| { | |
| "best_metric": 0.23266415805525312, | |
| "best_model_checkpoint": "results/punjabi_model/20240813_012221/checkpoint-5000", | |
| "epoch": 4.815177439288638, | |
| "eval_steps": 5000, | |
| "global_step": 75000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03210118292859092, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.24e-05, | |
| "loss": 2.1834, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06420236585718184, | |
| "grad_norm": 9.310239791870117, | |
| "learning_rate": 2.4900000000000002e-05, | |
| "loss": 0.676, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09630354878577276, | |
| "grad_norm": 8.342071533203125, | |
| "learning_rate": 3.74e-05, | |
| "loss": 0.577, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.12840473171436367, | |
| "grad_norm": 8.718728065490723, | |
| "learning_rate": 4.99e-05, | |
| "loss": 0.5662, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1605059146429546, | |
| "grad_norm": 8.425736427307129, | |
| "learning_rate": 4.98390243902439e-05, | |
| "loss": 0.5139, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.1926070975715455, | |
| "grad_norm": 7.612177848815918, | |
| "learning_rate": 4.967642276422764e-05, | |
| "loss": 0.4737, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.22470828050013644, | |
| "grad_norm": 6.828492164611816, | |
| "learning_rate": 4.951382113821139e-05, | |
| "loss": 0.4696, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.25680946342872735, | |
| "grad_norm": 25.135059356689453, | |
| "learning_rate": 4.9351219512195126e-05, | |
| "loss": 0.4487, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2889106463573183, | |
| "grad_norm": 6.97437047958374, | |
| "learning_rate": 4.9188617886178865e-05, | |
| "loss": 0.4259, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.3210118292859092, | |
| "grad_norm": 4.6233744621276855, | |
| "learning_rate": 4.9026016260162604e-05, | |
| "loss": 0.3788, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.3210118292859092, | |
| "eval_loss": 0.4430150091648102, | |
| "eval_runtime": 597.4312, | |
| "eval_samples_per_second": 11.601, | |
| "eval_steps_per_second": 11.601, | |
| "eval_wer": 0.23266415805525312, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.3531130122145001, | |
| "grad_norm": 4.593179225921631, | |
| "learning_rate": 4.886341463414634e-05, | |
| "loss": 0.4035, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.385214195143091, | |
| "grad_norm": 6.695476055145264, | |
| "learning_rate": 4.870081300813009e-05, | |
| "loss": 0.3882, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.41731537807168195, | |
| "grad_norm": 14.509389877319336, | |
| "learning_rate": 4.853853658536586e-05, | |
| "loss": 0.3574, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.4494165610002729, | |
| "grad_norm": 38.900367736816406, | |
| "learning_rate": 4.8375934959349596e-05, | |
| "loss": 0.3723, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.48151774392886376, | |
| "grad_norm": 3.945736885070801, | |
| "learning_rate": 4.8213333333333335e-05, | |
| "loss": 0.3586, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.5136189268574547, | |
| "grad_norm": 77.0046615600586, | |
| "learning_rate": 4.8050731707317074e-05, | |
| "loss": 0.3679, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5457201097860456, | |
| "grad_norm": 16.720748901367188, | |
| "learning_rate": 4.788813008130081e-05, | |
| "loss": 0.3352, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5778212927146366, | |
| "grad_norm": 17.356231689453125, | |
| "learning_rate": 4.772552845528456e-05, | |
| "loss": 0.341, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6099224756432274, | |
| "grad_norm": 9.849752426147461, | |
| "learning_rate": 4.75629268292683e-05, | |
| "loss": 0.3354, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6420236585718184, | |
| "grad_norm": 9.619921684265137, | |
| "learning_rate": 4.7400650406504066e-05, | |
| "loss": 0.3288, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6420236585718184, | |
| "eval_loss": 0.35230785608291626, | |
| "eval_runtime": 596.403, | |
| "eval_samples_per_second": 11.621, | |
| "eval_steps_per_second": 11.621, | |
| "eval_wer": 0.18700615692192107, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6741248415004093, | |
| "grad_norm": 7.539575099945068, | |
| "learning_rate": 4.723804878048781e-05, | |
| "loss": 0.3314, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.7062260244290002, | |
| "grad_norm": 13.222583770751953, | |
| "learning_rate": 4.7075447154471544e-05, | |
| "loss": 0.3383, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.7383272073575912, | |
| "grad_norm": 2.821357011795044, | |
| "learning_rate": 4.691284552845528e-05, | |
| "loss": 0.3319, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.770428390286182, | |
| "grad_norm": 7.006532669067383, | |
| "learning_rate": 4.675024390243903e-05, | |
| "loss": 0.3218, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8025295732147729, | |
| "grad_norm": 6.498229503631592, | |
| "learning_rate": 4.658764227642277e-05, | |
| "loss": 0.3175, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.8346307561433639, | |
| "grad_norm": 8.40401840209961, | |
| "learning_rate": 4.6425365853658536e-05, | |
| "loss": 0.3083, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.8667319390719548, | |
| "grad_norm": 1.3094803094863892, | |
| "learning_rate": 4.626276422764228e-05, | |
| "loss": 0.3109, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.8988331220005458, | |
| "grad_norm": 3.827925443649292, | |
| "learning_rate": 4.610016260162602e-05, | |
| "loss": 0.3203, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.9309343049291366, | |
| "grad_norm": 2.329725503921509, | |
| "learning_rate": 4.593756097560975e-05, | |
| "loss": 0.2979, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.9630354878577275, | |
| "grad_norm": 3.5614166259765625, | |
| "learning_rate": 4.57749593495935e-05, | |
| "loss": 0.3024, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9630354878577275, | |
| "eval_loss": 0.30602961778640747, | |
| "eval_runtime": 592.7853, | |
| "eval_samples_per_second": 11.692, | |
| "eval_steps_per_second": 11.692, | |
| "eval_wer": 0.16886647692905213, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9951366707863185, | |
| "grad_norm": 87.19066619873047, | |
| "learning_rate": 4.561235772357724e-05, | |
| "loss": 0.3088, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.0272378537149094, | |
| "grad_norm": 2.4488048553466797, | |
| "learning_rate": 4.544975609756098e-05, | |
| "loss": 0.2761, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.0593390366435003, | |
| "grad_norm": 2.8186426162719727, | |
| "learning_rate": 4.5287154471544715e-05, | |
| "loss": 0.2708, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.0914402195720911, | |
| "grad_norm": 1.2367055416107178, | |
| "learning_rate": 4.5124552845528454e-05, | |
| "loss": 0.2814, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.1235414025006822, | |
| "grad_norm": 1.350964903831482, | |
| "learning_rate": 4.496227642276423e-05, | |
| "loss": 0.2943, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.1556425854292731, | |
| "grad_norm": 3.4485456943511963, | |
| "learning_rate": 4.479967479674797e-05, | |
| "loss": 0.2814, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.187743768357864, | |
| "grad_norm": 2.5614452362060547, | |
| "learning_rate": 4.463707317073171e-05, | |
| "loss": 0.2624, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.2198449512864549, | |
| "grad_norm": 5.0949625968933105, | |
| "learning_rate": 4.447447154471545e-05, | |
| "loss": 0.2808, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.2519461342150457, | |
| "grad_norm": 1.0048319101333618, | |
| "learning_rate": 4.431186991869919e-05, | |
| "loss": 0.2534, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.2840473171436368, | |
| "grad_norm": 1.9653562307357788, | |
| "learning_rate": 4.4149268292682924e-05, | |
| "loss": 0.2465, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.2840473171436368, | |
| "eval_loss": 0.28473955392837524, | |
| "eval_runtime": 595.99, | |
| "eval_samples_per_second": 11.629, | |
| "eval_steps_per_second": 11.629, | |
| "eval_wer": 0.14723766227978022, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.3161485000722277, | |
| "grad_norm": 0.9238675832748413, | |
| "learning_rate": 4.3987317073170736e-05, | |
| "loss": 0.2812, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.3482496830008186, | |
| "grad_norm": 1.7164336442947388, | |
| "learning_rate": 4.3824715447154475e-05, | |
| "loss": 0.2716, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.3803508659294095, | |
| "grad_norm": 1.9793908596038818, | |
| "learning_rate": 4.3662113821138214e-05, | |
| "loss": 0.2779, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.4124520488580004, | |
| "grad_norm": 2.866766929626465, | |
| "learning_rate": 4.349951219512195e-05, | |
| "loss": 0.2467, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.4445532317865912, | |
| "grad_norm": 1.6922576427459717, | |
| "learning_rate": 4.333691056910569e-05, | |
| "loss": 0.2616, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.4766544147151823, | |
| "grad_norm": 2.3735485076904297, | |
| "learning_rate": 4.317430894308943e-05, | |
| "loss": 0.273, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.5087555976437732, | |
| "grad_norm": 1.882097601890564, | |
| "learning_rate": 4.3012032520325206e-05, | |
| "loss": 0.2541, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.540856780572364, | |
| "grad_norm": 1.2024834156036377, | |
| "learning_rate": 4.2849430894308945e-05, | |
| "loss": 0.2679, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.572957963500955, | |
| "grad_norm": 1.467214584350586, | |
| "learning_rate": 4.2686829268292684e-05, | |
| "loss": 0.2788, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.605059146429546, | |
| "grad_norm": 0.9247603416442871, | |
| "learning_rate": 4.252422764227643e-05, | |
| "loss": 0.2575, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.605059146429546, | |
| "eval_loss": 0.26871952414512634, | |
| "eval_runtime": 596.0677, | |
| "eval_samples_per_second": 11.628, | |
| "eval_steps_per_second": 11.628, | |
| "eval_wer": 0.13321745331372287, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.637160329358137, | |
| "grad_norm": 1.6064761877059937, | |
| "learning_rate": 4.236162601626016e-05, | |
| "loss": 0.2325, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.6692615122867278, | |
| "grad_norm": 1.4238929748535156, | |
| "learning_rate": 4.21990243902439e-05, | |
| "loss": 0.2463, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.7013626952153187, | |
| "grad_norm": 0.7233666181564331, | |
| "learning_rate": 4.2036422764227646e-05, | |
| "loss": 0.2537, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.7334638781439096, | |
| "grad_norm": 0.9025558233261108, | |
| "learning_rate": 4.1873821138211385e-05, | |
| "loss": 0.2524, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.7655650610725004, | |
| "grad_norm": 2.8646883964538574, | |
| "learning_rate": 4.1711219512195124e-05, | |
| "loss": 0.2536, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.7976662440010913, | |
| "grad_norm": 1.5429887771606445, | |
| "learning_rate": 4.154861788617886e-05, | |
| "loss": 0.2664, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.8297674269296822, | |
| "grad_norm": 2.8908677101135254, | |
| "learning_rate": 4.138634146341464e-05, | |
| "loss": 0.2636, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.8618686098582733, | |
| "grad_norm": 2.929805278778076, | |
| "learning_rate": 4.122373983739838e-05, | |
| "loss": 0.257, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.8939697927868642, | |
| "grad_norm": 1.2041730880737305, | |
| "learning_rate": 4.1061138211382116e-05, | |
| "loss": 0.2459, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.9260709757154553, | |
| "grad_norm": 1.7089420557022095, | |
| "learning_rate": 4.0898536585365855e-05, | |
| "loss": 0.2608, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.9260709757154553, | |
| "eval_loss": 0.2512572705745697, | |
| "eval_runtime": 594.6622, | |
| "eval_samples_per_second": 11.655, | |
| "eval_steps_per_second": 11.655, | |
| "eval_wer": 0.1310080925002706, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.9581721586440461, | |
| "grad_norm": 2.743481397628784, | |
| "learning_rate": 4.073626016260163e-05, | |
| "loss": 0.2381, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.990273341572637, | |
| "grad_norm": 0.6640022993087769, | |
| "learning_rate": 4.05739837398374e-05, | |
| "loss": 0.2399, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.022374524501228, | |
| "grad_norm": 1.1567021608352661, | |
| "learning_rate": 4.041138211382114e-05, | |
| "loss": 0.2238, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 2.0544757074298188, | |
| "grad_norm": 1.3281149864196777, | |
| "learning_rate": 4.0248780487804884e-05, | |
| "loss": 0.2302, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.0865768903584097, | |
| "grad_norm": 3.654982328414917, | |
| "learning_rate": 4.008617886178862e-05, | |
| "loss": 0.2195, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.1186780732870005, | |
| "grad_norm": 1.8765735626220703, | |
| "learning_rate": 3.9923577235772355e-05, | |
| "loss": 0.2375, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.1507792562155914, | |
| "grad_norm": 1.4444390535354614, | |
| "learning_rate": 3.97609756097561e-05, | |
| "loss": 0.2172, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 2.1828804391441823, | |
| "grad_norm": 0.9096426367759705, | |
| "learning_rate": 3.959837398373984e-05, | |
| "loss": 0.2258, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.2149816220727736, | |
| "grad_norm": 1.1495827436447144, | |
| "learning_rate": 3.943577235772358e-05, | |
| "loss": 0.223, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 2.2470828050013645, | |
| "grad_norm": 4.880052089691162, | |
| "learning_rate": 3.9273495934959354e-05, | |
| "loss": 0.2465, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.2470828050013645, | |
| "eval_loss": 0.2504861056804657, | |
| "eval_runtime": 588.4465, | |
| "eval_samples_per_second": 11.778, | |
| "eval_steps_per_second": 11.778, | |
| "eval_wer": 0.12438637709395832, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.2791839879299554, | |
| "grad_norm": 0.8823996782302856, | |
| "learning_rate": 3.911089430894309e-05, | |
| "loss": 0.2123, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 2.3112851708585462, | |
| "grad_norm": 0.8461039066314697, | |
| "learning_rate": 3.894861788617886e-05, | |
| "loss": 0.2255, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.343386353787137, | |
| "grad_norm": 1.3658907413482666, | |
| "learning_rate": 3.878601626016261e-05, | |
| "loss": 0.2166, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 2.375487536715728, | |
| "grad_norm": 5.125720500946045, | |
| "learning_rate": 3.8623414634146346e-05, | |
| "loss": 0.2174, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.407588719644319, | |
| "grad_norm": 0.960725724697113, | |
| "learning_rate": 3.846081300813008e-05, | |
| "loss": 0.2148, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 2.4396899025729097, | |
| "grad_norm": 3.487365484237671, | |
| "learning_rate": 3.829853658536586e-05, | |
| "loss": 0.225, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.4717910855015006, | |
| "grad_norm": 2.1244943141937256, | |
| "learning_rate": 3.813593495934959e-05, | |
| "loss": 0.2237, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 2.5038922684300915, | |
| "grad_norm": 128.5044708251953, | |
| "learning_rate": 3.797333333333333e-05, | |
| "loss": 0.2298, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.5359934513586824, | |
| "grad_norm": 2.0013089179992676, | |
| "learning_rate": 3.781073170731708e-05, | |
| "loss": 0.2184, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 2.5680946342872737, | |
| "grad_norm": 4.2432541847229, | |
| "learning_rate": 3.7648130081300816e-05, | |
| "loss": 0.211, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.5680946342872737, | |
| "eval_loss": 0.2507112920284271, | |
| "eval_runtime": 589.626, | |
| "eval_samples_per_second": 11.755, | |
| "eval_steps_per_second": 11.755, | |
| "eval_wer": 0.12312570435314117, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.6001958172158646, | |
| "grad_norm": 7.432988166809082, | |
| "learning_rate": 3.7485528455284555e-05, | |
| "loss": 0.2079, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 2.6322970001444554, | |
| "grad_norm": 0.998031735420227, | |
| "learning_rate": 3.732325203252033e-05, | |
| "loss": 0.2193, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 2.6643981830730463, | |
| "grad_norm": 0.9233012199401855, | |
| "learning_rate": 3.716065040650407e-05, | |
| "loss": 0.2219, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 2.696499366001637, | |
| "grad_norm": 2.8505234718322754, | |
| "learning_rate": 3.69980487804878e-05, | |
| "loss": 0.2117, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.728600548930228, | |
| "grad_norm": 5.0437912940979, | |
| "learning_rate": 3.683544715447155e-05, | |
| "loss": 0.195, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 2.760701731858819, | |
| "grad_norm": 3.6365315914154053, | |
| "learning_rate": 3.6672845528455286e-05, | |
| "loss": 0.2184, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 2.79280291478741, | |
| "grad_norm": 6.058934211730957, | |
| "learning_rate": 3.651024390243903e-05, | |
| "loss": 0.2132, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 2.8249040977160007, | |
| "grad_norm": 34.89470291137695, | |
| "learning_rate": 3.6347642276422764e-05, | |
| "loss": 0.2241, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.857005280644592, | |
| "grad_norm": 1.1790027618408203, | |
| "learning_rate": 3.61850406504065e-05, | |
| "loss": 0.2196, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 2.8891064635731825, | |
| "grad_norm": 1.6652076244354248, | |
| "learning_rate": 3.602243902439025e-05, | |
| "loss": 0.2178, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.8891064635731825, | |
| "eval_loss": 0.2256719172000885, | |
| "eval_runtime": 582.4547, | |
| "eval_samples_per_second": 11.9, | |
| "eval_steps_per_second": 11.9, | |
| "eval_wer": 0.11872608382837023, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.921207646501774, | |
| "grad_norm": 1.9215317964553833, | |
| "learning_rate": 3.585983739837399e-05, | |
| "loss": 0.2213, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 2.9533088294303647, | |
| "grad_norm": 13.67776107788086, | |
| "learning_rate": 3.5697560975609756e-05, | |
| "loss": 0.2233, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 2.9854100123589555, | |
| "grad_norm": 5.060420989990234, | |
| "learning_rate": 3.55349593495935e-05, | |
| "loss": 0.2118, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 3.0175111952875464, | |
| "grad_norm": 0.8249053955078125, | |
| "learning_rate": 3.537268292682927e-05, | |
| "loss": 0.2147, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 3.0496123782161373, | |
| "grad_norm": 0.7037186622619629, | |
| "learning_rate": 3.521008130081301e-05, | |
| "loss": 0.1941, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 3.081713561144728, | |
| "grad_norm": 0.7166640758514404, | |
| "learning_rate": 3.5047479674796755e-05, | |
| "loss": 0.1903, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 3.113814744073319, | |
| "grad_norm": 0.8476791977882385, | |
| "learning_rate": 3.488487804878049e-05, | |
| "loss": 0.2004, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 3.14591592700191, | |
| "grad_norm": 0.4231201410293579, | |
| "learning_rate": 3.472260162601626e-05, | |
| "loss": 0.1894, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 3.178017109930501, | |
| "grad_norm": 0.40067389607429504, | |
| "learning_rate": 3.456e-05, | |
| "loss": 0.1773, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 3.2101182928590917, | |
| "grad_norm": 0.7153622508049011, | |
| "learning_rate": 3.439739837398374e-05, | |
| "loss": 0.2046, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.2101182928590917, | |
| "eval_loss": 0.22042368352413177, | |
| "eval_runtime": 575.7584, | |
| "eval_samples_per_second": 12.038, | |
| "eval_steps_per_second": 12.038, | |
| "eval_wer": 0.11246728936259622, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 3.2422194757876825, | |
| "grad_norm": 2.9656982421875, | |
| "learning_rate": 3.4235121951219515e-05, | |
| "loss": 0.1929, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 3.274320658716274, | |
| "grad_norm": 1.7078142166137695, | |
| "learning_rate": 3.4072520325203254e-05, | |
| "loss": 0.193, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 3.3064218416448647, | |
| "grad_norm": 7.298386573791504, | |
| "learning_rate": 3.390991869918699e-05, | |
| "loss": 0.1933, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 3.3385230245734556, | |
| "grad_norm": 0.4513390064239502, | |
| "learning_rate": 3.374731707317073e-05, | |
| "loss": 0.2013, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 3.3706242075020465, | |
| "grad_norm": 0.4610413610935211, | |
| "learning_rate": 3.358471544715447e-05, | |
| "loss": 0.1828, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 3.4027253904306374, | |
| "grad_norm": 0.921280026435852, | |
| "learning_rate": 3.342211382113821e-05, | |
| "loss": 0.1897, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 3.4348265733592283, | |
| "grad_norm": 0.9576189517974854, | |
| "learning_rate": 3.325951219512195e-05, | |
| "loss": 0.184, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 3.466927756287819, | |
| "grad_norm": 0.47692379355430603, | |
| "learning_rate": 3.3096910569105695e-05, | |
| "loss": 0.1904, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 3.49902893921641, | |
| "grad_norm": 1.5692068338394165, | |
| "learning_rate": 3.2934308943089434e-05, | |
| "loss": 0.1905, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 3.531130122145001, | |
| "grad_norm": 6.266742706298828, | |
| "learning_rate": 3.277170731707317e-05, | |
| "loss": 0.1887, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.531130122145001, | |
| "eval_loss": 0.2165077179670334, | |
| "eval_runtime": 597.4689, | |
| "eval_samples_per_second": 11.601, | |
| "eval_steps_per_second": 11.601, | |
| "eval_wer": 0.11000324718736271, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.563231305073592, | |
| "grad_norm": 1.031356692314148, | |
| "learning_rate": 3.260910569105691e-05, | |
| "loss": 0.1883, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 3.5953324880021826, | |
| "grad_norm": 1.9417179822921753, | |
| "learning_rate": 3.244682926829268e-05, | |
| "loss": 0.1956, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 3.627433670930774, | |
| "grad_norm": 0.7637033462524414, | |
| "learning_rate": 3.2284227642276426e-05, | |
| "loss": 0.1877, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 3.659534853859365, | |
| "grad_norm": 0.5973896384239197, | |
| "learning_rate": 3.2121626016260165e-05, | |
| "loss": 0.1756, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 3.6916360367879557, | |
| "grad_norm": 1.629115343093872, | |
| "learning_rate": 3.1959024390243904e-05, | |
| "loss": 0.1867, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 3.7237372197165466, | |
| "grad_norm": 0.4055944085121155, | |
| "learning_rate": 3.179642276422764e-05, | |
| "loss": 0.1766, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 3.7558384026451375, | |
| "grad_norm": 3.1317808628082275, | |
| "learning_rate": 3.163382113821138e-05, | |
| "loss": 0.1928, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 3.7879395855737283, | |
| "grad_norm": 1.9987761974334717, | |
| "learning_rate": 3.147121951219512e-05, | |
| "loss": 0.1775, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 3.820040768502319, | |
| "grad_norm": 1.7839641571044922, | |
| "learning_rate": 3.1308617886178867e-05, | |
| "loss": 0.171, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 3.85214195143091, | |
| "grad_norm": 0.4871620535850525, | |
| "learning_rate": 3.1146016260162605e-05, | |
| "loss": 0.1874, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 3.85214195143091, | |
| "eval_loss": 0.20520555973052979, | |
| "eval_runtime": 598.1599, | |
| "eval_samples_per_second": 11.587, | |
| "eval_steps_per_second": 11.587, | |
| "eval_wer": 0.1052216046199199, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 3.884243134359501, | |
| "grad_norm": 1.415724754333496, | |
| "learning_rate": 3.098341463414634e-05, | |
| "loss": 0.1839, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 3.9163443172880923, | |
| "grad_norm": 0.7966309189796448, | |
| "learning_rate": 3.0820813008130083e-05, | |
| "loss": 0.1676, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 3.9484455002166827, | |
| "grad_norm": 0.7828078866004944, | |
| "learning_rate": 3.065821138211382e-05, | |
| "loss": 0.175, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 3.980546683145274, | |
| "grad_norm": 0.8037332892417908, | |
| "learning_rate": 3.049593495934959e-05, | |
| "loss": 0.1993, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 4.0126478660738645, | |
| "grad_norm": 1.3568751811981201, | |
| "learning_rate": 3.0333333333333337e-05, | |
| "loss": 0.1747, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 4.044749049002456, | |
| "grad_norm": 4.724090576171875, | |
| "learning_rate": 3.0170731707317072e-05, | |
| "loss": 0.16, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 4.076850231931046, | |
| "grad_norm": 4.713302135467529, | |
| "learning_rate": 3.0008130081300818e-05, | |
| "loss": 0.1544, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 4.1089514148596376, | |
| "grad_norm": 2.2513558864593506, | |
| "learning_rate": 2.9845528455284553e-05, | |
| "loss": 0.1673, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 4.141052597788229, | |
| "grad_norm": 0.5379019379615784, | |
| "learning_rate": 2.9682926829268292e-05, | |
| "loss": 0.1598, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 4.173153780716819, | |
| "grad_norm": 2.381577968597412, | |
| "learning_rate": 2.9520325203252035e-05, | |
| "loss": 0.1598, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.173153780716819, | |
| "eval_loss": 0.20186389982700348, | |
| "eval_runtime": 597.2934, | |
| "eval_samples_per_second": 11.604, | |
| "eval_steps_per_second": 11.604, | |
| "eval_wer": 0.10251561515099421, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 4.205254963645411, | |
| "grad_norm": 0.5360216498374939, | |
| "learning_rate": 2.9357723577235774e-05, | |
| "loss": 0.1459, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 4.237356146574001, | |
| "grad_norm": 0.7315216064453125, | |
| "learning_rate": 2.9195121951219513e-05, | |
| "loss": 0.1702, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 4.269457329502592, | |
| "grad_norm": 1.0547746419906616, | |
| "learning_rate": 2.9032520325203255e-05, | |
| "loss": 0.1632, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 4.301558512431183, | |
| "grad_norm": 2.5075976848602295, | |
| "learning_rate": 2.8869918699186994e-05, | |
| "loss": 0.1595, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 4.333659695359774, | |
| "grad_norm": 1.5994772911071777, | |
| "learning_rate": 2.870731707317073e-05, | |
| "loss": 0.1665, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 4.365760878288365, | |
| "grad_norm": 0.7722116112709045, | |
| "learning_rate": 2.8544715447154475e-05, | |
| "loss": 0.1642, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 4.397862061216956, | |
| "grad_norm": 1.0250564813613892, | |
| "learning_rate": 2.838211382113821e-05, | |
| "loss": 0.1687, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 4.429963244145547, | |
| "grad_norm": 0.7502961158752441, | |
| "learning_rate": 2.8219837398373983e-05, | |
| "loss": 0.15, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 4.462064427074138, | |
| "grad_norm": 2.5044519901275635, | |
| "learning_rate": 2.8057560975609758e-05, | |
| "loss": 0.1748, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 4.494165610002729, | |
| "grad_norm": 0.7988201975822449, | |
| "learning_rate": 2.7894959349593497e-05, | |
| "loss": 0.1565, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 4.494165610002729, | |
| "eval_loss": 0.19880281388759613, | |
| "eval_runtime": 595.5489, | |
| "eval_samples_per_second": 11.638, | |
| "eval_steps_per_second": 11.638, | |
| "eval_wer": 0.09975232237566774, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 4.526266792931319, | |
| "grad_norm": 0.47625595331192017, | |
| "learning_rate": 2.7732357723577236e-05, | |
| "loss": 0.1603, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 4.558367975859911, | |
| "grad_norm": 0.46868839859962463, | |
| "learning_rate": 2.7569756097560978e-05, | |
| "loss": 0.1623, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 4.590469158788501, | |
| "grad_norm": 7.4738335609436035, | |
| "learning_rate": 2.7407154471544717e-05, | |
| "loss": 0.152, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 4.6225703417170925, | |
| "grad_norm": 1.6319257020950317, | |
| "learning_rate": 2.724455284552846e-05, | |
| "loss": 0.1473, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 4.654671524645683, | |
| "grad_norm": 1.34164559841156, | |
| "learning_rate": 2.70819512195122e-05, | |
| "loss": 0.1428, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 4.686772707574274, | |
| "grad_norm": 0.5560343861579895, | |
| "learning_rate": 2.6919349593495934e-05, | |
| "loss": 0.1599, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 4.718873890502865, | |
| "grad_norm": 0.8048732280731201, | |
| "learning_rate": 2.6757073170731713e-05, | |
| "loss": 0.1656, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 4.750975073431456, | |
| "grad_norm": 0.577629804611206, | |
| "learning_rate": 2.6594471544715448e-05, | |
| "loss": 0.1533, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 4.783076256360047, | |
| "grad_norm": 1.2069358825683594, | |
| "learning_rate": 2.6431869918699187e-05, | |
| "loss": 0.1411, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 4.815177439288638, | |
| "grad_norm": 2.0387120246887207, | |
| "learning_rate": 2.626926829268293e-05, | |
| "loss": 0.1498, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 4.815177439288638, | |
| "eval_loss": 0.18954835832118988, | |
| "eval_runtime": 595.37, | |
| "eval_samples_per_second": 11.642, | |
| "eval_steps_per_second": 11.642, | |
| "eval_wer": 0.09902011346054668, | |
| "step": 75000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 155750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.299548382938926e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |