797937479858203e-07, | |
| "loss": 0.0799, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.816425120772947, | |
| "grad_norm": 0.326251745223999, | |
| "learning_rate": 8.572349339349017e-07, | |
| "loss": 0.0691, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 2.821256038647343, | |
| "grad_norm": 0.42289331555366516, | |
| "learning_rate": 8.346761198839832e-07, | |
| "loss": 0.0746, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.8260869565217392, | |
| "grad_norm": 0.28735774755477905, | |
| "learning_rate": 8.121173058330648e-07, | |
| "loss": 0.0782, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 2.8309178743961354, | |
| "grad_norm": 0.29395702481269836, | |
| "learning_rate": 7.895584917821463e-07, | |
| "loss": 0.08, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 2.8357487922705316, | |
| "grad_norm": 0.3306836187839508, | |
| "learning_rate": 7.669996777312278e-07, | |
| "loss": 0.0869, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 2.8405797101449277, | |
| "grad_norm": 0.2740659713745117, | |
| "learning_rate": 7.444408636803094e-07, | |
| "loss": 0.064, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.845410628019324, | |
| "grad_norm": 0.28304237127304077, | |
| "learning_rate": 7.218820496293909e-07, | |
| "loss": 0.0769, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 2.85024154589372, | |
| "grad_norm": 0.3081373870372772, | |
| "learning_rate": 6.993232355784725e-07, | |
| "loss": 0.0783, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.855072463768116, | |
| "grad_norm": 0.3063504099845886, | |
| "learning_rate": 6.76764421527554e-07, | |
| "loss": 0.0643, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 2.8599033816425123, | |
| "grad_norm": 0.2641620635986328, | |
| "learning_rate": 6.542056074766355e-07, | |
| "loss": 0.0658, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.864734299516908, | |
| "grad_norm": 0.3239176869392395, | |
| "learning_rate": 6.316467934257171e-07, | |
| "loss": 0.0677, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 2.869565217391304, | |
| "grad_norm": 0.23815782368183136, | |
| "learning_rate": 6.090879793747986e-07, | |
| "loss": 0.0686, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.8743961352657004, | |
| "grad_norm": 0.26518934965133667, | |
| "learning_rate": 5.8652916532388e-07, | |
| "loss": 0.073, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 2.8792270531400965, | |
| "grad_norm": 0.2455345243215561, | |
| "learning_rate": 5.639703512729617e-07, | |
| "loss": 0.0664, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.8840579710144927, | |
| "grad_norm": 0.2730591893196106, | |
| "learning_rate": 5.414115372220432e-07, | |
| "loss": 0.0745, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.3046686351299286, | |
| "learning_rate": 5.188527231711246e-07, | |
| "loss": 0.0637, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 2.893719806763285, | |
| "grad_norm": 0.26765045523643494, | |
| "learning_rate": 4.962939091202063e-07, | |
| "loss": 0.0818, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 2.898550724637681, | |
| "grad_norm": 0.2611401677131653, | |
| "learning_rate": 4.7373509506928777e-07, | |
| "loss": 0.0871, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.9033816425120773, | |
| "grad_norm": 0.3256029486656189, | |
| "learning_rate": 4.5117628101836934e-07, | |
| "loss": 0.0772, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 2.9082125603864735, | |
| "grad_norm": 0.3779186010360718, | |
| "learning_rate": 4.2861746696745086e-07, | |
| "loss": 0.0709, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 2.9130434782608696, | |
| "grad_norm": 0.248891681432724, | |
| "learning_rate": 4.060586529165324e-07, | |
| "loss": 0.0836, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 2.917874396135266, | |
| "grad_norm": 0.27647843956947327, | |
| "learning_rate": 3.834998388656139e-07, | |
| "loss": 0.0636, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.922705314009662, | |
| "grad_norm": 0.28876233100891113, | |
| "learning_rate": 3.6094102481469546e-07, | |
| "loss": 0.0648, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 2.927536231884058, | |
| "grad_norm": 0.26836660504341125, | |
| "learning_rate": 3.38382210763777e-07, | |
| "loss": 0.0726, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.9323671497584543, | |
| "grad_norm": 0.2655857503414154, | |
| "learning_rate": 3.1582339671285855e-07, | |
| "loss": 0.0736, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 2.9371980676328504, | |
| "grad_norm": 0.30681997537612915, | |
| "learning_rate": 2.9326458266194e-07, | |
| "loss": 0.0688, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.942028985507246, | |
| "grad_norm": 0.3034045994281769, | |
| "learning_rate": 2.707057686110216e-07, | |
| "loss": 0.0611, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 2.9468599033816423, | |
| "grad_norm": 0.24807259440422058, | |
| "learning_rate": 2.4814695456010315e-07, | |
| "loss": 0.0782, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.9516908212560384, | |
| "grad_norm": 0.34220463037490845, | |
| "learning_rate": 2.2558814050918467e-07, | |
| "loss": 0.0751, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "grad_norm": 0.2882407009601593, | |
| "learning_rate": 2.030293264582662e-07, | |
| "loss": 0.0686, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.9613526570048307, | |
| "grad_norm": 0.31148266792297363, | |
| "learning_rate": 1.8047051240734773e-07, | |
| "loss": 0.0668, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 2.966183574879227, | |
| "grad_norm": 0.2847365736961365, | |
| "learning_rate": 1.5791169835642927e-07, | |
| "loss": 0.0785, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.971014492753623, | |
| "grad_norm": 0.2872695028781891, | |
| "learning_rate": 1.353528843055108e-07, | |
| "loss": 0.0723, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 2.975845410628019, | |
| "grad_norm": 0.24350111186504364, | |
| "learning_rate": 1.1279407025459234e-07, | |
| "loss": 0.0669, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.9806763285024154, | |
| "grad_norm": 0.2746003270149231, | |
| "learning_rate": 9.023525620367387e-08, | |
| "loss": 0.0814, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 2.9855072463768115, | |
| "grad_norm": 0.255521684885025, | |
| "learning_rate": 6.76764421527554e-08, | |
| "loss": 0.0774, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.9903381642512077, | |
| "grad_norm": 0.35289525985717773, | |
| "learning_rate": 4.511762810183693e-08, | |
| "loss": 0.0645, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 2.995169082125604, | |
| "grad_norm": 0.279884934425354, | |
| "learning_rate": 2.2558814050918466e-08, | |
| "loss": 0.0738, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4045591652393341, | |
| "learning_rate": 0.0, | |
| "loss": 0.0711, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_runtime": 338.6263, | |
| "eval_samples_per_second": 3.054, | |
| "eval_steps_per_second": 0.384, | |
| "step": 3105 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 3105, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1255832139272192e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |