| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9969356486210419, | |
| "eval_steps": 82, | |
| "global_step": 244, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0040858018386108275, | |
| "grad_norm": 4.75867223739624, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 1.3989, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0040858018386108275, | |
| "eval_loss": 1.7111468315124512, | |
| "eval_runtime": 5.4436, | |
| "eval_samples_per_second": 14.512, | |
| "eval_steps_per_second": 1.837, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008171603677221655, | |
| "grad_norm": 4.975377559661865, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.4837, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012257405515832482, | |
| "grad_norm": 5.219729900360107, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.5181, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01634320735444331, | |
| "grad_norm": 4.57335901260376, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 1.4106, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020429009193054137, | |
| "grad_norm": 3.840559720993042, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.3763, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024514811031664963, | |
| "grad_norm": 3.2056212425231934, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.1876, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.028600612870275793, | |
| "grad_norm": 2.6987595558166504, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 1.2154, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03268641470888662, | |
| "grad_norm": 2.378502130508423, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 1.1594, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03677221654749745, | |
| "grad_norm": 1.7688865661621094, | |
| "learning_rate": 6e-06, | |
| "loss": 0.8435, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04085801838610827, | |
| "grad_norm": 1.3263744115829468, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.7219, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0449438202247191, | |
| "grad_norm": 1.3509997129440308, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.8172, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.049029622063329927, | |
| "grad_norm": 1.4541417360305786, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.7393, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05311542390194075, | |
| "grad_norm": 1.181699275970459, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.664, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05720122574055159, | |
| "grad_norm": 0.9503294825553894, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.6222, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06128702757916241, | |
| "grad_norm": 0.7614471316337585, | |
| "learning_rate": 1e-05, | |
| "loss": 0.56, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06537282941777324, | |
| "grad_norm": 0.9878801107406616, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 0.5548, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06945863125638406, | |
| "grad_norm": 0.8131901025772095, | |
| "learning_rate": 1.1333333333333334e-05, | |
| "loss": 0.4878, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0735444330949949, | |
| "grad_norm": 0.7322743535041809, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.5159, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07763023493360573, | |
| "grad_norm": 0.6428759098052979, | |
| "learning_rate": 1.2666666666666667e-05, | |
| "loss": 0.4575, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08171603677221655, | |
| "grad_norm": 0.562318742275238, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.4571, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08580183861082738, | |
| "grad_norm": 0.5707699060440063, | |
| "learning_rate": 1.4e-05, | |
| "loss": 0.4592, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0898876404494382, | |
| "grad_norm": 0.5272228717803955, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 0.4457, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09397344228804903, | |
| "grad_norm": 0.5120903253555298, | |
| "learning_rate": 1.5333333333333334e-05, | |
| "loss": 0.4034, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09805924412665985, | |
| "grad_norm": 0.46359285712242126, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.4037, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10214504596527069, | |
| "grad_norm": 0.49431198835372925, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.3875, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1062308478038815, | |
| "grad_norm": 0.4450273811817169, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 0.3797, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.11031664964249234, | |
| "grad_norm": 0.4551868140697479, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.3512, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11440245148110317, | |
| "grad_norm": 0.5083736777305603, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 0.3906, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.118488253319714, | |
| "grad_norm": 0.47295963764190674, | |
| "learning_rate": 1.9333333333333333e-05, | |
| "loss": 0.3554, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.12257405515832483, | |
| "grad_norm": 0.4848616123199463, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3712, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12665985699693566, | |
| "grad_norm": 0.4398118555545807, | |
| "learning_rate": 1.999989986294826e-05, | |
| "loss": 0.3694, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.13074565883554648, | |
| "grad_norm": 0.41183602809906006, | |
| "learning_rate": 1.9999599453798523e-05, | |
| "loss": 0.3336, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1348314606741573, | |
| "grad_norm": 0.492713987827301, | |
| "learning_rate": 1.999909877856721e-05, | |
| "loss": 0.3657, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13891726251276812, | |
| "grad_norm": 0.4517015516757965, | |
| "learning_rate": 1.9998397847281548e-05, | |
| "loss": 0.367, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.14300306435137897, | |
| "grad_norm": 0.4641965627670288, | |
| "learning_rate": 1.9997496673979375e-05, | |
| "loss": 0.3565, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1470888661899898, | |
| "grad_norm": 0.4812065064907074, | |
| "learning_rate": 1.9996395276708856e-05, | |
| "loss": 0.3773, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1511746680286006, | |
| "grad_norm": 0.42300987243652344, | |
| "learning_rate": 1.999509367752813e-05, | |
| "loss": 0.3643, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15526046986721145, | |
| "grad_norm": 0.4512963593006134, | |
| "learning_rate": 1.9993591902504854e-05, | |
| "loss": 0.3409, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15934627170582227, | |
| "grad_norm": 0.41626426577568054, | |
| "learning_rate": 1.9991889981715696e-05, | |
| "loss": 0.3546, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1634320735444331, | |
| "grad_norm": 0.43549367785453796, | |
| "learning_rate": 1.9989987949245725e-05, | |
| "loss": 0.3091, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1675178753830439, | |
| "grad_norm": 0.4042600393295288, | |
| "learning_rate": 1.9987885843187717e-05, | |
| "loss": 0.3174, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.17160367722165476, | |
| "grad_norm": 0.4394363462924957, | |
| "learning_rate": 1.9985583705641418e-05, | |
| "loss": 0.3601, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17568947906026558, | |
| "grad_norm": 0.4294170141220093, | |
| "learning_rate": 1.9983081582712684e-05, | |
| "loss": 0.3283, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1797752808988764, | |
| "grad_norm": 0.44452300667762756, | |
| "learning_rate": 1.998037952451255e-05, | |
| "loss": 0.3367, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18386108273748722, | |
| "grad_norm": 0.4113090932369232, | |
| "learning_rate": 1.9977477585156252e-05, | |
| "loss": 0.2986, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18794688457609807, | |
| "grad_norm": 0.44443050026893616, | |
| "learning_rate": 1.9974375822762117e-05, | |
| "loss": 0.3463, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1920326864147089, | |
| "grad_norm": 0.4303809106349945, | |
| "learning_rate": 1.9971074299450414e-05, | |
| "loss": 0.3281, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1961184882533197, | |
| "grad_norm": 0.4178621470928192, | |
| "learning_rate": 1.9967573081342103e-05, | |
| "loss": 0.3629, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.20020429009193055, | |
| "grad_norm": 0.38657113909721375, | |
| "learning_rate": 1.9963872238557516e-05, | |
| "loss": 0.3225, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20429009193054137, | |
| "grad_norm": 0.5300270915031433, | |
| "learning_rate": 1.9959971845214953e-05, | |
| "loss": 0.3279, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2083758937691522, | |
| "grad_norm": 0.4061177968978882, | |
| "learning_rate": 1.9955871979429188e-05, | |
| "loss": 0.3278, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.212461695607763, | |
| "grad_norm": 0.41504785418510437, | |
| "learning_rate": 1.9951572723309918e-05, | |
| "loss": 0.3096, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.21654749744637386, | |
| "grad_norm": 0.4208971858024597, | |
| "learning_rate": 1.9947074162960113e-05, | |
| "loss": 0.3187, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.22063329928498468, | |
| "grad_norm": 0.36819201707839966, | |
| "learning_rate": 1.9942376388474282e-05, | |
| "loss": 0.3167, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2247191011235955, | |
| "grad_norm": 0.43327596783638, | |
| "learning_rate": 1.993747949393668e-05, | |
| "loss": 0.3188, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22880490296220635, | |
| "grad_norm": 0.4377865791320801, | |
| "learning_rate": 1.9932383577419432e-05, | |
| "loss": 0.3478, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23289070480081717, | |
| "grad_norm": 0.43336397409439087, | |
| "learning_rate": 1.992708874098054e-05, | |
| "loss": 0.3025, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.236976506639428, | |
| "grad_norm": 0.4399135410785675, | |
| "learning_rate": 1.9921595090661872e-05, | |
| "loss": 0.3098, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2410623084780388, | |
| "grad_norm": 0.4253901243209839, | |
| "learning_rate": 1.991590273648702e-05, | |
| "loss": 0.3303, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.24514811031664965, | |
| "grad_norm": 0.39254307746887207, | |
| "learning_rate": 1.9910011792459086e-05, | |
| "loss": 0.3018, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24923391215526047, | |
| "grad_norm": 0.4217659831047058, | |
| "learning_rate": 1.9903922376558432e-05, | |
| "loss": 0.285, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2533197139938713, | |
| "grad_norm": 0.48558109998703003, | |
| "learning_rate": 1.989763461074029e-05, | |
| "loss": 0.3221, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2574055158324821, | |
| "grad_norm": 0.47454214096069336, | |
| "learning_rate": 1.989114862093232e-05, | |
| "loss": 0.3056, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.26149131767109296, | |
| "grad_norm": 0.4013993442058563, | |
| "learning_rate": 1.9884464537032103e-05, | |
| "loss": 0.3376, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.26557711950970375, | |
| "grad_norm": 0.4264606237411499, | |
| "learning_rate": 1.9877582492904533e-05, | |
| "loss": 0.3158, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2696629213483146, | |
| "grad_norm": 0.5440453886985779, | |
| "learning_rate": 1.9870502626379127e-05, | |
| "loss": 0.3056, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27374872318692545, | |
| "grad_norm": 0.40003377199172974, | |
| "learning_rate": 1.9863225079247286e-05, | |
| "loss": 0.3357, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27783452502553624, | |
| "grad_norm": 0.39155763387680054, | |
| "learning_rate": 1.985574999725943e-05, | |
| "loss": 0.2819, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2819203268641471, | |
| "grad_norm": 0.4461009204387665, | |
| "learning_rate": 1.9848077530122083e-05, | |
| "loss": 0.2732, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.28600612870275793, | |
| "grad_norm": 0.38970062136650085, | |
| "learning_rate": 1.9840207831494903e-05, | |
| "loss": 0.2957, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2900919305413687, | |
| "grad_norm": 0.4369664788246155, | |
| "learning_rate": 1.983214105898757e-05, | |
| "loss": 0.3158, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2941777323799796, | |
| "grad_norm": 0.4734659492969513, | |
| "learning_rate": 1.9823877374156647e-05, | |
| "loss": 0.3054, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2982635342185904, | |
| "grad_norm": 0.3933468461036682, | |
| "learning_rate": 1.9815416942502346e-05, | |
| "loss": 0.286, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3023493360572012, | |
| "grad_norm": 0.4472273290157318, | |
| "learning_rate": 1.98067599334652e-05, | |
| "loss": 0.3149, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30643513789581206, | |
| "grad_norm": 0.43143752217292786, | |
| "learning_rate": 1.979790652042268e-05, | |
| "loss": 0.2792, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3105209397344229, | |
| "grad_norm": 0.4325246512889862, | |
| "learning_rate": 1.978885688068572e-05, | |
| "loss": 0.3024, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3146067415730337, | |
| "grad_norm": 0.48796600103378296, | |
| "learning_rate": 1.9779611195495177e-05, | |
| "loss": 0.3343, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.31869254341164455, | |
| "grad_norm": 0.40505748987197876, | |
| "learning_rate": 1.977016965001817e-05, | |
| "loss": 0.2753, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.32277834525025534, | |
| "grad_norm": 0.40753036737442017, | |
| "learning_rate": 1.976053243334442e-05, | |
| "loss": 0.3073, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3268641470888662, | |
| "grad_norm": 0.4000149071216583, | |
| "learning_rate": 1.9750699738482403e-05, | |
| "loss": 0.284, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33094994892747703, | |
| "grad_norm": 0.42099907994270325, | |
| "learning_rate": 1.9740671762355548e-05, | |
| "loss": 0.2881, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3350357507660878, | |
| "grad_norm": 0.4155902564525604, | |
| "learning_rate": 1.973044870579824e-05, | |
| "loss": 0.2969, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3350357507660878, | |
| "eval_loss": 0.31923907995224, | |
| "eval_runtime": 5.81, | |
| "eval_samples_per_second": 13.597, | |
| "eval_steps_per_second": 1.721, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3391215526046987, | |
| "grad_norm": 0.39282551407814026, | |
| "learning_rate": 1.972003077355183e-05, | |
| "loss": 0.2948, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3432073544433095, | |
| "grad_norm": 0.4381943643093109, | |
| "learning_rate": 1.9709418174260523e-05, | |
| "loss": 0.3454, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3472931562819203, | |
| "grad_norm": 0.4093382954597473, | |
| "learning_rate": 1.9698611120467196e-05, | |
| "loss": 0.2962, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.35137895812053116, | |
| "grad_norm": 0.450135737657547, | |
| "learning_rate": 1.9687609828609156e-05, | |
| "loss": 0.3243, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.355464759959142, | |
| "grad_norm": 0.4139018654823303, | |
| "learning_rate": 1.9676414519013782e-05, | |
| "loss": 0.2996, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3595505617977528, | |
| "grad_norm": 0.40026575326919556, | |
| "learning_rate": 1.966502541589414e-05, | |
| "loss": 0.2788, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.36627820134162903, | |
| "learning_rate": 1.965344274734447e-05, | |
| "loss": 0.2857, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.36772216547497444, | |
| "grad_norm": 0.42685478925704956, | |
| "learning_rate": 1.9641666745335626e-05, | |
| "loss": 0.2995, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3718079673135853, | |
| "grad_norm": 0.374288946390152, | |
| "learning_rate": 1.9629697645710432e-05, | |
| "loss": 0.3056, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.37589376915219613, | |
| "grad_norm": 0.3649786114692688, | |
| "learning_rate": 1.961753568817896e-05, | |
| "loss": 0.2854, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3799795709908069, | |
| "grad_norm": 0.38573023676872253, | |
| "learning_rate": 1.9605181116313725e-05, | |
| "loss": 0.2667, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3840653728294178, | |
| "grad_norm": 0.37577807903289795, | |
| "learning_rate": 1.9592634177544803e-05, | |
| "loss": 0.2815, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3881511746680286, | |
| "grad_norm": 0.4320047199726105, | |
| "learning_rate": 1.957989512315489e-05, | |
| "loss": 0.3094, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3922369765066394, | |
| "grad_norm": 0.3816889524459839, | |
| "learning_rate": 1.9566964208274254e-05, | |
| "loss": 0.292, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39632277834525026, | |
| "grad_norm": 0.3946669399738312, | |
| "learning_rate": 1.9553841691875632e-05, | |
| "loss": 0.3002, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.4004085801838611, | |
| "grad_norm": 0.36885613203048706, | |
| "learning_rate": 1.9540527836769047e-05, | |
| "loss": 0.2583, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4044943820224719, | |
| "grad_norm": 0.37865176796913147, | |
| "learning_rate": 1.9527022909596537e-05, | |
| "loss": 0.2787, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.40858018386108275, | |
| "grad_norm": 0.4429585337638855, | |
| "learning_rate": 1.951332718082682e-05, | |
| "loss": 0.3226, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.41266598569969354, | |
| "grad_norm": 0.3926009237766266, | |
| "learning_rate": 1.9499440924749878e-05, | |
| "loss": 0.2914, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4167517875383044, | |
| "grad_norm": 0.3467339277267456, | |
| "learning_rate": 1.9485364419471454e-05, | |
| "loss": 0.266, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.42083758937691523, | |
| "grad_norm": 0.4126642644405365, | |
| "learning_rate": 1.9471097946907506e-05, | |
| "loss": 0.2775, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.424923391215526, | |
| "grad_norm": 0.44586020708084106, | |
| "learning_rate": 1.9456641792778527e-05, | |
| "loss": 0.2884, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4290091930541369, | |
| "grad_norm": 0.3969588279724121, | |
| "learning_rate": 1.9441996246603848e-05, | |
| "loss": 0.2835, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4330949948927477, | |
| "grad_norm": 0.38928356766700745, | |
| "learning_rate": 1.9427161601695833e-05, | |
| "loss": 0.2826, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4371807967313585, | |
| "grad_norm": 0.4089799225330353, | |
| "learning_rate": 1.9412138155154e-05, | |
| "loss": 0.2817, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.44126659856996936, | |
| "grad_norm": 0.375505656003952, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.2882, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4453524004085802, | |
| "grad_norm": 0.406118780374527, | |
| "learning_rate": 1.9381526064466995e-05, | |
| "loss": 0.2861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.449438202247191, | |
| "grad_norm": 0.3882409334182739, | |
| "learning_rate": 1.9365938033402715e-05, | |
| "loss": 0.261, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.45352400408580185, | |
| "grad_norm": 0.4351583421230316, | |
| "learning_rate": 1.9350162426854152e-05, | |
| "loss": 0.3014, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4576098059244127, | |
| "grad_norm": 0.3621097505092621, | |
| "learning_rate": 1.933419956076584e-05, | |
| "loss": 0.2728, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4616956077630235, | |
| "grad_norm": 0.3881032466888428, | |
| "learning_rate": 1.9318049754832656e-05, | |
| "loss": 0.2736, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46578140960163433, | |
| "grad_norm": 0.37627285718917847, | |
| "learning_rate": 1.9301713332493386e-05, | |
| "loss": 0.2707, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4698672114402451, | |
| "grad_norm": 0.4285913109779358, | |
| "learning_rate": 1.9285190620924267e-05, | |
| "loss": 0.2815, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.473953013278856, | |
| "grad_norm": 0.35718926787376404, | |
| "learning_rate": 1.926848195103242e-05, | |
| "loss": 0.2621, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4780388151174668, | |
| "grad_norm": 0.3852044641971588, | |
| "learning_rate": 1.925158765744924e-05, | |
| "loss": 0.283, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4821246169560776, | |
| "grad_norm": 0.3884032368659973, | |
| "learning_rate": 1.923450807852367e-05, | |
| "loss": 0.2711, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.48621041879468846, | |
| "grad_norm": 0.4398249685764313, | |
| "learning_rate": 1.9217243556315445e-05, | |
| "loss": 0.2757, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.4902962206332993, | |
| "grad_norm": 0.36689624190330505, | |
| "learning_rate": 1.9199794436588244e-05, | |
| "loss": 0.2669, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4943820224719101, | |
| "grad_norm": 0.46398666501045227, | |
| "learning_rate": 1.9182161068802742e-05, | |
| "loss": 0.2683, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.49846782431052095, | |
| "grad_norm": 0.40020987391471863, | |
| "learning_rate": 1.916434380610963e-05, | |
| "loss": 0.2927, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5025536261491318, | |
| "grad_norm": 0.4032459259033203, | |
| "learning_rate": 1.9146343005342546e-05, | |
| "loss": 0.31, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5066394279877426, | |
| "grad_norm": 0.44166550040245056, | |
| "learning_rate": 1.912815902701091e-05, | |
| "loss": 0.2842, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5107252298263534, | |
| "grad_norm": 0.39895153045654297, | |
| "learning_rate": 1.9109792235292715e-05, | |
| "loss": 0.2766, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5148110316649642, | |
| "grad_norm": 0.3415013253688812, | |
| "learning_rate": 1.909124299802724e-05, | |
| "loss": 0.2761, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5188968335035751, | |
| "grad_norm": 0.3837663531303406, | |
| "learning_rate": 1.9072511686707663e-05, | |
| "loss": 0.2797, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5229826353421859, | |
| "grad_norm": 0.4030819833278656, | |
| "learning_rate": 1.9053598676473656e-05, | |
| "loss": 0.2932, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5270684371807968, | |
| "grad_norm": 0.40120938420295715, | |
| "learning_rate": 1.9034504346103825e-05, | |
| "loss": 0.2698, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5311542390194075, | |
| "grad_norm": 0.3621327579021454, | |
| "learning_rate": 1.9015229078008163e-05, | |
| "loss": 0.298, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5352400408580184, | |
| "grad_norm": 0.33476150035858154, | |
| "learning_rate": 1.8995773258220374e-05, | |
| "loss": 0.2612, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5393258426966292, | |
| "grad_norm": 0.3523140549659729, | |
| "learning_rate": 1.8976137276390145e-05, | |
| "loss": 0.2671, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.54341164453524, | |
| "grad_norm": 0.3624558746814728, | |
| "learning_rate": 1.8956321525775337e-05, | |
| "loss": 0.2687, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5474974463738509, | |
| "grad_norm": 0.35892072319984436, | |
| "learning_rate": 1.8936326403234125e-05, | |
| "loss": 0.2755, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5515832482124617, | |
| "grad_norm": 0.3678256869316101, | |
| "learning_rate": 1.891615230921703e-05, | |
| "loss": 0.278, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5556690500510725, | |
| "grad_norm": 0.38125160336494446, | |
| "learning_rate": 1.8895799647758912e-05, | |
| "loss": 0.2765, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5597548518896833, | |
| "grad_norm": 0.40152257680892944, | |
| "learning_rate": 1.8875268826470875e-05, | |
| "loss": 0.3239, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5638406537282942, | |
| "grad_norm": 0.3935178816318512, | |
| "learning_rate": 1.8854560256532098e-05, | |
| "loss": 0.2956, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.567926455566905, | |
| "grad_norm": 0.4389478266239166, | |
| "learning_rate": 1.8833674352681613e-05, | |
| "loss": 0.2968, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5720122574055159, | |
| "grad_norm": 0.3884355127811432, | |
| "learning_rate": 1.881261153320999e-05, | |
| "loss": 0.3074, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5760980592441267, | |
| "grad_norm": 0.4054373502731323, | |
| "learning_rate": 1.879137221995095e-05, | |
| "loss": 0.2996, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5801838610827375, | |
| "grad_norm": 0.4423893690109253, | |
| "learning_rate": 1.8769956838272937e-05, | |
| "loss": 0.3082, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5842696629213483, | |
| "grad_norm": 0.42978307604789734, | |
| "learning_rate": 1.8748365817070586e-05, | |
| "loss": 0.2878, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5883554647599591, | |
| "grad_norm": 0.38182228803634644, | |
| "learning_rate": 1.8726599588756144e-05, | |
| "loss": 0.2649, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.59244126659857, | |
| "grad_norm": 0.43477413058280945, | |
| "learning_rate": 1.8704658589250795e-05, | |
| "loss": 0.271, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5965270684371808, | |
| "grad_norm": 0.3876926898956299, | |
| "learning_rate": 1.868254325797594e-05, | |
| "loss": 0.2804, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6006128702757916, | |
| "grad_norm": 0.39310601353645325, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.2767, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6046986721144024, | |
| "grad_norm": 0.421290785074234, | |
| "learning_rate": 1.8637791375251505e-05, | |
| "loss": 0.2668, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6087844739530133, | |
| "grad_norm": 0.450023353099823, | |
| "learning_rate": 1.8615155720066247e-05, | |
| "loss": 0.2888, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6128702757916241, | |
| "grad_norm": 0.3645341396331787, | |
| "learning_rate": 1.859234752562217e-05, | |
| "loss": 0.2828, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.616956077630235, | |
| "grad_norm": 0.41853606700897217, | |
| "learning_rate": 1.8569367248708343e-05, | |
| "loss": 0.284, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6210418794688458, | |
| "grad_norm": 0.3675737679004669, | |
| "learning_rate": 1.8546215349560204e-05, | |
| "loss": 0.2933, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6251276813074566, | |
| "grad_norm": 0.3668256998062134, | |
| "learning_rate": 1.8522892291850335e-05, | |
| "loss": 0.2729, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6292134831460674, | |
| "grad_norm": 0.34576019644737244, | |
| "learning_rate": 1.849939854267919e-05, | |
| "loss": 0.2612, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6332992849846782, | |
| "grad_norm": 0.41370126605033875, | |
| "learning_rate": 1.847573457256571e-05, | |
| "loss": 0.2693, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6373850868232891, | |
| "grad_norm": 0.4205566644668579, | |
| "learning_rate": 1.845190085543795e-05, | |
| "loss": 0.2746, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6414708886618999, | |
| "grad_norm": 0.3997614085674286, | |
| "learning_rate": 1.8427897868623535e-05, | |
| "loss": 0.2813, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6455566905005107, | |
| "grad_norm": 0.41005200147628784, | |
| "learning_rate": 1.840372609284013e-05, | |
| "loss": 0.2647, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6496424923391215, | |
| "grad_norm": 0.4547550678253174, | |
| "learning_rate": 1.8379386012185813e-05, | |
| "loss": 0.2791, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6537282941777324, | |
| "grad_norm": 0.4075047969818115, | |
| "learning_rate": 1.8354878114129368e-05, | |
| "loss": 0.2769, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6578140960163432, | |
| "grad_norm": 0.37060046195983887, | |
| "learning_rate": 1.8330202889500518e-05, | |
| "loss": 0.3028, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6618998978549541, | |
| "grad_norm": 0.35541340708732605, | |
| "learning_rate": 1.8305360832480118e-05, | |
| "loss": 0.2981, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6659856996935649, | |
| "grad_norm": 0.3970625400543213, | |
| "learning_rate": 1.8280352440590236e-05, | |
| "loss": 0.2634, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6700715015321757, | |
| "grad_norm": 0.4075865149497986, | |
| "learning_rate": 1.82551782146842e-05, | |
| "loss": 0.3027, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6700715015321757, | |
| "eval_loss": 0.291363924741745, | |
| "eval_runtime": 5.7936, | |
| "eval_samples_per_second": 13.636, | |
| "eval_steps_per_second": 1.726, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6741573033707865, | |
| "grad_norm": 0.34390076994895935, | |
| "learning_rate": 1.8229838658936566e-05, | |
| "loss": 0.2536, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6782431052093973, | |
| "grad_norm": 0.3729197084903717, | |
| "learning_rate": 1.8204334280833005e-05, | |
| "loss": 0.2739, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6823289070480082, | |
| "grad_norm": 0.3974601924419403, | |
| "learning_rate": 1.817866559116017e-05, | |
| "loss": 0.2858, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.686414708886619, | |
| "grad_norm": 0.3424644470214844, | |
| "learning_rate": 1.8152833103995443e-05, | |
| "loss": 0.2305, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6905005107252298, | |
| "grad_norm": 0.4293709397315979, | |
| "learning_rate": 1.8126837336696645e-05, | |
| "loss": 0.3179, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6945863125638406, | |
| "grad_norm": 0.3259459435939789, | |
| "learning_rate": 1.8100678809891668e-05, | |
| "loss": 0.2589, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6986721144024515, | |
| "grad_norm": 0.40771302580833435, | |
| "learning_rate": 1.807435804746807e-05, | |
| "loss": 0.2637, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7027579162410623, | |
| "grad_norm": 0.3847212493419647, | |
| "learning_rate": 1.8047875576562556e-05, | |
| "loss": 0.2782, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7068437180796732, | |
| "grad_norm": 0.35547974705696106, | |
| "learning_rate": 1.802123192755044e-05, | |
| "loss": 0.2695, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.710929519918284, | |
| "grad_norm": 0.3954298198223114, | |
| "learning_rate": 1.7994427634035016e-05, | |
| "loss": 0.3005, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7150153217568948, | |
| "grad_norm": 0.3506409525871277, | |
| "learning_rate": 1.796746323283686e-05, | |
| "loss": 0.2716, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7191011235955056, | |
| "grad_norm": 0.42227277159690857, | |
| "learning_rate": 1.7940339263983112e-05, | |
| "loss": 0.2915, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7231869254341164, | |
| "grad_norm": 0.3948259949684143, | |
| "learning_rate": 1.791305627069662e-05, | |
| "loss": 0.2883, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.3580792248249054, | |
| "learning_rate": 1.7885614799385086e-05, | |
| "loss": 0.2782, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7313585291113381, | |
| "grad_norm": 0.39698660373687744, | |
| "learning_rate": 1.785801539963012e-05, | |
| "loss": 0.2657, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7354443309499489, | |
| "grad_norm": 0.3663792610168457, | |
| "learning_rate": 1.7830258624176224e-05, | |
| "loss": 0.2686, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7395301327885597, | |
| "grad_norm": 0.38216930627822876, | |
| "learning_rate": 1.7802345028919728e-05, | |
| "loss": 0.2706, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7436159346271706, | |
| "grad_norm": 0.4187450706958771, | |
| "learning_rate": 1.777427517289766e-05, | |
| "loss": 0.2573, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7477017364657814, | |
| "grad_norm": 0.34619036316871643, | |
| "learning_rate": 1.7746049618276545e-05, | |
| "loss": 0.269, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7517875383043923, | |
| "grad_norm": 0.35370582342147827, | |
| "learning_rate": 1.7717668930341152e-05, | |
| "loss": 0.2552, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7558733401430031, | |
| "grad_norm": 0.4264880418777466, | |
| "learning_rate": 1.768913367748316e-05, | |
| "loss": 0.2952, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7599591419816139, | |
| "grad_norm": 0.39135676622390747, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.2661, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7640449438202247, | |
| "grad_norm": 0.39061596989631653, | |
| "learning_rate": 1.7631601766032337e-05, | |
| "loss": 0.2737, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7681307456588355, | |
| "grad_norm": 0.3799816966056824, | |
| "learning_rate": 1.7602606259654704e-05, | |
| "loss": 0.2767, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7722165474974464, | |
| "grad_norm": 0.3592148721218109, | |
| "learning_rate": 1.7573458492761802e-05, | |
| "loss": 0.2448, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7763023493360572, | |
| "grad_norm": 0.39084604382514954, | |
| "learning_rate": 1.7544159049107902e-05, | |
| "loss": 0.275, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.780388151174668, | |
| "grad_norm": 0.36443451046943665, | |
| "learning_rate": 1.7514708515485002e-05, | |
| "loss": 0.2645, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7844739530132788, | |
| "grad_norm": 0.4001200497150421, | |
| "learning_rate": 1.7485107481711014e-05, | |
| "loss": 0.2724, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7885597548518897, | |
| "grad_norm": 0.39093396067619324, | |
| "learning_rate": 1.7455356540617988e-05, | |
| "loss": 0.2712, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7926455566905005, | |
| "grad_norm": 0.3430577218532562, | |
| "learning_rate": 1.7425456288040236e-05, | |
| "loss": 0.2489, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7967313585291114, | |
| "grad_norm": 0.3573733866214752, | |
| "learning_rate": 1.7395407322802374e-05, | |
| "loss": 0.2696, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.8008171603677222, | |
| "grad_norm": 0.38158077001571655, | |
| "learning_rate": 1.736521024670737e-05, | |
| "loss": 0.2814, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.804902962206333, | |
| "grad_norm": 0.366470068693161, | |
| "learning_rate": 1.733486566452446e-05, | |
| "loss": 0.2529, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8089887640449438, | |
| "grad_norm": 0.3718278408050537, | |
| "learning_rate": 1.7304374183977032e-05, | |
| "loss": 0.2747, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8130745658835546, | |
| "grad_norm": 0.3395809233188629, | |
| "learning_rate": 1.7273736415730488e-05, | |
| "loss": 0.2693, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8171603677221655, | |
| "grad_norm": 0.307731032371521, | |
| "learning_rate": 1.7242952973379983e-05, | |
| "loss": 0.2081, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8212461695607763, | |
| "grad_norm": 0.3522433936595917, | |
| "learning_rate": 1.7212024473438145e-05, | |
| "loss": 0.2495, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8253319713993871, | |
| "grad_norm": 0.35946980118751526, | |
| "learning_rate": 1.7180951535322742e-05, | |
| "loss": 0.2702, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8294177732379979, | |
| "grad_norm": 0.3933047950267792, | |
| "learning_rate": 1.7149734781344247e-05, | |
| "loss": 0.2629, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8335035750766088, | |
| "grad_norm": 0.3658384084701538, | |
| "learning_rate": 1.7118374836693407e-05, | |
| "loss": 0.2538, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8375893769152196, | |
| "grad_norm": 0.3532220423221588, | |
| "learning_rate": 1.7086872329428702e-05, | |
| "loss": 0.2587, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8416751787538305, | |
| "grad_norm": 0.3619686961174011, | |
| "learning_rate": 1.705522789046377e-05, | |
| "loss": 0.2658, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8457609805924413, | |
| "grad_norm": 0.4083801209926605, | |
| "learning_rate": 1.7023442153554776e-05, | |
| "loss": 0.2614, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.849846782431052, | |
| "grad_norm": 0.3868924081325531, | |
| "learning_rate": 1.6991515755287715e-05, | |
| "loss": 0.2831, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8539325842696629, | |
| "grad_norm": 0.38413897156715393, | |
| "learning_rate": 1.695944933506567e-05, | |
| "loss": 0.2596, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8580183861082737, | |
| "grad_norm": 0.34999531507492065, | |
| "learning_rate": 1.6927243535095995e-05, | |
| "loss": 0.2842, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8621041879468846, | |
| "grad_norm": 0.328204482793808, | |
| "learning_rate": 1.6894899000377462e-05, | |
| "loss": 0.2332, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8661899897854954, | |
| "grad_norm": 0.3802552819252014, | |
| "learning_rate": 1.686241637868734e-05, | |
| "loss": 0.2709, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8702757916241062, | |
| "grad_norm": 0.35758858919143677, | |
| "learning_rate": 1.6829796320568416e-05, | |
| "loss": 0.279, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.874361593462717, | |
| "grad_norm": 0.3561984896659851, | |
| "learning_rate": 1.6797039479315994e-05, | |
| "loss": 0.2868, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8784473953013279, | |
| "grad_norm": 0.32591065764427185, | |
| "learning_rate": 1.6764146510964762e-05, | |
| "loss": 0.2485, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8825331971399387, | |
| "grad_norm": 0.36409640312194824, | |
| "learning_rate": 1.67311180742757e-05, | |
| "loss": 0.2577, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8866189989785496, | |
| "grad_norm": 0.34685492515563965, | |
| "learning_rate": 1.669795483072287e-05, | |
| "loss": 0.247, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8907048008171604, | |
| "grad_norm": 0.3445712625980377, | |
| "learning_rate": 1.6664657444480145e-05, | |
| "loss": 0.2565, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8947906026557712, | |
| "grad_norm": 0.34710460901260376, | |
| "learning_rate": 1.6631226582407954e-05, | |
| "loss": 0.2363, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.898876404494382, | |
| "grad_norm": 0.33726766705513, | |
| "learning_rate": 1.6597662914039885e-05, | |
| "loss": 0.2483, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9029622063329928, | |
| "grad_norm": 0.34024032950401306, | |
| "learning_rate": 1.65639671115693e-05, | |
| "loss": 0.2474, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.9070480081716037, | |
| "grad_norm": 0.38807395100593567, | |
| "learning_rate": 1.653013984983585e-05, | |
| "loss": 0.2726, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9111338100102145, | |
| "grad_norm": 0.36375290155410767, | |
| "learning_rate": 1.6496181806312005e-05, | |
| "loss": 0.2726, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9152196118488254, | |
| "grad_norm": 0.36927178502082825, | |
| "learning_rate": 1.6462093661089432e-05, | |
| "loss": 0.2518, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9193054136874361, | |
| "grad_norm": 0.3809269070625305, | |
| "learning_rate": 1.6427876096865394e-05, | |
| "loss": 0.2449, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.923391215526047, | |
| "grad_norm": 0.34634968638420105, | |
| "learning_rate": 1.6393529798929103e-05, | |
| "loss": 0.2575, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9274770173646578, | |
| "grad_norm": 0.33054831624031067, | |
| "learning_rate": 1.635905545514795e-05, | |
| "loss": 0.2639, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9315628192032687, | |
| "grad_norm": 0.35482174158096313, | |
| "learning_rate": 1.6324453755953772e-05, | |
| "loss": 0.2667, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9356486210418795, | |
| "grad_norm": 0.3657509684562683, | |
| "learning_rate": 1.6289725394328998e-05, | |
| "loss": 0.255, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9397344228804902, | |
| "grad_norm": 0.3343275785446167, | |
| "learning_rate": 1.6254871065792776e-05, | |
| "loss": 0.2336, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9438202247191011, | |
| "grad_norm": 0.3493170142173767, | |
| "learning_rate": 1.621989146838704e-05, | |
| "loss": 0.2649, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.947906026557712, | |
| "grad_norm": 0.3305867612361908, | |
| "learning_rate": 1.618478730266255e-05, | |
| "loss": 0.2767, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9519918283963228, | |
| "grad_norm": 0.35817259550094604, | |
| "learning_rate": 1.6149559271664835e-05, | |
| "loss": 0.2817, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9560776302349336, | |
| "grad_norm": 0.37733370065689087, | |
| "learning_rate": 1.6114208080920125e-05, | |
| "loss": 0.2809, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.9601634320735445, | |
| "grad_norm": 0.3227766156196594, | |
| "learning_rate": 1.607873443842122e-05, | |
| "loss": 0.2545, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9642492339121552, | |
| "grad_norm": 0.3445710241794586, | |
| "learning_rate": 1.6043139054613326e-05, | |
| "loss": 0.2476, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9683350357507661, | |
| "grad_norm": 0.3375508785247803, | |
| "learning_rate": 1.600742264237979e-05, | |
| "loss": 0.2502, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9724208375893769, | |
| "grad_norm": 0.356039434671402, | |
| "learning_rate": 1.5971585917027864e-05, | |
| "loss": 0.268, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9765066394279878, | |
| "grad_norm": 0.34852373600006104, | |
| "learning_rate": 1.5935629596274345e-05, | |
| "loss": 0.2605, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.9805924412665986, | |
| "grad_norm": 0.3376101851463318, | |
| "learning_rate": 1.5899554400231233e-05, | |
| "loss": 0.2567, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9846782431052093, | |
| "grad_norm": 0.32361170649528503, | |
| "learning_rate": 1.586336105139127e-05, | |
| "loss": 0.2481, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.9887640449438202, | |
| "grad_norm": 0.35558903217315674, | |
| "learning_rate": 1.5827050274613512e-05, | |
| "loss": 0.2514, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.992849846782431, | |
| "grad_norm": 0.31636619567871094, | |
| "learning_rate": 1.579062279710879e-05, | |
| "loss": 0.2237, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9969356486210419, | |
| "grad_norm": 0.3540779948234558, | |
| "learning_rate": 1.5754079348425137e-05, | |
| "loss": 0.2381, | |
| "step": 244 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 732, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 244, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.127797840707584e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |