| { | |
| "best_metric": 10.294066429138184, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 3.0047846889952154, | |
| "eval_steps": 50, | |
| "global_step": 157, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.019138755980861243, | |
| "grad_norm": 0.4266333281993866, | |
| "learning_rate": 1e-05, | |
| "loss": 10.3784, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019138755980861243, | |
| "eval_loss": 10.377665519714355, | |
| "eval_runtime": 0.2079, | |
| "eval_samples_per_second": 423.375, | |
| "eval_steps_per_second": 105.844, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03827751196172249, | |
| "grad_norm": 0.4478803873062134, | |
| "learning_rate": 2e-05, | |
| "loss": 10.3778, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05741626794258373, | |
| "grad_norm": 0.41060081124305725, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3799, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07655502392344497, | |
| "grad_norm": 0.3675904870033264, | |
| "learning_rate": 4e-05, | |
| "loss": 10.3776, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09569377990430622, | |
| "grad_norm": 0.3921853303909302, | |
| "learning_rate": 5e-05, | |
| "loss": 10.3772, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11483253588516747, | |
| "grad_norm": 0.4090043306350708, | |
| "learning_rate": 6e-05, | |
| "loss": 10.3755, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1339712918660287, | |
| "grad_norm": 0.39872977137565613, | |
| "learning_rate": 7e-05, | |
| "loss": 10.3738, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15311004784688995, | |
| "grad_norm": 0.38402482867240906, | |
| "learning_rate": 8e-05, | |
| "loss": 10.3735, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1722488038277512, | |
| "grad_norm": 0.3583473861217499, | |
| "learning_rate": 9e-05, | |
| "loss": 10.3738, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.19138755980861244, | |
| "grad_norm": 0.3627314865589142, | |
| "learning_rate": 0.0001, | |
| "loss": 10.3702, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.42793917655944824, | |
| "learning_rate": 9.99885820390154e-05, | |
| "loss": 10.3661, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.22966507177033493, | |
| "grad_norm": 0.39294254779815674, | |
| "learning_rate": 9.995433337085491e-05, | |
| "loss": 10.3655, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.24880382775119617, | |
| "grad_norm": 0.3837736248970032, | |
| "learning_rate": 9.989726963751682e-05, | |
| "loss": 10.3663, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2679425837320574, | |
| "grad_norm": 0.5134721994400024, | |
| "learning_rate": 9.981741690106034e-05, | |
| "loss": 10.359, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.28708133971291866, | |
| "grad_norm": 0.5079248547554016, | |
| "learning_rate": 9.971481163170268e-05, | |
| "loss": 10.3572, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3062200956937799, | |
| "grad_norm": 0.5069385766983032, | |
| "learning_rate": 9.95895006911623e-05, | |
| "loss": 10.3597, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3253588516746411, | |
| "grad_norm": 0.47848770022392273, | |
| "learning_rate": 9.944154131125642e-05, | |
| "loss": 10.355, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3444976076555024, | |
| "grad_norm": 0.5119503140449524, | |
| "learning_rate": 9.927100106776212e-05, | |
| "loss": 10.353, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.46988987922668457, | |
| "learning_rate": 9.907795784955327e-05, | |
| "loss": 10.3484, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3827751196172249, | |
| "grad_norm": 0.5642258524894714, | |
| "learning_rate": 9.88624998230272e-05, | |
| "loss": 10.3472, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4019138755980861, | |
| "grad_norm": 0.47275394201278687, | |
| "learning_rate": 9.862472539183756e-05, | |
| "loss": 10.3452, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.4831475019454956, | |
| "learning_rate": 9.836474315195147e-05, | |
| "loss": 10.342, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.44019138755980863, | |
| "grad_norm": 0.4092586636543274, | |
| "learning_rate": 9.808267184205183e-05, | |
| "loss": 10.3422, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.45933014354066987, | |
| "grad_norm": 0.46105170249938965, | |
| "learning_rate": 9.777864028930705e-05, | |
| "loss": 10.336, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4784688995215311, | |
| "grad_norm": 0.4268898069858551, | |
| "learning_rate": 9.745278735053343e-05, | |
| "loss": 10.333, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.49760765550239233, | |
| "grad_norm": 0.3566943109035492, | |
| "learning_rate": 9.710526184877667e-05, | |
| "loss": 10.335, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5167464114832536, | |
| "grad_norm": 0.40495947003364563, | |
| "learning_rate": 9.673622250534156e-05, | |
| "loss": 10.3247, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5358851674641149, | |
| "grad_norm": 0.38109859824180603, | |
| "learning_rate": 9.63458378673011e-05, | |
| "loss": 10.3239, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5550239234449761, | |
| "grad_norm": 0.36544325947761536, | |
| "learning_rate": 9.593428623051792e-05, | |
| "loss": 10.3227, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5741626794258373, | |
| "grad_norm": 0.36144474148750305, | |
| "learning_rate": 9.550175555821333e-05, | |
| "loss": 10.3243, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5933014354066986, | |
| "grad_norm": 0.3017820715904236, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 10.3222, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6124401913875598, | |
| "grad_norm": 0.25235190987586975, | |
| "learning_rate": 9.457455677726448e-05, | |
| "loss": 10.3197, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.270778626203537, | |
| "learning_rate": 9.408031213740045e-05, | |
| "loss": 10.3177, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6507177033492823, | |
| "grad_norm": 0.22615878283977509, | |
| "learning_rate": 9.356593520616948e-05, | |
| "loss": 10.3199, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6698564593301436, | |
| "grad_norm": 0.18281053006649017, | |
| "learning_rate": 9.303166090900082e-05, | |
| "loss": 10.319, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6889952153110048, | |
| "grad_norm": 0.1660808026790619, | |
| "learning_rate": 9.24777332588177e-05, | |
| "loss": 10.3191, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7081339712918661, | |
| "grad_norm": 0.19208800792694092, | |
| "learning_rate": 9.190440524459203e-05, | |
| "loss": 10.3193, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.1481315642595291, | |
| "learning_rate": 9.131193871579975e-05, | |
| "loss": 10.316, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7464114832535885, | |
| "grad_norm": 0.10044525563716888, | |
| "learning_rate": 9.070060426282925e-05, | |
| "loss": 10.3234, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7655502392344498, | |
| "grad_norm": 0.1636795997619629, | |
| "learning_rate": 9.007068109339784e-05, | |
| "loss": 10.3094, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.784688995215311, | |
| "grad_norm": 0.1244809702038765, | |
| "learning_rate": 8.942245690503239e-05, | |
| "loss": 10.3069, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8038277511961722, | |
| "grad_norm": 0.1734219640493393, | |
| "learning_rate": 8.87562277536726e-05, | |
| "loss": 10.3117, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8229665071770335, | |
| "grad_norm": 0.12269428372383118, | |
| "learning_rate": 8.807229791845673e-05, | |
| "loss": 10.3143, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.1280500292778015, | |
| "learning_rate": 8.737097976275178e-05, | |
| "loss": 10.3103, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.861244019138756, | |
| "grad_norm": 0.11635402590036392, | |
| "learning_rate": 8.665259359149132e-05, | |
| "loss": 10.3126, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8803827751196173, | |
| "grad_norm": 0.14325037598609924, | |
| "learning_rate": 8.591746750488639e-05, | |
| "loss": 10.3113, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8995215311004785, | |
| "grad_norm": 0.08365687727928162, | |
| "learning_rate": 8.516593724857598e-05, | |
| "loss": 10.3121, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.9186602870813397, | |
| "grad_norm": 0.08075929433107376, | |
| "learning_rate": 8.439834606028594e-05, | |
| "loss": 10.3117, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.937799043062201, | |
| "grad_norm": 0.09325200319290161, | |
| "learning_rate": 8.361504451306585e-05, | |
| "loss": 10.3172, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "grad_norm": 0.07876553386449814, | |
| "learning_rate": 8.28163903551759e-05, | |
| "loss": 10.3132, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "eval_loss": 10.310466766357422, | |
| "eval_runtime": 0.2053, | |
| "eval_samples_per_second": 428.635, | |
| "eval_steps_per_second": 107.159, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9760765550239234, | |
| "grad_norm": 0.08856772631406784, | |
| "learning_rate": 8.200274834669675e-05, | |
| "loss": 10.3106, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9952153110047847, | |
| "grad_norm": 0.09694826602935791, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 10.3171, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.014354066985646, | |
| "grad_norm": 0.15517449378967285, | |
| "learning_rate": 8.033199387471277e-05, | |
| "loss": 18.3693, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.0334928229665072, | |
| "grad_norm": 0.0667949989438057, | |
| "learning_rate": 7.9475644475583e-05, | |
| "loss": 10.0332, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.0705876350402832, | |
| "learning_rate": 7.860583300610849e-05, | |
| "loss": 10.3075, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.0717703349282297, | |
| "grad_norm": 0.08071315288543701, | |
| "learning_rate": 7.772295672522615e-05, | |
| "loss": 10.3503, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.060574401170015335, | |
| "learning_rate": 7.682741885881315e-05, | |
| "loss": 10.2436, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.1100478468899522, | |
| "grad_norm": 0.06491580605506897, | |
| "learning_rate": 7.591962841552627e-05, | |
| "loss": 10.377, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.1291866028708135, | |
| "grad_norm": 0.08106613159179688, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.2996, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.1483253588516746, | |
| "grad_norm": 0.08917899429798126, | |
| "learning_rate": 7.406895362348916e-05, | |
| "loss": 10.2626, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.167464114832536, | |
| "grad_norm": 0.10297100991010666, | |
| "learning_rate": 7.312691451204178e-05, | |
| "loss": 10.3349, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.186602870813397, | |
| "grad_norm": 0.07042112201452255, | |
| "learning_rate": 7.217431291229067e-05, | |
| "loss": 10.3964, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.2057416267942584, | |
| "grad_norm": 0.07486451417207718, | |
| "learning_rate": 7.121158389495186e-05, | |
| "loss": 10.1931, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.2248803827751196, | |
| "grad_norm": 0.09859184175729752, | |
| "learning_rate": 7.023916715611969e-05, | |
| "loss": 10.3828, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.244019138755981, | |
| "grad_norm": 0.07667276263237, | |
| "learning_rate": 6.925750681644953e-05, | |
| "loss": 10.7305, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.07506383955478668, | |
| "learning_rate": 6.826705121831976e-05, | |
| "loss": 10.1215, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2822966507177034, | |
| "grad_norm": 0.06321374326944351, | |
| "learning_rate": 6.726825272106538e-05, | |
| "loss": 10.0464, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3014354066985647, | |
| "grad_norm": 0.07606147229671478, | |
| "learning_rate": 6.626156749437736e-05, | |
| "loss": 10.3409, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.3205741626794258, | |
| "grad_norm": 0.08716686069965363, | |
| "learning_rate": 6.524745530996137e-05, | |
| "loss": 10.3083, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.339712918660287, | |
| "grad_norm": 0.07832231372594833, | |
| "learning_rate": 6.422637933155162e-05, | |
| "loss": 10.2791, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3588516746411483, | |
| "grad_norm": 0.09248408675193787, | |
| "learning_rate": 6.319880590337549e-05, | |
| "loss": 10.3614, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.3779904306220097, | |
| "grad_norm": 0.08951514214277267, | |
| "learning_rate": 6.216520433716545e-05, | |
| "loss": 10.2633, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3971291866028708, | |
| "grad_norm": 0.08380934596061707, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 10.2203, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.4162679425837321, | |
| "grad_norm": 0.08385058492422104, | |
| "learning_rate": 6.008180758778167e-05, | |
| "loss": 10.4143, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.4354066985645932, | |
| "grad_norm": 0.07687491178512573, | |
| "learning_rate": 5.903296393031995e-05, | |
| "loss": 10.2941, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.07478059083223343, | |
| "learning_rate": 5.7979994751668964e-05, | |
| "loss": 10.2666, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.10892575234174728, | |
| "learning_rate": 5.69233809622687e-05, | |
| "loss": 10.3199, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.492822966507177, | |
| "grad_norm": 0.09505303204059601, | |
| "learning_rate": 5.58636051371201e-05, | |
| "loss": 10.6274, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.5119617224880382, | |
| "grad_norm": 0.09463223069906235, | |
| "learning_rate": 5.480115129538409e-05, | |
| "loss": 10.2876, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.5311004784688995, | |
| "grad_norm": 0.08493571728467941, | |
| "learning_rate": 5.373650467932122e-05, | |
| "loss": 9.994, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.5502392344497609, | |
| "grad_norm": 0.093894362449646, | |
| "learning_rate": 5.267015153267245e-05, | |
| "loss": 10.3278, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.569377990430622, | |
| "grad_norm": 0.10776396840810776, | |
| "learning_rate": 5.1602578878582776e-05, | |
| "loss": 10.2704, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.588516746411483, | |
| "grad_norm": 0.10382504016160965, | |
| "learning_rate": 5.053427429716867e-05, | |
| "loss": 10.2904, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.6076555023923444, | |
| "grad_norm": 0.1062113493680954, | |
| "learning_rate": 4.9465725702831346e-05, | |
| "loss": 10.3657, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.6267942583732058, | |
| "grad_norm": 0.12500600516796112, | |
| "learning_rate": 4.839742112141724e-05, | |
| "loss": 10.323, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.6459330143540671, | |
| "grad_norm": 0.12947624921798706, | |
| "learning_rate": 4.732984846732755e-05, | |
| "loss": 10.2952, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.6650717703349283, | |
| "grad_norm": 0.10035474598407745, | |
| "learning_rate": 4.626349532067879e-05, | |
| "loss": 10.2569, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.10493069142103195, | |
| "learning_rate": 4.5198848704615914e-05, | |
| "loss": 10.3804, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.7033492822966507, | |
| "grad_norm": 0.09970294684171677, | |
| "learning_rate": 4.4136394862879914e-05, | |
| "loss": 10.2641, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.722488038277512, | |
| "grad_norm": 0.1537674069404602, | |
| "learning_rate": 4.307661903773129e-05, | |
| "loss": 10.3269, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.7416267942583732, | |
| "grad_norm": 0.1353168934583664, | |
| "learning_rate": 4.2020005248331054e-05, | |
| "loss": 10.7181, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.7607655502392343, | |
| "grad_norm": 0.10454723984003067, | |
| "learning_rate": 4.096703606968006e-05, | |
| "loss": 10.1224, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.7799043062200957, | |
| "grad_norm": 0.10816597938537598, | |
| "learning_rate": 3.991819241221835e-05, | |
| "loss": 10.0276, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.799043062200957, | |
| "grad_norm": 0.12994703650474548, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 10.3349, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.1489221602678299, | |
| "learning_rate": 3.783479566283457e-05, | |
| "loss": 10.2663, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.8373205741626795, | |
| "grad_norm": 0.11641079187393188, | |
| "learning_rate": 3.680119409662452e-05, | |
| "loss": 10.2685, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.8564593301435406, | |
| "grad_norm": 0.13673876225948334, | |
| "learning_rate": 3.5773620668448384e-05, | |
| "loss": 10.3558, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.875598086124402, | |
| "grad_norm": 0.1347990185022354, | |
| "learning_rate": 3.4752544690038647e-05, | |
| "loss": 10.2981, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.11974027007818222, | |
| "learning_rate": 3.373843250562265e-05, | |
| "loss": 10.2657, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.9138755980861244, | |
| "grad_norm": 0.11552689224481583, | |
| "learning_rate": 3.273174727893463e-05, | |
| "loss": 10.3134, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.9138755980861244, | |
| "eval_loss": 10.300241470336914, | |
| "eval_runtime": 0.2059, | |
| "eval_samples_per_second": 427.459, | |
| "eval_steps_per_second": 106.865, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.9330143540669855, | |
| "grad_norm": 0.14589062333106995, | |
| "learning_rate": 3.173294878168025e-05, | |
| "loss": 10.4054, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.9521531100478469, | |
| "grad_norm": 0.13475301861763, | |
| "learning_rate": 3.074249318355046e-05, | |
| "loss": 10.1998, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.9712918660287082, | |
| "grad_norm": 0.14838005602359772, | |
| "learning_rate": 2.976083284388031e-05, | |
| "loss": 10.3857, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.9904306220095693, | |
| "grad_norm": 0.15033847093582153, | |
| "learning_rate": 2.8788416105048122e-05, | |
| "loss": 10.591, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.0095693779904304, | |
| "grad_norm": 0.22326327860355377, | |
| "learning_rate": 2.7825687087709328e-05, | |
| "loss": 18.1498, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.028708133971292, | |
| "grad_norm": 0.14351670444011688, | |
| "learning_rate": 2.687308548795825e-05, | |
| "loss": 9.8884, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.047846889952153, | |
| "grad_norm": 0.12684540450572968, | |
| "learning_rate": 2.5931046376510877e-05, | |
| "loss": 10.2986, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.0669856459330145, | |
| "grad_norm": 0.14553603529930115, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 10.2596, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.0861244019138754, | |
| "grad_norm": 0.13355287909507751, | |
| "learning_rate": 2.4080371584473748e-05, | |
| "loss": 10.2515, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.1391756683588028, | |
| "learning_rate": 2.317258114118686e-05, | |
| "loss": 10.3434, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.124401913875598, | |
| "grad_norm": 0.1632007211446762, | |
| "learning_rate": 2.2277043274773857e-05, | |
| "loss": 10.3774, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.1435406698564594, | |
| "grad_norm": 0.11006581783294678, | |
| "learning_rate": 2.139416699389153e-05, | |
| "loss": 10.237, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.1626794258373208, | |
| "grad_norm": 0.11960204690694809, | |
| "learning_rate": 2.0524355524417017e-05, | |
| "loss": 10.2388, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.13326919078826904, | |
| "learning_rate": 1.966800612528723e-05, | |
| "loss": 10.3773, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.200956937799043, | |
| "grad_norm": 0.12038934975862503, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 10.2409, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.2200956937799043, | |
| "grad_norm": 0.15435650944709778, | |
| "learning_rate": 1.7997251653303248e-05, | |
| "loss": 10.3346, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.2392344497607657, | |
| "grad_norm": 0.14944836497306824, | |
| "learning_rate": 1.7183609644824096e-05, | |
| "loss": 10.6496, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.258373205741627, | |
| "grad_norm": 0.12455789744853973, | |
| "learning_rate": 1.6384955486934156e-05, | |
| "loss": 10.3332, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.277511961722488, | |
| "grad_norm": 0.11132438480854034, | |
| "learning_rate": 1.5601653939714074e-05, | |
| "loss": 9.8656, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.2966507177033493, | |
| "grad_norm": 0.13063865900039673, | |
| "learning_rate": 1.4834062751424015e-05, | |
| "loss": 10.3357, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.11636406928300858, | |
| "learning_rate": 1.4082532495113626e-05, | |
| "loss": 10.2896, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.334928229665072, | |
| "grad_norm": 0.1426575481891632, | |
| "learning_rate": 1.3347406408508695e-05, | |
| "loss": 10.2307, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.354066985645933, | |
| "grad_norm": 0.14399978518486023, | |
| "learning_rate": 1.262902023724824e-05, | |
| "loss": 10.3345, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.373205741626794, | |
| "grad_norm": 0.1384967416524887, | |
| "learning_rate": 1.1927702081543279e-05, | |
| "loss": 10.3376, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.3923444976076556, | |
| "grad_norm": 0.10496751964092255, | |
| "learning_rate": 1.1243772246327416e-05, | |
| "loss": 10.221, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.411483253588517, | |
| "grad_norm": 0.13061700761318207, | |
| "learning_rate": 1.0577543094967612e-05, | |
| "loss": 10.3965, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.430622009569378, | |
| "grad_norm": 0.12669506669044495, | |
| "learning_rate": 9.929318906602175e-06, | |
| "loss": 10.2988, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.449760765550239, | |
| "grad_norm": 0.10814495384693146, | |
| "learning_rate": 9.299395737170757e-06, | |
| "loss": 10.2391, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.4688995215311005, | |
| "grad_norm": 0.12308809906244278, | |
| "learning_rate": 8.688061284200266e-06, | |
| "loss": 10.328, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.488038277511962, | |
| "grad_norm": 0.14173907041549683, | |
| "learning_rate": 8.09559475540797e-06, | |
| "loss": 10.8135, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.507177033492823, | |
| "grad_norm": 0.12133615463972092, | |
| "learning_rate": 7.522266741182305e-06, | |
| "loss": 10.2866, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.10576584190130234, | |
| "learning_rate": 6.968339090999187e-06, | |
| "loss": 9.7405, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.10492077469825745, | |
| "learning_rate": 6.43406479383053e-06, | |
| "loss": 10.3311, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.5645933014354068, | |
| "grad_norm": 0.13342593610286713, | |
| "learning_rate": 5.919687862599549e-06, | |
| "loss": 10.2719, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.583732057416268, | |
| "grad_norm": 0.12062268704175949, | |
| "learning_rate": 5.425443222735527e-06, | |
| "loss": 10.2956, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.6028708133971294, | |
| "grad_norm": 0.11691385507583618, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 10.4012, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.6220095693779903, | |
| "grad_norm": 0.1241801381111145, | |
| "learning_rate": 4.498244441786675e-06, | |
| "loss": 10.2174, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.6411483253588517, | |
| "grad_norm": 0.09667439758777618, | |
| "learning_rate": 4.065713769482082e-06, | |
| "loss": 10.2467, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.660287081339713, | |
| "grad_norm": 0.09752865880727768, | |
| "learning_rate": 3.654162132698918e-06, | |
| "loss": 10.3165, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.679425837320574, | |
| "grad_norm": 0.1245708093047142, | |
| "learning_rate": 3.2637774946584486e-06, | |
| "loss": 10.3543, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.6985645933014353, | |
| "grad_norm": 0.12063057720661163, | |
| "learning_rate": 2.894738151223331e-06, | |
| "loss": 10.2164, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.7177033492822966, | |
| "grad_norm": 0.14080384373664856, | |
| "learning_rate": 2.547212649466568e-06, | |
| "loss": 10.3849, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.13528694212436676, | |
| "learning_rate": 2.221359710692961e-06, | |
| "loss": 10.7423, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.7559808612440193, | |
| "grad_norm": 0.10822786390781403, | |
| "learning_rate": 1.9173281579481892e-06, | |
| "loss": 10.0585, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.77511961722488, | |
| "grad_norm": 0.10477904230356216, | |
| "learning_rate": 1.6352568480485276e-06, | |
| "loss": 10.0791, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.7942583732057416, | |
| "grad_norm": 0.1134805977344513, | |
| "learning_rate": 1.3752746081624467e-06, | |
| "loss": 10.2513, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.813397129186603, | |
| "grad_norm": 0.14501194655895233, | |
| "learning_rate": 1.1375001769727999e-06, | |
| "loss": 10.2943, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.8325358851674642, | |
| "grad_norm": 0.11421407014131546, | |
| "learning_rate": 9.220421504467281e-07, | |
| "loss": 10.2723, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.8516746411483256, | |
| "grad_norm": 0.12981349229812622, | |
| "learning_rate": 7.289989322378732e-07, | |
| "loss": 10.4036, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.8708133971291865, | |
| "grad_norm": 0.1481410712003708, | |
| "learning_rate": 5.584586887435739e-07, | |
| "loss": 10.2066, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.8708133971291865, | |
| "eval_loss": 10.294066429138184, | |
| "eval_runtime": 0.2055, | |
| "eval_samples_per_second": 428.249, | |
| "eval_steps_per_second": 107.062, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.889952153110048, | |
| "grad_norm": 0.11171100288629532, | |
| "learning_rate": 4.104993088376974e-07, | |
| "loss": 10.2872, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.0931149572134018, | |
| "learning_rate": 2.851883682973233e-07, | |
| "loss": 10.2623, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.92822966507177, | |
| "grad_norm": 0.09574086219072342, | |
| "learning_rate": 1.8258309893965375e-07, | |
| "loss": 10.4194, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.1195647194981575, | |
| "learning_rate": 1.0273036248318324e-07, | |
| "loss": 10.1691, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.9665071770334928, | |
| "grad_norm": 0.1469721496105194, | |
| "learning_rate": 4.566662914508579e-08, | |
| "loss": 10.379, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.985645933014354, | |
| "grad_norm": 0.1375855803489685, | |
| "learning_rate": 1.1417960984605458e-08, | |
| "loss": 10.6679, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.0047846889952154, | |
| "grad_norm": 0.17474159598350525, | |
| "learning_rate": 0.0, | |
| "loss": 18.0126, | |
| "step": 157 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 157, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 17513453715456.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |