| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9631901840490797, | |
| "eval_steps": 41, | |
| "global_step": 326, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006134969325153374, | |
| "grad_norm": 0.10048680007457733, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.6143, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006134969325153374, | |
| "eval_loss": 0.49658429622650146, | |
| "eval_runtime": 22.5462, | |
| "eval_samples_per_second": 8.161, | |
| "eval_steps_per_second": 1.02, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012269938650306749, | |
| "grad_norm": 0.08257655799388885, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.4536, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.018404907975460124, | |
| "grad_norm": 0.10207108408212662, | |
| "learning_rate": 3e-06, | |
| "loss": 0.5374, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.024539877300613498, | |
| "grad_norm": 0.07208588719367981, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.4189, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03067484662576687, | |
| "grad_norm": 0.08989891409873962, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4773, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03680981595092025, | |
| "grad_norm": 0.07754336297512054, | |
| "learning_rate": 6e-06, | |
| "loss": 0.5562, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04294478527607362, | |
| "grad_norm": 0.10505057871341705, | |
| "learning_rate": 7e-06, | |
| "loss": 0.5093, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.049079754601226995, | |
| "grad_norm": 0.07071765512228012, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4337, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05521472392638037, | |
| "grad_norm": 0.0959077998995781, | |
| "learning_rate": 9e-06, | |
| "loss": 0.5716, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06134969325153374, | |
| "grad_norm": 0.08926673978567123, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5525, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06748466257668712, | |
| "grad_norm": 0.09146387130022049, | |
| "learning_rate": 9.999752906107043e-06, | |
| "loss": 0.4638, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0736196319018405, | |
| "grad_norm": 0.08329001069068909, | |
| "learning_rate": 9.999011648850328e-06, | |
| "loss": 0.3994, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07975460122699386, | |
| "grad_norm": 0.09074167907238007, | |
| "learning_rate": 9.997776301493914e-06, | |
| "loss": 0.6013, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08588957055214724, | |
| "grad_norm": 0.08691710978746414, | |
| "learning_rate": 9.99604698613651e-06, | |
| "loss": 0.4411, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09202453987730061, | |
| "grad_norm": 0.09161891788244247, | |
| "learning_rate": 9.993823873699427e-06, | |
| "loss": 0.5256, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09815950920245399, | |
| "grad_norm": 0.10627757757902145, | |
| "learning_rate": 9.991107183909665e-06, | |
| "loss": 0.6318, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10429447852760736, | |
| "grad_norm": 0.09262832999229431, | |
| "learning_rate": 9.98789718527821e-06, | |
| "loss": 0.5003, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11042944785276074, | |
| "grad_norm": 0.11589068174362183, | |
| "learning_rate": 9.98419419507348e-06, | |
| "loss": 0.7078, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1165644171779141, | |
| "grad_norm": 0.10624095797538757, | |
| "learning_rate": 9.979998579289985e-06, | |
| "loss": 0.6702, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12269938650306748, | |
| "grad_norm": 0.06414097547531128, | |
| "learning_rate": 9.975310752612138e-06, | |
| "loss": 0.353, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12883435582822086, | |
| "grad_norm": 0.11902766674757004, | |
| "learning_rate": 9.970131178373276e-06, | |
| "loss": 0.5749, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.13496932515337423, | |
| "grad_norm": 0.09129296988248825, | |
| "learning_rate": 9.964460368509868e-06, | |
| "loss": 0.4984, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1411042944785276, | |
| "grad_norm": 0.09726156294345856, | |
| "learning_rate": 9.958298883510904e-06, | |
| "loss": 0.476, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.147239263803681, | |
| "grad_norm": 0.09123346209526062, | |
| "learning_rate": 9.951647332362511e-06, | |
| "loss": 0.4627, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.15337423312883436, | |
| "grad_norm": 0.11005334556102753, | |
| "learning_rate": 9.944506372487754e-06, | |
| "loss": 0.6342, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15950920245398773, | |
| "grad_norm": 0.08414352685213089, | |
| "learning_rate": 9.936876709681668e-06, | |
| "loss": 0.4311, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1656441717791411, | |
| "grad_norm": 0.07989637553691864, | |
| "learning_rate": 9.928759098041482e-06, | |
| "loss": 0.5064, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.17177914110429449, | |
| "grad_norm": 0.08479982614517212, | |
| "learning_rate": 9.920154339892104e-06, | |
| "loss": 0.406, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.17791411042944785, | |
| "grad_norm": 0.09588364511728287, | |
| "learning_rate": 9.911063285706808e-06, | |
| "loss": 0.5324, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.18404907975460122, | |
| "grad_norm": 0.10101877897977829, | |
| "learning_rate": 9.901486834023182e-06, | |
| "loss": 0.5151, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1901840490797546, | |
| "grad_norm": 0.10507562756538391, | |
| "learning_rate": 9.891425931354316e-06, | |
| "loss": 0.5903, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.19631901840490798, | |
| "grad_norm": 0.09090801328420639, | |
| "learning_rate": 9.880881572095255e-06, | |
| "loss": 0.326, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.20245398773006135, | |
| "grad_norm": 0.08976439386606216, | |
| "learning_rate": 9.869854798424709e-06, | |
| "loss": 0.3729, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2085889570552147, | |
| "grad_norm": 0.10010894387960434, | |
| "learning_rate": 9.85834670020205e-06, | |
| "loss": 0.4083, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2147239263803681, | |
| "grad_norm": 0.11611686646938324, | |
| "learning_rate": 9.846358414859598e-06, | |
| "loss": 0.46, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22085889570552147, | |
| "grad_norm": 0.08281069248914719, | |
| "learning_rate": 9.833891127290186e-06, | |
| "loss": 0.4027, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.22699386503067484, | |
| "grad_norm": 0.07023394107818604, | |
| "learning_rate": 9.820946069730067e-06, | |
| "loss": 0.3775, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2331288343558282, | |
| "grad_norm": 0.07550126314163208, | |
| "learning_rate": 9.807524521637103e-06, | |
| "loss": 0.3316, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2392638036809816, | |
| "grad_norm": 0.09305860102176666, | |
| "learning_rate": 9.793627809564324e-06, | |
| "loss": 0.5796, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.24539877300613497, | |
| "grad_norm": 0.07182462513446808, | |
| "learning_rate": 9.779257307028805e-06, | |
| "loss": 0.389, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.25153374233128833, | |
| "grad_norm": 0.0985904186964035, | |
| "learning_rate": 9.76441443437591e-06, | |
| "loss": 0.5509, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.25153374233128833, | |
| "eval_loss": 0.4608064889907837, | |
| "eval_runtime": 22.8858, | |
| "eval_samples_per_second": 8.04, | |
| "eval_steps_per_second": 1.005, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.25766871165644173, | |
| "grad_norm": 0.07043192535638809, | |
| "learning_rate": 9.749100658638914e-06, | |
| "loss": 0.3804, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.26380368098159507, | |
| "grad_norm": 0.0967957079410553, | |
| "learning_rate": 9.733317493394004e-06, | |
| "loss": 0.5316, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.26993865030674846, | |
| "grad_norm": 0.07897109538316727, | |
| "learning_rate": 9.717066498610673e-06, | |
| "loss": 0.4192, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.27607361963190186, | |
| "grad_norm": 0.10683301091194153, | |
| "learning_rate": 9.700349280497552e-06, | |
| "loss": 0.5923, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2822085889570552, | |
| "grad_norm": 0.07347994297742844, | |
| "learning_rate": 9.68316749134364e-06, | |
| "loss": 0.3696, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2883435582822086, | |
| "grad_norm": 0.0915137529373169, | |
| "learning_rate": 9.665522829355005e-06, | |
| "loss": 0.5799, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.294478527607362, | |
| "grad_norm": 0.08099963515996933, | |
| "learning_rate": 9.647417038486936e-06, | |
| "loss": 0.4504, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3006134969325153, | |
| "grad_norm": 0.11759970337152481, | |
| "learning_rate": 9.628851908271572e-06, | |
| "loss": 0.6128, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3067484662576687, | |
| "grad_norm": 0.11251598596572876, | |
| "learning_rate": 9.609829273641034e-06, | |
| "loss": 0.5497, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3128834355828221, | |
| "grad_norm": 0.08890489488840103, | |
| "learning_rate": 9.590351014746059e-06, | |
| "loss": 0.5293, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.31901840490797545, | |
| "grad_norm": 0.08688578754663467, | |
| "learning_rate": 9.570419056770174e-06, | |
| "loss": 0.4923, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.32515337423312884, | |
| "grad_norm": 0.08014731109142303, | |
| "learning_rate": 9.550035369739416e-06, | |
| "loss": 0.4639, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3312883435582822, | |
| "grad_norm": 0.08636078238487244, | |
| "learning_rate": 9.529201968327618e-06, | |
| "loss": 0.3038, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3374233128834356, | |
| "grad_norm": 0.06179236248135567, | |
| "learning_rate": 9.50792091165728e-06, | |
| "loss": 0.3302, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.34355828220858897, | |
| "grad_norm": 0.06599953025579453, | |
| "learning_rate": 9.486194303096062e-06, | |
| "loss": 0.2632, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3496932515337423, | |
| "grad_norm": 0.06800541281700134, | |
| "learning_rate": 9.464024290048879e-06, | |
| "loss": 0.4134, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3558282208588957, | |
| "grad_norm": 0.07703054696321487, | |
| "learning_rate": 9.44141306374566e-06, | |
| "loss": 0.4175, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3619631901840491, | |
| "grad_norm": 0.06939573585987091, | |
| "learning_rate": 9.418362859024781e-06, | |
| "loss": 0.3076, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.36809815950920244, | |
| "grad_norm": 0.07432107627391815, | |
| "learning_rate": 9.39487595411217e-06, | |
| "loss": 0.4667, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.37423312883435583, | |
| "grad_norm": 0.08580990135669708, | |
| "learning_rate": 9.37095467039613e-06, | |
| "loss": 0.4878, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3803680981595092, | |
| "grad_norm": 0.07800393551588058, | |
| "learning_rate": 9.346601372197914e-06, | |
| "loss": 0.4484, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.38650306748466257, | |
| "grad_norm": 0.08790121227502823, | |
| "learning_rate": 9.32181846653802e-06, | |
| "loss": 0.5667, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.39263803680981596, | |
| "grad_norm": 0.07062779366970062, | |
| "learning_rate": 9.296608402898306e-06, | |
| "loss": 0.3125, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3987730061349693, | |
| "grad_norm": 0.08514374494552612, | |
| "learning_rate": 9.270973672979877e-06, | |
| "loss": 0.5389, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4049079754601227, | |
| "grad_norm": 0.09363362193107605, | |
| "learning_rate": 9.244916810456822e-06, | |
| "loss": 0.4613, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4110429447852761, | |
| "grad_norm": 0.08358146995306015, | |
| "learning_rate": 9.218440390725772e-06, | |
| "loss": 0.422, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4171779141104294, | |
| "grad_norm": 0.06681355088949203, | |
| "learning_rate": 9.191547030651383e-06, | |
| "loss": 0.4099, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4233128834355828, | |
| "grad_norm": 0.07471272349357605, | |
| "learning_rate": 9.164239388307668e-06, | |
| "loss": 0.3876, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4294478527607362, | |
| "grad_norm": 0.08724360167980194, | |
| "learning_rate": 9.136520162715288e-06, | |
| "loss": 0.493, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.43558282208588955, | |
| "grad_norm": 0.06159456446766853, | |
| "learning_rate": 9.108392093574785e-06, | |
| "loss": 0.2673, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.44171779141104295, | |
| "grad_norm": 0.0887589380145073, | |
| "learning_rate": 9.079857960995806e-06, | |
| "loss": 0.4282, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.44785276073619634, | |
| "grad_norm": 0.08436176180839539, | |
| "learning_rate": 9.050920585222309e-06, | |
| "loss": 0.503, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4539877300613497, | |
| "grad_norm": 0.08223728090524673, | |
| "learning_rate": 9.021582826353825e-06, | |
| "loss": 0.5367, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4601226993865031, | |
| "grad_norm": 0.07190907746553421, | |
| "learning_rate": 8.991847584062776e-06, | |
| "loss": 0.4085, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4662576687116564, | |
| "grad_norm": 0.0850868821144104, | |
| "learning_rate": 8.961717797307872e-06, | |
| "loss": 0.4974, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4723926380368098, | |
| "grad_norm": 0.1070251539349556, | |
| "learning_rate": 8.931196444043635e-06, | |
| "loss": 0.5619, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4785276073619632, | |
| "grad_norm": 0.07099828869104385, | |
| "learning_rate": 8.900286540926062e-06, | |
| "loss": 0.431, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.48466257668711654, | |
| "grad_norm": 0.06902176886796951, | |
| "learning_rate": 8.868991143014469e-06, | |
| "loss": 0.3375, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.49079754601226994, | |
| "grad_norm": 0.11233729869127274, | |
| "learning_rate": 8.83731334346954e-06, | |
| "loss": 0.5159, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.49693251533742333, | |
| "grad_norm": 0.07816348224878311, | |
| "learning_rate": 8.805256273247597e-06, | |
| "loss": 0.4657, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5030674846625767, | |
| "grad_norm": 0.08207116276025772, | |
| "learning_rate": 8.772823100791152e-06, | |
| "loss": 0.5891, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5030674846625767, | |
| "eval_loss": 0.4219532012939453, | |
| "eval_runtime": 52.8567, | |
| "eval_samples_per_second": 3.481, | |
| "eval_steps_per_second": 0.435, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.50920245398773, | |
| "grad_norm": 0.1063007116317749, | |
| "learning_rate": 8.74001703171574e-06, | |
| "loss": 0.4719, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5153374233128835, | |
| "grad_norm": 0.06525219976902008, | |
| "learning_rate": 8.706841308493092e-06, | |
| "loss": 0.2945, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5214723926380368, | |
| "grad_norm": 0.084320567548275, | |
| "learning_rate": 8.673299210130647e-06, | |
| "loss": 0.5193, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5276073619631901, | |
| "grad_norm": 0.09182199090719223, | |
| "learning_rate": 8.639394051847472e-06, | |
| "loss": 0.4065, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5337423312883436, | |
| "grad_norm": 0.08341530710458755, | |
| "learning_rate": 8.605129184746586e-06, | |
| "loss": 0.5126, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5398773006134969, | |
| "grad_norm": 0.0803583562374115, | |
| "learning_rate": 8.57050799548375e-06, | |
| "loss": 0.4048, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5460122699386503, | |
| "grad_norm": 0.10489048063755035, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.4255, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5521472392638037, | |
| "grad_norm": 0.09766017645597458, | |
| "learning_rate": 8.500210372847128e-06, | |
| "loss": 0.5091, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.558282208588957, | |
| "grad_norm": 0.06364760547876358, | |
| "learning_rate": 8.464540887518638e-06, | |
| "loss": 0.2591, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5644171779141104, | |
| "grad_norm": 0.07464785873889923, | |
| "learning_rate": 8.428528975432067e-06, | |
| "loss": 0.3407, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5705521472392638, | |
| "grad_norm": 0.08410008996725082, | |
| "learning_rate": 8.392178195916832e-06, | |
| "loss": 0.4741, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5766871165644172, | |
| "grad_norm": 0.08058732002973557, | |
| "learning_rate": 8.355492141795185e-06, | |
| "loss": 0.373, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5828220858895705, | |
| "grad_norm": 0.10042490810155869, | |
| "learning_rate": 8.318474439027096e-06, | |
| "loss": 0.4643, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.588957055214724, | |
| "grad_norm": 0.06268122047185898, | |
| "learning_rate": 8.281128746351878e-06, | |
| "loss": 0.3071, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5950920245398773, | |
| "grad_norm": 0.08433736115694046, | |
| "learning_rate": 8.24345875492657e-06, | |
| "loss": 0.4042, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6012269938650306, | |
| "grad_norm": 0.10664895176887512, | |
| "learning_rate": 8.2054681879611e-06, | |
| "loss": 0.5345, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6073619631901841, | |
| "grad_norm": 0.08828828483819962, | |
| "learning_rate": 8.167160800350306e-06, | |
| "loss": 0.4128, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6134969325153374, | |
| "grad_norm": 0.09366817027330399, | |
| "learning_rate": 8.1285403783028e-06, | |
| "loss": 0.6111, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6196319018404908, | |
| "grad_norm": 0.07751886546611786, | |
| "learning_rate": 8.089610738966754e-06, | |
| "loss": 0.4008, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6257668711656442, | |
| "grad_norm": 0.09029467403888702, | |
| "learning_rate": 8.050375730052622e-06, | |
| "loss": 0.5109, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6319018404907976, | |
| "grad_norm": 0.09237638860940933, | |
| "learning_rate": 8.010839229452843e-06, | |
| "loss": 0.3751, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6380368098159509, | |
| "grad_norm": 0.09511607140302658, | |
| "learning_rate": 7.971005144858554e-06, | |
| "loss": 0.3808, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6441717791411042, | |
| "grad_norm": 0.07342271506786346, | |
| "learning_rate": 7.930877413373369e-06, | |
| "loss": 0.3106, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6503067484662577, | |
| "grad_norm": 0.07932678610086441, | |
| "learning_rate": 7.890460001124242e-06, | |
| "loss": 0.3932, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.656441717791411, | |
| "grad_norm": 0.07940706610679626, | |
| "learning_rate": 7.849756902869471e-06, | |
| "loss": 0.4068, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6625766871165644, | |
| "grad_norm": 0.09030912816524506, | |
| "learning_rate": 7.808772141603855e-06, | |
| "loss": 0.431, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6687116564417178, | |
| "grad_norm": 0.091177798807621, | |
| "learning_rate": 7.767509768161079e-06, | |
| "loss": 0.4878, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6748466257668712, | |
| "grad_norm": 0.09804334491491318, | |
| "learning_rate": 7.725973860813338e-06, | |
| "loss": 0.4954, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6809815950920245, | |
| "grad_norm": 0.06426946073770523, | |
| "learning_rate": 7.684168524868253e-06, | |
| "loss": 0.2363, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6871165644171779, | |
| "grad_norm": 0.06771596521139145, | |
| "learning_rate": 7.642097892263098e-06, | |
| "loss": 0.3587, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6932515337423313, | |
| "grad_norm": 0.08947895467281342, | |
| "learning_rate": 7.599766121156436e-06, | |
| "loss": 0.3753, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6993865030674846, | |
| "grad_norm": 0.06724036484956741, | |
| "learning_rate": 7.5571773955171124e-06, | |
| "loss": 0.3099, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7055214723926381, | |
| "grad_norm": 0.08312314003705978, | |
| "learning_rate": 7.5143359247107314e-06, | |
| "loss": 0.3134, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7116564417177914, | |
| "grad_norm": 0.0804123803973198, | |
| "learning_rate": 7.471245943083615e-06, | |
| "loss": 0.3077, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7177914110429447, | |
| "grad_norm": 0.08391325175762177, | |
| "learning_rate": 7.427911709544288e-06, | |
| "loss": 0.416, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7239263803680982, | |
| "grad_norm": 0.067812480032444, | |
| "learning_rate": 7.3843375071425315e-06, | |
| "loss": 0.2535, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7300613496932515, | |
| "grad_norm": 0.103757344186306, | |
| "learning_rate": 7.340527642646069e-06, | |
| "loss": 0.5859, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7361963190184049, | |
| "grad_norm": 0.07385145872831345, | |
| "learning_rate": 7.2964864461148895e-06, | |
| "loss": 0.3527, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7423312883435583, | |
| "grad_norm": 0.08877512067556381, | |
| "learning_rate": 7.252218270473274e-06, | |
| "loss": 0.5138, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7484662576687117, | |
| "grad_norm": 0.06157367303967476, | |
| "learning_rate": 7.2077274910795605e-06, | |
| "loss": 0.2458, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.754601226993865, | |
| "grad_norm": 0.17377746105194092, | |
| "learning_rate": 7.163018505293703e-06, | |
| "loss": 0.3547, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.754601226993865, | |
| "eval_loss": 0.40320152044296265, | |
| "eval_runtime": 22.6808, | |
| "eval_samples_per_second": 8.113, | |
| "eval_steps_per_second": 1.014, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7607361963190185, | |
| "grad_norm": 0.08620224893093109, | |
| "learning_rate": 7.118095732042643e-06, | |
| "loss": 0.5084, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7668711656441718, | |
| "grad_norm": 0.07054693251848221, | |
| "learning_rate": 7.072963611383545e-06, | |
| "loss": 0.2391, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7730061349693251, | |
| "grad_norm": 0.1339060664176941, | |
| "learning_rate": 7.02762660406497e-06, | |
| "loss": 0.6351, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7791411042944786, | |
| "grad_norm": 0.07386178523302078, | |
| "learning_rate": 6.982089191085971e-06, | |
| "loss": 0.3047, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7852760736196319, | |
| "grad_norm": 0.10029296576976776, | |
| "learning_rate": 6.936355873253207e-06, | |
| "loss": 0.4328, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7914110429447853, | |
| "grad_norm": 0.09643431752920151, | |
| "learning_rate": 6.8904311707360914e-06, | |
| "loss": 0.4234, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7975460122699386, | |
| "grad_norm": 0.08629649877548218, | |
| "learning_rate": 6.844319622620039e-06, | |
| "loss": 0.3454, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.803680981595092, | |
| "grad_norm": 0.09917541593313217, | |
| "learning_rate": 6.798025786457825e-06, | |
| "loss": 0.4828, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8098159509202454, | |
| "grad_norm": 0.08652715384960175, | |
| "learning_rate": 6.751554237819122e-06, | |
| "loss": 0.3921, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8159509202453987, | |
| "grad_norm": 0.08848355710506439, | |
| "learning_rate": 6.704909569838281e-06, | |
| "loss": 0.3573, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8220858895705522, | |
| "grad_norm": 0.09000125527381897, | |
| "learning_rate": 6.65809639276034e-06, | |
| "loss": 0.4486, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8282208588957055, | |
| "grad_norm": 0.08975492417812347, | |
| "learning_rate": 6.611119333485364e-06, | |
| "loss": 0.3867, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8343558282208589, | |
| "grad_norm": 0.0796804279088974, | |
| "learning_rate": 6.563983035111136e-06, | |
| "loss": 0.381, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8404907975460123, | |
| "grad_norm": 0.08723866939544678, | |
| "learning_rate": 6.516692156474243e-06, | |
| "loss": 0.4316, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8466257668711656, | |
| "grad_norm": 0.11449507623910904, | |
| "learning_rate": 6.469251371689606e-06, | |
| "loss": 0.5042, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.852760736196319, | |
| "grad_norm": 0.127399280667305, | |
| "learning_rate": 6.421665369688501e-06, | |
| "loss": 0.56, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.8588957055214724, | |
| "grad_norm": 0.0761680155992508, | |
| "learning_rate": 6.373938853755126e-06, | |
| "loss": 0.2881, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8650306748466258, | |
| "grad_norm": 0.08487329632043839, | |
| "learning_rate": 6.326076541061729e-06, | |
| "loss": 0.3171, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8711656441717791, | |
| "grad_norm": 0.09721650928258896, | |
| "learning_rate": 6.278083162202374e-06, | |
| "loss": 0.4289, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8773006134969326, | |
| "grad_norm": 0.08022255450487137, | |
| "learning_rate": 6.22996346072539e-06, | |
| "loss": 0.327, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8834355828220859, | |
| "grad_norm": 0.11367864906787872, | |
| "learning_rate": 6.181722192664526e-06, | |
| "loss": 0.4132, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8895705521472392, | |
| "grad_norm": 0.08727509528398514, | |
| "learning_rate": 6.133364126068867e-06, | |
| "loss": 0.3467, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8957055214723927, | |
| "grad_norm": 0.09527825564146042, | |
| "learning_rate": 6.084894040531591e-06, | |
| "loss": 0.5279, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.901840490797546, | |
| "grad_norm": 0.07519163191318512, | |
| "learning_rate": 6.036316726717546e-06, | |
| "loss": 0.3221, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9079754601226994, | |
| "grad_norm": 0.08815211802721024, | |
| "learning_rate": 5.987636985889764e-06, | |
| "loss": 0.347, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9141104294478528, | |
| "grad_norm": 0.10302853584289551, | |
| "learning_rate": 5.938859629434913e-06, | |
| "loss": 0.4378, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9202453987730062, | |
| "grad_norm": 0.09466961026191711, | |
| "learning_rate": 5.8899894783877536e-06, | |
| "loss": 0.3053, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9263803680981595, | |
| "grad_norm": 0.09226138889789581, | |
| "learning_rate": 5.841031362954629e-06, | |
| "loss": 0.4522, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9325153374233128, | |
| "grad_norm": 0.10703295469284058, | |
| "learning_rate": 5.791990122036075e-06, | |
| "loss": 0.4592, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9386503067484663, | |
| "grad_norm": 0.1230417788028717, | |
| "learning_rate": 5.742870602748547e-06, | |
| "loss": 0.5402, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9447852760736196, | |
| "grad_norm": 0.09710447490215302, | |
| "learning_rate": 5.693677659945343e-06, | |
| "loss": 0.4628, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.950920245398773, | |
| "grad_norm": 0.11172331869602203, | |
| "learning_rate": 5.6444161557367534e-06, | |
| "loss": 0.3569, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9570552147239264, | |
| "grad_norm": 0.09823194146156311, | |
| "learning_rate": 5.595090959009525e-06, | |
| "loss": 0.4716, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.9631901840490797, | |
| "grad_norm": 0.08101484179496765, | |
| "learning_rate": 5.5457069449456055e-06, | |
| "loss": 0.4065, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9693251533742331, | |
| "grad_norm": 0.08857468515634537, | |
| "learning_rate": 5.496268994540309e-06, | |
| "loss": 0.3606, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9754601226993865, | |
| "grad_norm": 0.07157395780086517, | |
| "learning_rate": 5.446781994119886e-06, | |
| "loss": 0.2453, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9815950920245399, | |
| "grad_norm": 0.07760554552078247, | |
| "learning_rate": 5.397250834858573e-06, | |
| "loss": 0.364, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9877300613496932, | |
| "grad_norm": 0.07792805880308151, | |
| "learning_rate": 5.347680412295152e-06, | |
| "loss": 0.3496, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9938650306748467, | |
| "grad_norm": 0.08258286118507385, | |
| "learning_rate": 5.2980756258491e-06, | |
| "loss": 0.3455, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.11803517490625381, | |
| "learning_rate": 5.2484413783363335e-06, | |
| "loss": 0.4871, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0061349693251533, | |
| "grad_norm": 0.10544019937515259, | |
| "learning_rate": 5.19878257548463e-06, | |
| "loss": 0.4369, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0061349693251533, | |
| "eval_loss": 0.39176106452941895, | |
| "eval_runtime": 22.8177, | |
| "eval_samples_per_second": 8.064, | |
| "eval_steps_per_second": 1.008, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0122699386503067, | |
| "grad_norm": 0.09844586253166199, | |
| "learning_rate": 5.149104125448752e-06, | |
| "loss": 0.4385, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.01840490797546, | |
| "grad_norm": 0.07265810668468475, | |
| "learning_rate": 5.099410938325351e-06, | |
| "loss": 0.2595, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0245398773006136, | |
| "grad_norm": 0.12025801837444305, | |
| "learning_rate": 5.04970792566765e-06, | |
| "loss": 0.5587, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.030674846625767, | |
| "grad_norm": 0.1306670904159546, | |
| "learning_rate": 5e-06, | |
| "loss": 0.603, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0368098159509203, | |
| "grad_norm": 0.10355333238840103, | |
| "learning_rate": 4.9502920743323525e-06, | |
| "loss": 0.3555, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0061349693251533, | |
| "grad_norm": 0.10698702186346054, | |
| "learning_rate": 4.900589061674649e-06, | |
| "loss": 0.4452, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0122699386503067, | |
| "grad_norm": 0.0719093456864357, | |
| "learning_rate": 4.850895874551248e-06, | |
| "loss": 0.3209, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.01840490797546, | |
| "grad_norm": 0.07391523569822311, | |
| "learning_rate": 4.801217424515373e-06, | |
| "loss": 0.2961, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.0245398773006136, | |
| "grad_norm": 0.07925672829151154, | |
| "learning_rate": 4.751558621663668e-06, | |
| "loss": 0.3446, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.030674846625767, | |
| "grad_norm": 0.10210183262825012, | |
| "learning_rate": 4.701924374150901e-06, | |
| "loss": 0.49, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.0368098159509203, | |
| "grad_norm": 0.11506979912519455, | |
| "learning_rate": 4.6523195877048495e-06, | |
| "loss": 0.2977, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.0429447852760736, | |
| "grad_norm": 0.09373599290847778, | |
| "learning_rate": 4.602749165141429e-06, | |
| "loss": 0.3853, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.049079754601227, | |
| "grad_norm": 0.09830620139837265, | |
| "learning_rate": 4.5532180058801145e-06, | |
| "loss": 0.4363, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.0552147239263803, | |
| "grad_norm": 0.08626256883144379, | |
| "learning_rate": 4.5037310054596936e-06, | |
| "loss": 0.4265, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.0613496932515338, | |
| "grad_norm": 0.11579606682062149, | |
| "learning_rate": 4.454293055054397e-06, | |
| "loss": 0.5565, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.0674846625766872, | |
| "grad_norm": 0.10661870241165161, | |
| "learning_rate": 4.404909040990477e-06, | |
| "loss": 0.5136, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0736196319018405, | |
| "grad_norm": 0.09875867515802383, | |
| "learning_rate": 4.355583844263247e-06, | |
| "loss": 0.5337, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.0797546012269938, | |
| "grad_norm": 0.2183881402015686, | |
| "learning_rate": 4.30632234005466e-06, | |
| "loss": 0.3483, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.0858895705521472, | |
| "grad_norm": 0.0771089717745781, | |
| "learning_rate": 4.257129397251453e-06, | |
| "loss": 0.3027, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.0920245398773005, | |
| "grad_norm": 0.10682545602321625, | |
| "learning_rate": 4.2080098779639255e-06, | |
| "loss": 0.4408, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.098159509202454, | |
| "grad_norm": 0.08865021914243698, | |
| "learning_rate": 4.158968637045374e-06, | |
| "loss": 0.4054, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1042944785276074, | |
| "grad_norm": 0.11121812462806702, | |
| "learning_rate": 4.11001052161225e-06, | |
| "loss": 0.3993, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.1104294478527608, | |
| "grad_norm": 0.10253646969795227, | |
| "learning_rate": 4.061140370565088e-06, | |
| "loss": 0.4388, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.116564417177914, | |
| "grad_norm": 0.09401492774486542, | |
| "learning_rate": 4.012363014110237e-06, | |
| "loss": 0.4846, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.1226993865030674, | |
| "grad_norm": 0.1021651104092598, | |
| "learning_rate": 3.9636832732824555e-06, | |
| "loss": 0.312, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.1288343558282208, | |
| "grad_norm": 0.12308547645807266, | |
| "learning_rate": 3.91510595946841e-06, | |
| "loss": 0.5238, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.1349693251533743, | |
| "grad_norm": 0.1113848015666008, | |
| "learning_rate": 3.866635873931133e-06, | |
| "loss": 0.4818, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.1411042944785277, | |
| "grad_norm": 0.10525127500295639, | |
| "learning_rate": 3.818277807335477e-06, | |
| "loss": 0.381, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.147239263803681, | |
| "grad_norm": 0.12201809883117676, | |
| "learning_rate": 3.7700365392746106e-06, | |
| "loss": 0.3412, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.1533742331288344, | |
| "grad_norm": 0.08116313070058823, | |
| "learning_rate": 3.721916837797627e-06, | |
| "loss": 0.3486, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.1595092024539877, | |
| "grad_norm": 0.11550577729940414, | |
| "learning_rate": 3.6739234589382722e-06, | |
| "loss": 0.3337, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.165644171779141, | |
| "grad_norm": 0.08885617554187775, | |
| "learning_rate": 3.6260611462448736e-06, | |
| "loss": 0.27, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.1717791411042944, | |
| "grad_norm": 0.1086370199918747, | |
| "learning_rate": 3.5783346303114986e-06, | |
| "loss": 0.3044, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.177914110429448, | |
| "grad_norm": 0.10726229846477509, | |
| "learning_rate": 3.5307486283103966e-06, | |
| "loss": 0.4503, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.1840490797546013, | |
| "grad_norm": 0.0863179937005043, | |
| "learning_rate": 3.4833078435257584e-06, | |
| "loss": 0.2563, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.1901840490797546, | |
| "grad_norm": 0.14450977742671967, | |
| "learning_rate": 3.4360169648888653e-06, | |
| "loss": 0.5502, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.196319018404908, | |
| "grad_norm": 0.10789606720209122, | |
| "learning_rate": 3.388880666514637e-06, | |
| "loss": 0.3767, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.2024539877300613, | |
| "grad_norm": 0.11021628230810165, | |
| "learning_rate": 3.3419036072396614e-06, | |
| "loss": 0.4526, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2085889570552146, | |
| "grad_norm": 0.10493913292884827, | |
| "learning_rate": 3.29509043016172e-06, | |
| "loss": 0.3998, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.2147239263803682, | |
| "grad_norm": 0.12165309488773346, | |
| "learning_rate": 3.2484457621808787e-06, | |
| "loss": 0.383, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.2208588957055215, | |
| "grad_norm": 0.09394080936908722, | |
| "learning_rate": 3.201974213542178e-06, | |
| "loss": 0.4129, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.2208588957055215, | |
| "eval_loss": 0.38451722264289856, | |
| "eval_runtime": 22.6218, | |
| "eval_samples_per_second": 8.134, | |
| "eval_steps_per_second": 1.017, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.2269938650306749, | |
| "grad_norm": 0.09802668541669846, | |
| "learning_rate": 3.1556803773799616e-06, | |
| "loss": 0.408, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.2331288343558282, | |
| "grad_norm": 0.13217565417289734, | |
| "learning_rate": 3.1095688292639094e-06, | |
| "loss": 0.3198, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.2392638036809815, | |
| "grad_norm": 0.10313939303159714, | |
| "learning_rate": 3.0636441267467955e-06, | |
| "loss": 0.3128, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.2453987730061349, | |
| "grad_norm": 0.0762137621641159, | |
| "learning_rate": 3.01791080891403e-06, | |
| "loss": 0.2638, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.2515337423312882, | |
| "grad_norm": 0.11147941648960114, | |
| "learning_rate": 2.972373395935031e-06, | |
| "loss": 0.4534, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.2576687116564418, | |
| "grad_norm": 0.09862508624792099, | |
| "learning_rate": 2.927036388616457e-06, | |
| "loss": 0.4292, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.2638036809815951, | |
| "grad_norm": 0.09868728369474411, | |
| "learning_rate": 2.8819042679573618e-06, | |
| "loss": 0.3276, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.2699386503067485, | |
| "grad_norm": 0.08501134812831879, | |
| "learning_rate": 2.8369814947062994e-06, | |
| "loss": 0.3095, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.2760736196319018, | |
| "grad_norm": 0.11438091844320297, | |
| "learning_rate": 2.792272508920443e-06, | |
| "loss": 0.295, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.2822085889570551, | |
| "grad_norm": 0.11010921746492386, | |
| "learning_rate": 2.7477817295267273e-06, | |
| "loss": 0.433, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.2883435582822087, | |
| "grad_norm": 0.08820465952157974, | |
| "learning_rate": 2.70351355388511e-06, | |
| "loss": 0.3045, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.294478527607362, | |
| "grad_norm": 0.08259084820747375, | |
| "learning_rate": 2.6594723573539307e-06, | |
| "loss": 0.3812, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.3006134969325154, | |
| "grad_norm": 0.11154909431934357, | |
| "learning_rate": 2.615662492857471e-06, | |
| "loss": 0.3513, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.3067484662576687, | |
| "grad_norm": 0.0986664667725563, | |
| "learning_rate": 2.5720882904557156e-06, | |
| "loss": 0.3658, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.312883435582822, | |
| "grad_norm": 0.2315550446510315, | |
| "learning_rate": 2.528754056916386e-06, | |
| "loss": 0.5219, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.3190184049079754, | |
| "grad_norm": 0.10897372663021088, | |
| "learning_rate": 2.4856640752892702e-06, | |
| "loss": 0.4178, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.3251533742331287, | |
| "grad_norm": 0.14039960503578186, | |
| "learning_rate": 2.4428226044828896e-06, | |
| "loss": 0.4113, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.331288343558282, | |
| "grad_norm": 0.10824128240346909, | |
| "learning_rate": 2.4002338788435654e-06, | |
| "loss": 0.5079, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.3374233128834356, | |
| "grad_norm": 0.10082822293043137, | |
| "learning_rate": 2.3579021077369047e-06, | |
| "loss": 0.3356, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.343558282208589, | |
| "grad_norm": 0.1202516257762909, | |
| "learning_rate": 2.315831475131751e-06, | |
| "loss": 0.3236, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.3496932515337423, | |
| "grad_norm": 0.1029471606016159, | |
| "learning_rate": 2.2740261391866634e-06, | |
| "loss": 0.3491, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.3558282208588956, | |
| "grad_norm": 0.12975585460662842, | |
| "learning_rate": 2.232490231838923e-06, | |
| "loss": 0.402, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.3619631901840492, | |
| "grad_norm": 0.12482481449842453, | |
| "learning_rate": 2.1912278583961454e-06, | |
| "loss": 0.4321, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.3680981595092025, | |
| "grad_norm": 0.1507352739572525, | |
| "learning_rate": 2.1502430971305288e-06, | |
| "loss": 0.5751, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.3742331288343559, | |
| "grad_norm": 0.0954100489616394, | |
| "learning_rate": 2.1095399988757574e-06, | |
| "loss": 0.3933, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.3803680981595092, | |
| "grad_norm": 0.10602657496929169, | |
| "learning_rate": 2.0691225866266335e-06, | |
| "loss": 0.4803, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.3865030674846626, | |
| "grad_norm": 0.11936960369348526, | |
| "learning_rate": 2.0289948551414486e-06, | |
| "loss": 0.3983, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.392638036809816, | |
| "grad_norm": 0.08489519357681274, | |
| "learning_rate": 1.989160770547159e-06, | |
| "loss": 0.2552, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.3987730061349692, | |
| "grad_norm": 0.10064487159252167, | |
| "learning_rate": 1.949624269947378e-06, | |
| "loss": 0.4482, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.4049079754601226, | |
| "grad_norm": 0.10288871824741364, | |
| "learning_rate": 1.9103892610332467e-06, | |
| "loss": 0.3787, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.4110429447852761, | |
| "grad_norm": 0.09945371747016907, | |
| "learning_rate": 1.8714596216972008e-06, | |
| "loss": 0.3845, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.4171779141104295, | |
| "grad_norm": 0.11584262549877167, | |
| "learning_rate": 1.8328391996496942e-06, | |
| "loss": 0.5336, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.4233128834355828, | |
| "grad_norm": 0.08862798660993576, | |
| "learning_rate": 1.794531812038901e-06, | |
| "loss": 0.3253, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.4294478527607362, | |
| "grad_norm": 0.12273416668176651, | |
| "learning_rate": 1.756541245073432e-06, | |
| "loss": 0.4336, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.4355828220858895, | |
| "grad_norm": 0.08479359745979309, | |
| "learning_rate": 1.7188712536481233e-06, | |
| "loss": 0.3385, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.441717791411043, | |
| "grad_norm": 0.08785238116979599, | |
| "learning_rate": 1.6815255609729047e-06, | |
| "loss": 0.3856, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.4478527607361964, | |
| "grad_norm": 0.11194564402103424, | |
| "learning_rate": 1.6445078582048158e-06, | |
| "loss": 0.4059, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.4539877300613497, | |
| "grad_norm": 0.10265929996967316, | |
| "learning_rate": 1.6078218040831678e-06, | |
| "loss": 0.5095, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.460122699386503, | |
| "grad_norm": 0.1036793664097786, | |
| "learning_rate": 1.5714710245679348e-06, | |
| "loss": 0.3271, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.4662576687116564, | |
| "grad_norm": 0.16661570966243744, | |
| "learning_rate": 1.5354591124813628e-06, | |
| "loss": 0.4891, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.4723926380368098, | |
| "grad_norm": 0.08812650293111801, | |
| "learning_rate": 1.499789627152874e-06, | |
| "loss": 0.335, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.4723926380368098, | |
| "eval_loss": 0.3808572590351105, | |
| "eval_runtime": 22.6328, | |
| "eval_samples_per_second": 8.13, | |
| "eval_steps_per_second": 1.016, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.478527607361963, | |
| "grad_norm": 0.1271064132452011, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.4646, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.4846625766871164, | |
| "grad_norm": 0.10755060613155365, | |
| "learning_rate": 1.4294920045162514e-06, | |
| "loss": 0.4369, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.49079754601227, | |
| "grad_norm": 0.10873299837112427, | |
| "learning_rate": 1.3948708152534163e-06, | |
| "loss": 0.3521, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.4969325153374233, | |
| "grad_norm": 0.15240046381950378, | |
| "learning_rate": 1.3606059481525296e-06, | |
| "loss": 0.4, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.5030674846625767, | |
| "grad_norm": 0.158911794424057, | |
| "learning_rate": 1.3267007898693552e-06, | |
| "loss": 0.3103, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.50920245398773, | |
| "grad_norm": 0.11236365884542465, | |
| "learning_rate": 1.2931586915069106e-06, | |
| "loss": 0.3715, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.5153374233128836, | |
| "grad_norm": 0.1483219861984253, | |
| "learning_rate": 1.2599829682842618e-06, | |
| "loss": 0.4046, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.521472392638037, | |
| "grad_norm": 0.09588169306516647, | |
| "learning_rate": 1.227176899208849e-06, | |
| "loss": 0.3077, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.5276073619631902, | |
| "grad_norm": 0.10631563514471054, | |
| "learning_rate": 1.194743726752403e-06, | |
| "loss": 0.4418, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.5337423312883436, | |
| "grad_norm": 0.10569003224372864, | |
| "learning_rate": 1.1626866565304594e-06, | |
| "loss": 0.4616, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.539877300613497, | |
| "grad_norm": 0.11209447681903839, | |
| "learning_rate": 1.1310088569855315e-06, | |
| "loss": 0.4556, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.5460122699386503, | |
| "grad_norm": 0.14903852343559265, | |
| "learning_rate": 1.09971345907394e-06, | |
| "loss": 0.3381, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.5521472392638036, | |
| "grad_norm": 0.10578689724206924, | |
| "learning_rate": 1.068803555956367e-06, | |
| "loss": 0.3798, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.558282208588957, | |
| "grad_norm": 0.12829791009426117, | |
| "learning_rate": 1.0382822026921291e-06, | |
| "loss": 0.5217, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.5644171779141103, | |
| "grad_norm": 0.12076660990715027, | |
| "learning_rate": 1.0081524159372246e-06, | |
| "loss": 0.4557, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.5705521472392638, | |
| "grad_norm": 0.11094032227993011, | |
| "learning_rate": 9.784171736461762e-07, | |
| "loss": 0.3339, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.5766871165644172, | |
| "grad_norm": 0.09728420525789261, | |
| "learning_rate": 9.490794147776927e-07, | |
| "loss": 0.3506, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.5828220858895705, | |
| "grad_norm": 0.1307501345872879, | |
| "learning_rate": 9.201420390041965e-07, | |
| "loss": 0.3652, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.588957055214724, | |
| "grad_norm": 0.11151021718978882, | |
| "learning_rate": 8.916079064252164e-07, | |
| "loss": 0.3796, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.5950920245398774, | |
| "grad_norm": 0.091631218791008, | |
| "learning_rate": 8.634798372847148e-07, | |
| "loss": 0.2796, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.6012269938650308, | |
| "grad_norm": 0.10448212176561356, | |
| "learning_rate": 8.357606116923328e-07, | |
| "loss": 0.2626, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.607361963190184, | |
| "grad_norm": 0.11236970871686935, | |
| "learning_rate": 8.084529693486171e-07, | |
| "loss": 0.3601, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.6134969325153374, | |
| "grad_norm": 0.11304070055484772, | |
| "learning_rate": 7.815596092742278e-07, | |
| "loss": 0.3641, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.6196319018404908, | |
| "grad_norm": 0.4070085883140564, | |
| "learning_rate": 7.550831895431799e-07, | |
| "loss": 0.4316, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.6257668711656441, | |
| "grad_norm": 0.10678707808256149, | |
| "learning_rate": 7.290263270201231e-07, | |
| "loss": 0.4281, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.6319018404907975, | |
| "grad_norm": 0.10431778430938721, | |
| "learning_rate": 7.033915971016952e-07, | |
| "loss": 0.3907, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.6380368098159508, | |
| "grad_norm": 0.10240163654088974, | |
| "learning_rate": 6.781815334619812e-07, | |
| "loss": 0.3344, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.6441717791411041, | |
| "grad_norm": 0.10372646898031235, | |
| "learning_rate": 6.533986278020876e-07, | |
| "loss": 0.373, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.6503067484662577, | |
| "grad_norm": 0.12202878296375275, | |
| "learning_rate": 6.290453296038702e-07, | |
| "loss": 0.4087, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.656441717791411, | |
| "grad_norm": 0.09939797222614288, | |
| "learning_rate": 6.051240458878316e-07, | |
| "loss": 0.3611, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.6625766871165644, | |
| "grad_norm": 0.12352439016103745, | |
| "learning_rate": 5.816371409752203e-07, | |
| "loss": 0.5242, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.668711656441718, | |
| "grad_norm": 0.11458908766508102, | |
| "learning_rate": 5.585869362543416e-07, | |
| "loss": 0.4827, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.6748466257668713, | |
| "grad_norm": 0.14145226776599884, | |
| "learning_rate": 5.359757099511237e-07, | |
| "loss": 0.3911, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.6809815950920246, | |
| "grad_norm": 0.09614825993776321, | |
| "learning_rate": 5.138056969039384e-07, | |
| "loss": 0.3729, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.687116564417178, | |
| "grad_norm": 0.10174663364887238, | |
| "learning_rate": 4.920790883427201e-07, | |
| "loss": 0.4666, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.6932515337423313, | |
| "grad_norm": 0.09838556498289108, | |
| "learning_rate": 4.707980316723837e-07, | |
| "loss": 0.4527, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.6993865030674846, | |
| "grad_norm": 0.09283680468797684, | |
| "learning_rate": 4.4996463026058476e-07, | |
| "loss": 0.3644, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.705521472392638, | |
| "grad_norm": 0.1122664213180542, | |
| "learning_rate": 4.2958094322982703e-07, | |
| "loss": 0.344, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.7116564417177913, | |
| "grad_norm": 0.15890415012836456, | |
| "learning_rate": 4.096489852539426e-07, | |
| "loss": 0.647, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.7177914110429446, | |
| "grad_norm": 0.13360188901424408, | |
| "learning_rate": 3.9017072635896716e-07, | |
| "loss": 0.5654, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.7239263803680982, | |
| "grad_norm": 0.10913598537445068, | |
| "learning_rate": 3.7114809172842827e-07, | |
| "loss": 0.3842, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.7239263803680982, | |
| "eval_loss": 0.379504919052124, | |
| "eval_runtime": 22.6056, | |
| "eval_samples_per_second": 8.14, | |
| "eval_steps_per_second": 1.017, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.7300613496932515, | |
| "grad_norm": 0.09505796432495117, | |
| "learning_rate": 3.5258296151306495e-07, | |
| "loss": 0.2472, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.7361963190184049, | |
| "grad_norm": 0.10274416208267212, | |
| "learning_rate": 3.3447717064499565e-07, | |
| "loss": 0.4665, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.7423312883435584, | |
| "grad_norm": 0.11168382316827774, | |
| "learning_rate": 3.168325086563612e-07, | |
| "loss": 0.4719, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.7484662576687118, | |
| "grad_norm": 0.121711865067482, | |
| "learning_rate": 2.996507195024495e-07, | |
| "loss": 0.3143, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.7546012269938651, | |
| "grad_norm": 0.0906308963894844, | |
| "learning_rate": 2.8293350138932805e-07, | |
| "loss": 0.4019, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.7607361963190185, | |
| "grad_norm": 0.12006894499063492, | |
| "learning_rate": 2.666825066059986e-07, | |
| "loss": 0.3327, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.7668711656441718, | |
| "grad_norm": 0.12550675868988037, | |
| "learning_rate": 2.5089934136108665e-07, | |
| "loss": 0.4567, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.7730061349693251, | |
| "grad_norm": 0.11931753158569336, | |
| "learning_rate": 2.3558556562409074e-07, | |
| "loss": 0.5051, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.7791411042944785, | |
| "grad_norm": 0.13427190482616425, | |
| "learning_rate": 2.2074269297119588e-07, | |
| "loss": 0.4541, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.7852760736196318, | |
| "grad_norm": 0.09217509627342224, | |
| "learning_rate": 2.0637219043567636e-07, | |
| "loss": 0.3891, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.7914110429447851, | |
| "grad_norm": 0.11418317258358002, | |
| "learning_rate": 1.9247547836289792e-07, | |
| "loss": 0.3991, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.7975460122699385, | |
| "grad_norm": 0.09878183901309967, | |
| "learning_rate": 1.7905393026993513e-07, | |
| "loss": 0.3447, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.803680981595092, | |
| "grad_norm": 0.10060778260231018, | |
| "learning_rate": 1.6610887270981425e-07, | |
| "loss": 0.3269, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.8098159509202454, | |
| "grad_norm": 0.09531107544898987, | |
| "learning_rate": 1.5364158514040328e-07, | |
| "loss": 0.3163, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.8159509202453987, | |
| "grad_norm": 0.11034092307090759, | |
| "learning_rate": 1.4165329979794972e-07, | |
| "loss": 0.349, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.8220858895705523, | |
| "grad_norm": 0.10453791916370392, | |
| "learning_rate": 1.3014520157529244e-07, | |
| "loss": 0.3591, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.8282208588957056, | |
| "grad_norm": 0.11782266944646835, | |
| "learning_rate": 1.1911842790474637e-07, | |
| "loss": 0.4248, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.834355828220859, | |
| "grad_norm": 0.13830620050430298, | |
| "learning_rate": 1.0857406864568488e-07, | |
| "loss": 0.5286, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.8404907975460123, | |
| "grad_norm": 0.08683877438306808, | |
| "learning_rate": 9.851316597681959e-08, | |
| "loss": 0.2167, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.8466257668711656, | |
| "grad_norm": 0.1016291156411171, | |
| "learning_rate": 8.893671429319294e-08, | |
| "loss": 0.3983, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.852760736196319, | |
| "grad_norm": 0.1077284961938858, | |
| "learning_rate": 7.984566010789673e-08, | |
| "loss": 0.3073, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.8588957055214723, | |
| "grad_norm": 0.08103854209184647, | |
| "learning_rate": 7.124090195851807e-08, | |
| "loss": 0.3116, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.8650306748466257, | |
| "grad_norm": 0.11006593704223633, | |
| "learning_rate": 6.31232903183332e-08, | |
| "loss": 0.4318, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.871165644171779, | |
| "grad_norm": 0.09934885799884796, | |
| "learning_rate": 5.549362751224585e-08, | |
| "loss": 0.3649, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.8773006134969326, | |
| "grad_norm": 0.11354987323284149, | |
| "learning_rate": 4.8352667637490694e-08, | |
| "loss": 0.4092, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.883435582822086, | |
| "grad_norm": 0.10538670420646667, | |
| "learning_rate": 4.170111648909736e-08, | |
| "loss": 0.3204, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.8895705521472392, | |
| "grad_norm": 0.11542137712240219, | |
| "learning_rate": 3.553963149013295e-08, | |
| "loss": 0.4189, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.8957055214723928, | |
| "grad_norm": 0.11087686568498611, | |
| "learning_rate": 2.986882162672344e-08, | |
| "loss": 0.4718, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.9018404907975461, | |
| "grad_norm": 0.11411860585212708, | |
| "learning_rate": 2.4689247387862934e-08, | |
| "loss": 0.3454, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.9079754601226995, | |
| "grad_norm": 0.08472780138254166, | |
| "learning_rate": 2.000142071001632e-08, | |
| "loss": 0.2456, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.9141104294478528, | |
| "grad_norm": 0.10252390056848526, | |
| "learning_rate": 1.580580492652084e-08, | |
| "loss": 0.4087, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9202453987730062, | |
| "grad_norm": 0.12601801753044128, | |
| "learning_rate": 1.2102814721791645e-08, | |
| "loss": 0.4071, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.9263803680981595, | |
| "grad_norm": 0.08735030889511108, | |
| "learning_rate": 8.8928160903351e-09, | |
| "loss": 0.3744, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.9325153374233128, | |
| "grad_norm": 0.10152660310268402, | |
| "learning_rate": 6.176126300573848e-09, | |
| "loss": 0.4182, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.9386503067484662, | |
| "grad_norm": 0.11482395976781845, | |
| "learning_rate": 3.953013863490784e-09, | |
| "loss": 0.3801, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.9447852760736195, | |
| "grad_norm": 0.09311243146657944, | |
| "learning_rate": 2.223698506088612e-09, | |
| "loss": 0.2806, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.9509202453987728, | |
| "grad_norm": 0.1334279328584671, | |
| "learning_rate": 9.883511496722176e-10, | |
| "loss": 0.5107, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.9570552147239264, | |
| "grad_norm": 0.09790550917387009, | |
| "learning_rate": 2.470938929571842e-10, | |
| "loss": 0.2982, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.9631901840490797, | |
| "grad_norm": 0.11607832461595535, | |
| "learning_rate": 0.0, | |
| "loss": 0.3971, | |
| "step": 326 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 326, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 82, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8033259439603057e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |