| { | |
| "best_metric": 0.5420793793724458, | |
| "best_model_checkpoint": "md_d_l2_arctic/checkpoint-3700", | |
| "epoch": 100.0, | |
| "eval_steps": 100, | |
| "global_step": 7900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 92.3280258178711, | |
| "learning_rate": 2.4050632911392408e-06, | |
| "loss": 14.3657, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "eval_loss": 9.120972633361816, | |
| "eval_runtime": 15.883, | |
| "eval_samples_per_second": 28.332, | |
| "eval_steps_per_second": 1.448, | |
| "eval_wer": 2.4374870125372308, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 1.9874589443206787, | |
| "learning_rate": 4.9113924050632915e-06, | |
| "loss": 4.382, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_loss": 3.421921730041504, | |
| "eval_runtime": 10.9345, | |
| "eval_samples_per_second": 41.154, | |
| "eval_steps_per_second": 2.103, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 3.151137590408325, | |
| "learning_rate": 7.443037974683544e-06, | |
| "loss": 3.2514, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_loss": 2.788123369216919, | |
| "eval_runtime": 10.9315, | |
| "eval_samples_per_second": 41.165, | |
| "eval_steps_per_second": 2.104, | |
| "eval_wer": 0.9979912724250191, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 3.4454073905944824, | |
| "learning_rate": 9.974683544303799e-06, | |
| "loss": 2.4508, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "eval_loss": 1.8000441789627075, | |
| "eval_runtime": 10.9542, | |
| "eval_samples_per_second": 41.08, | |
| "eval_steps_per_second": 2.1, | |
| "eval_wer": 0.7380342176352428, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "grad_norm": 2.928997039794922, | |
| "learning_rate": 1.2506329113924051e-05, | |
| "loss": 1.6168, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "eval_loss": 1.131545901298523, | |
| "eval_runtime": 11.2822, | |
| "eval_samples_per_second": 39.886, | |
| "eval_steps_per_second": 2.039, | |
| "eval_wer": 0.967791092332202, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "grad_norm": 2.174651861190796, | |
| "learning_rate": 1.5037974683544306e-05, | |
| "loss": 1.1212, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "eval_loss": 0.8748846054077148, | |
| "eval_runtime": 11.1275, | |
| "eval_samples_per_second": 40.44, | |
| "eval_steps_per_second": 2.067, | |
| "eval_wer": 1.065803144697652, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "grad_norm": 7.105869770050049, | |
| "learning_rate": 1.7569620253164558e-05, | |
| "loss": 0.8953, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "eval_loss": 0.7655201554298401, | |
| "eval_runtime": 11.2119, | |
| "eval_samples_per_second": 40.136, | |
| "eval_steps_per_second": 2.051, | |
| "eval_wer": 0.9655052988848098, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 10.13, | |
| "grad_norm": 2.2692487239837646, | |
| "learning_rate": 1.99887482419128e-05, | |
| "loss": 0.7684, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 10.13, | |
| "eval_loss": 0.6687426567077637, | |
| "eval_runtime": 11.2107, | |
| "eval_samples_per_second": 40.14, | |
| "eval_steps_per_second": 2.052, | |
| "eval_wer": 0.7621389485350142, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 11.39, | |
| "grad_norm": 3.137871742248535, | |
| "learning_rate": 1.970745428973277e-05, | |
| "loss": 0.6661, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 11.39, | |
| "eval_loss": 0.6319410800933838, | |
| "eval_runtime": 11.2724, | |
| "eval_samples_per_second": 39.921, | |
| "eval_steps_per_second": 2.04, | |
| "eval_wer": 0.6755558634065248, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "grad_norm": 2.894819498062134, | |
| "learning_rate": 1.9426160337552744e-05, | |
| "loss": 0.6306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "eval_loss": 0.6195651292800903, | |
| "eval_runtime": 11.3009, | |
| "eval_samples_per_second": 39.82, | |
| "eval_steps_per_second": 2.035, | |
| "eval_wer": 0.6963358038373624, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 13.92, | |
| "grad_norm": 3.4255077838897705, | |
| "learning_rate": 1.9144866385372717e-05, | |
| "loss": 0.5759, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 13.92, | |
| "eval_loss": 0.5875259637832642, | |
| "eval_runtime": 11.1702, | |
| "eval_samples_per_second": 40.286, | |
| "eval_steps_per_second": 2.059, | |
| "eval_wer": 0.5965228233012398, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 15.19, | |
| "grad_norm": 2.189448118209839, | |
| "learning_rate": 1.8863572433192687e-05, | |
| "loss": 0.5417, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 15.19, | |
| "eval_loss": 0.5779715776443481, | |
| "eval_runtime": 11.2889, | |
| "eval_samples_per_second": 39.862, | |
| "eval_steps_per_second": 2.037, | |
| "eval_wer": 0.6528364618688093, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "grad_norm": 6.146909713745117, | |
| "learning_rate": 1.858227848101266e-05, | |
| "loss": 0.528, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "eval_loss": 0.5798078179359436, | |
| "eval_runtime": 11.2225, | |
| "eval_samples_per_second": 40.098, | |
| "eval_steps_per_second": 2.049, | |
| "eval_wer": 0.6539447253584539, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 17.72, | |
| "grad_norm": 3.1426494121551514, | |
| "learning_rate": 1.8300984528832633e-05, | |
| "loss": 0.4857, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 17.72, | |
| "eval_loss": 0.5568873286247253, | |
| "eval_runtime": 11.201, | |
| "eval_samples_per_second": 40.175, | |
| "eval_steps_per_second": 2.053, | |
| "eval_wer": 0.5724873588695713, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "grad_norm": 1.9612127542495728, | |
| "learning_rate": 1.8019690576652603e-05, | |
| "loss": 0.4655, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "eval_loss": 0.549960196018219, | |
| "eval_runtime": 11.3494, | |
| "eval_samples_per_second": 39.65, | |
| "eval_steps_per_second": 2.027, | |
| "eval_wer": 0.575535083466094, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "grad_norm": 2.946164131164551, | |
| "learning_rate": 1.7738396624472576e-05, | |
| "loss": 0.4526, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "eval_loss": 0.5582863092422485, | |
| "eval_runtime": 11.3026, | |
| "eval_samples_per_second": 39.814, | |
| "eval_steps_per_second": 2.035, | |
| "eval_wer": 0.5776130775091778, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 21.52, | |
| "grad_norm": 2.338858127593994, | |
| "learning_rate": 1.745710267229255e-05, | |
| "loss": 0.4287, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 21.52, | |
| "eval_loss": 0.555654764175415, | |
| "eval_runtime": 11.2714, | |
| "eval_samples_per_second": 39.924, | |
| "eval_steps_per_second": 2.041, | |
| "eval_wer": 0.5609891251645078, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 22.78, | |
| "grad_norm": 2.5261471271514893, | |
| "learning_rate": 1.717580872011252e-05, | |
| "loss": 0.4149, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 22.78, | |
| "eval_loss": 0.5575445294380188, | |
| "eval_runtime": 11.3966, | |
| "eval_samples_per_second": 39.486, | |
| "eval_steps_per_second": 2.018, | |
| "eval_wer": 0.5748424187850661, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 24.05, | |
| "grad_norm": 3.1321146488189697, | |
| "learning_rate": 1.689451476793249e-05, | |
| "loss": 0.3983, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 24.05, | |
| "eval_loss": 0.5648804306983948, | |
| "eval_runtime": 11.191, | |
| "eval_samples_per_second": 40.211, | |
| "eval_steps_per_second": 2.055, | |
| "eval_wer": 0.6003324790468934, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "grad_norm": 2.3534066677093506, | |
| "learning_rate": 1.661322081575246e-05, | |
| "loss": 0.4001, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "eval_loss": 0.567441463470459, | |
| "eval_runtime": 11.1757, | |
| "eval_samples_per_second": 40.266, | |
| "eval_steps_per_second": 2.058, | |
| "eval_wer": 0.5976310867908845, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 26.58, | |
| "grad_norm": 2.417196750640869, | |
| "learning_rate": 1.6331926863572434e-05, | |
| "loss": 0.3649, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 26.58, | |
| "eval_loss": 0.579703152179718, | |
| "eval_runtime": 11.126, | |
| "eval_samples_per_second": 40.446, | |
| "eval_steps_per_second": 2.067, | |
| "eval_wer": 0.5804530027013922, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 27.85, | |
| "grad_norm": 2.8033077716827393, | |
| "learning_rate": 1.6050632911392404e-05, | |
| "loss": 0.3711, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 27.85, | |
| "eval_loss": 0.5839091539382935, | |
| "eval_runtime": 11.2012, | |
| "eval_samples_per_second": 40.174, | |
| "eval_steps_per_second": 2.053, | |
| "eval_wer": 0.6546373900394818, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 29.11, | |
| "grad_norm": 2.2273740768432617, | |
| "learning_rate": 1.576933895921238e-05, | |
| "loss": 0.3547, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 29.11, | |
| "eval_loss": 0.5734866261482239, | |
| "eval_runtime": 11.079, | |
| "eval_samples_per_second": 40.617, | |
| "eval_steps_per_second": 2.076, | |
| "eval_wer": 0.5904273741081942, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 30.38, | |
| "grad_norm": 2.2852895259857178, | |
| "learning_rate": 1.548804500703235e-05, | |
| "loss": 0.3402, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 30.38, | |
| "eval_loss": 0.5698839426040649, | |
| "eval_runtime": 11.0702, | |
| "eval_samples_per_second": 40.65, | |
| "eval_steps_per_second": 2.078, | |
| "eval_wer": 0.5426335111172681, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "grad_norm": 1.869658350944519, | |
| "learning_rate": 1.5206751054852323e-05, | |
| "loss": 0.3414, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "eval_loss": 0.5700486302375793, | |
| "eval_runtime": 11.3836, | |
| "eval_samples_per_second": 39.531, | |
| "eval_steps_per_second": 2.02, | |
| "eval_wer": 0.5421486458405486, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 32.91, | |
| "grad_norm": 1.7649214267730713, | |
| "learning_rate": 1.4925457102672294e-05, | |
| "loss": 0.3255, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 32.91, | |
| "eval_loss": 0.5744786262512207, | |
| "eval_runtime": 11.2135, | |
| "eval_samples_per_second": 40.13, | |
| "eval_steps_per_second": 2.051, | |
| "eval_wer": 0.5663226432084228, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 34.18, | |
| "grad_norm": 2.0651187896728516, | |
| "learning_rate": 1.4644163150492266e-05, | |
| "loss": 0.3093, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 34.18, | |
| "eval_loss": 0.5957615971565247, | |
| "eval_runtime": 11.1504, | |
| "eval_samples_per_second": 40.357, | |
| "eval_steps_per_second": 2.063, | |
| "eval_wer": 0.5931980328323059, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 35.44, | |
| "grad_norm": 2.1554198265075684, | |
| "learning_rate": 1.4362869198312237e-05, | |
| "loss": 0.315, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 35.44, | |
| "eval_loss": 0.5933964848518372, | |
| "eval_runtime": 11.202, | |
| "eval_samples_per_second": 40.171, | |
| "eval_steps_per_second": 2.053, | |
| "eval_wer": 0.5905659070443998, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 36.71, | |
| "grad_norm": 2.880059003829956, | |
| "learning_rate": 1.4081575246132208e-05, | |
| "loss": 0.31, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 36.71, | |
| "eval_loss": 0.6071695685386658, | |
| "eval_runtime": 11.2335, | |
| "eval_samples_per_second": 40.059, | |
| "eval_steps_per_second": 2.047, | |
| "eval_wer": 0.601094410196024, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 37.97, | |
| "grad_norm": 2.7784523963928223, | |
| "learning_rate": 1.380028129395218e-05, | |
| "loss": 0.3026, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 37.97, | |
| "eval_loss": 0.6038002371788025, | |
| "eval_runtime": 11.1544, | |
| "eval_samples_per_second": 40.343, | |
| "eval_steps_per_second": 2.062, | |
| "eval_wer": 0.5760199487428136, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 39.24, | |
| "grad_norm": 2.2978756427764893, | |
| "learning_rate": 1.3518987341772155e-05, | |
| "loss": 0.2802, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 39.24, | |
| "eval_loss": 0.6079789400100708, | |
| "eval_runtime": 11.1769, | |
| "eval_samples_per_second": 40.262, | |
| "eval_steps_per_second": 2.058, | |
| "eval_wer": 0.5776823439772806, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 40.51, | |
| "grad_norm": 2.1417360305786133, | |
| "learning_rate": 1.3237693389592126e-05, | |
| "loss": 0.2835, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 40.51, | |
| "eval_loss": 0.6061974167823792, | |
| "eval_runtime": 11.1226, | |
| "eval_samples_per_second": 40.458, | |
| "eval_steps_per_second": 2.068, | |
| "eval_wer": 0.5743575535083466, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 41.77, | |
| "grad_norm": 1.6934860944747925, | |
| "learning_rate": 1.2956399437412097e-05, | |
| "loss": 0.2585, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 41.77, | |
| "eval_loss": 0.6224856972694397, | |
| "eval_runtime": 11.0743, | |
| "eval_samples_per_second": 40.634, | |
| "eval_steps_per_second": 2.077, | |
| "eval_wer": 0.5784442751264113, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 43.04, | |
| "grad_norm": 1.937164068222046, | |
| "learning_rate": 1.2675105485232069e-05, | |
| "loss": 0.2699, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 43.04, | |
| "eval_loss": 0.6225900650024414, | |
| "eval_runtime": 11.1534, | |
| "eval_samples_per_second": 40.347, | |
| "eval_steps_per_second": 2.062, | |
| "eval_wer": 0.5664611761446284, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 44.3, | |
| "grad_norm": 2.5404953956604004, | |
| "learning_rate": 1.239381153305204e-05, | |
| "loss": 0.2785, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 44.3, | |
| "eval_loss": 0.6240466833114624, | |
| "eval_runtime": 11.1649, | |
| "eval_samples_per_second": 40.305, | |
| "eval_steps_per_second": 2.06, | |
| "eval_wer": 0.5713790953799266, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 45.57, | |
| "grad_norm": 5.070058822631836, | |
| "learning_rate": 1.2112517580872011e-05, | |
| "loss": 0.2689, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 45.57, | |
| "eval_loss": 0.6294780969619751, | |
| "eval_runtime": 11.118, | |
| "eval_samples_per_second": 40.475, | |
| "eval_steps_per_second": 2.069, | |
| "eval_wer": 0.5648680473782642, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 46.84, | |
| "grad_norm": 2.5456833839416504, | |
| "learning_rate": 1.1831223628691983e-05, | |
| "loss": 0.2514, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 46.84, | |
| "eval_loss": 0.6424580812454224, | |
| "eval_runtime": 11.271, | |
| "eval_samples_per_second": 39.926, | |
| "eval_steps_per_second": 2.041, | |
| "eval_wer": 0.5420793793724458, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 48.1, | |
| "grad_norm": 2.153717279434204, | |
| "learning_rate": 1.1549929676511956e-05, | |
| "loss": 0.2433, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 48.1, | |
| "eval_loss": 0.6667928099632263, | |
| "eval_runtime": 11.226, | |
| "eval_samples_per_second": 40.086, | |
| "eval_steps_per_second": 2.049, | |
| "eval_wer": 0.606774260580453, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 49.37, | |
| "grad_norm": 2.5784971714019775, | |
| "learning_rate": 1.1268635724331929e-05, | |
| "loss": 0.2403, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 49.37, | |
| "eval_loss": 0.6562526226043701, | |
| "eval_runtime": 11.112, | |
| "eval_samples_per_second": 40.497, | |
| "eval_steps_per_second": 2.07, | |
| "eval_wer": 0.5749809517212717, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 50.63, | |
| "grad_norm": 2.656663656234741, | |
| "learning_rate": 1.09873417721519e-05, | |
| "loss": 0.2287, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 50.63, | |
| "eval_loss": 0.6695858240127563, | |
| "eval_runtime": 11.0907, | |
| "eval_samples_per_second": 40.575, | |
| "eval_steps_per_second": 2.074, | |
| "eval_wer": 0.5932672993004087, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 51.9, | |
| "grad_norm": 2.9128212928771973, | |
| "learning_rate": 1.0706047819971872e-05, | |
| "loss": 0.2366, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 51.9, | |
| "eval_loss": 0.6738879680633545, | |
| "eval_runtime": 11.1372, | |
| "eval_samples_per_second": 40.405, | |
| "eval_steps_per_second": 2.065, | |
| "eval_wer": 0.5731107570824964, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 53.16, | |
| "grad_norm": 1.8941991329193115, | |
| "learning_rate": 1.0424753867791843e-05, | |
| "loss": 0.2295, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 53.16, | |
| "eval_loss": 0.680944561958313, | |
| "eval_runtime": 11.1339, | |
| "eval_samples_per_second": 40.417, | |
| "eval_steps_per_second": 2.066, | |
| "eval_wer": 0.6090600540278451, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 54.43, | |
| "grad_norm": 2.2944602966308594, | |
| "learning_rate": 1.0143459915611814e-05, | |
| "loss": 0.2274, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 54.43, | |
| "eval_loss": 0.6874995827674866, | |
| "eval_runtime": 11.1569, | |
| "eval_samples_per_second": 40.334, | |
| "eval_steps_per_second": 2.061, | |
| "eval_wer": 0.5913971046616333, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 55.7, | |
| "grad_norm": 2.4684228897094727, | |
| "learning_rate": 9.862165963431787e-06, | |
| "loss": 0.2178, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 55.7, | |
| "eval_loss": 0.6899309158325195, | |
| "eval_runtime": 11.21, | |
| "eval_samples_per_second": 40.143, | |
| "eval_steps_per_second": 2.052, | |
| "eval_wer": 0.5949296945348757, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 56.96, | |
| "grad_norm": 2.0573887825012207, | |
| "learning_rate": 9.580872011251759e-06, | |
| "loss": 0.2176, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 56.96, | |
| "eval_loss": 0.6924750208854675, | |
| "eval_runtime": 11.1413, | |
| "eval_samples_per_second": 40.39, | |
| "eval_steps_per_second": 2.064, | |
| "eval_wer": 0.5828080626168871, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 58.23, | |
| "grad_norm": 1.9555515050888062, | |
| "learning_rate": 9.299578059071732e-06, | |
| "loss": 0.2064, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 58.23, | |
| "eval_loss": 0.7009023427963257, | |
| "eval_runtime": 11.0931, | |
| "eval_samples_per_second": 40.566, | |
| "eval_steps_per_second": 2.073, | |
| "eval_wer": 0.598462284408118, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 59.49, | |
| "grad_norm": 3.208376169204712, | |
| "learning_rate": 9.018284106891703e-06, | |
| "loss": 0.2081, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 59.49, | |
| "eval_loss": 0.701277494430542, | |
| "eval_runtime": 11.216, | |
| "eval_samples_per_second": 40.121, | |
| "eval_steps_per_second": 2.051, | |
| "eval_wer": 0.5995705478977627, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 60.76, | |
| "grad_norm": 2.5926976203918457, | |
| "learning_rate": 8.736990154711675e-06, | |
| "loss": 0.2093, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 60.76, | |
| "eval_loss": 0.725727379322052, | |
| "eval_runtime": 11.1473, | |
| "eval_samples_per_second": 40.369, | |
| "eval_steps_per_second": 2.063, | |
| "eval_wer": 0.6086444552192284, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 62.03, | |
| "grad_norm": 2.2040176391601562, | |
| "learning_rate": 8.455696202531646e-06, | |
| "loss": 0.2024, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 62.03, | |
| "eval_loss": 0.7215314507484436, | |
| "eval_runtime": 11.4213, | |
| "eval_samples_per_second": 39.4, | |
| "eval_steps_per_second": 2.014, | |
| "eval_wer": 0.6003324790468934, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 63.29, | |
| "grad_norm": 1.6947568655014038, | |
| "learning_rate": 8.174402250351619e-06, | |
| "loss": 0.1999, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 63.29, | |
| "eval_loss": 0.7332788109779358, | |
| "eval_runtime": 11.2504, | |
| "eval_samples_per_second": 39.999, | |
| "eval_steps_per_second": 2.044, | |
| "eval_wer": 0.6090600540278451, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 64.56, | |
| "grad_norm": 2.657949686050415, | |
| "learning_rate": 7.89310829817159e-06, | |
| "loss": 0.2064, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 64.56, | |
| "eval_loss": 0.7529835104942322, | |
| "eval_runtime": 11.2168, | |
| "eval_samples_per_second": 40.118, | |
| "eval_steps_per_second": 2.05, | |
| "eval_wer": 0.6397450993973818, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 65.82, | |
| "grad_norm": 2.161647081375122, | |
| "learning_rate": 7.611814345991562e-06, | |
| "loss": 0.186, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 65.82, | |
| "eval_loss": 0.7542085647583008, | |
| "eval_runtime": 11.1628, | |
| "eval_samples_per_second": 40.312, | |
| "eval_steps_per_second": 2.06, | |
| "eval_wer": 0.6348964466301863, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 67.09, | |
| "grad_norm": 1.5503740310668945, | |
| "learning_rate": 7.330520393811533e-06, | |
| "loss": 0.186, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 67.09, | |
| "eval_loss": 0.7416096925735474, | |
| "eval_runtime": 11.1272, | |
| "eval_samples_per_second": 40.441, | |
| "eval_steps_per_second": 2.067, | |
| "eval_wer": 0.6270000692664681, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 68.35, | |
| "grad_norm": 2.8439977169036865, | |
| "learning_rate": 7.049226441631506e-06, | |
| "loss": 0.1807, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 68.35, | |
| "eval_loss": 0.7548705339431763, | |
| "eval_runtime": 11.1833, | |
| "eval_samples_per_second": 40.239, | |
| "eval_steps_per_second": 2.057, | |
| "eval_wer": 0.6352427789707003, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 69.62, | |
| "grad_norm": 3.1191818714141846, | |
| "learning_rate": 6.7679324894514775e-06, | |
| "loss": 0.1784, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 69.62, | |
| "eval_loss": 0.7506438493728638, | |
| "eval_runtime": 11.1231, | |
| "eval_samples_per_second": 40.456, | |
| "eval_steps_per_second": 2.068, | |
| "eval_wer": 0.5844011913832514, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 70.89, | |
| "grad_norm": 2.1088929176330566, | |
| "learning_rate": 6.486638537271449e-06, | |
| "loss": 0.1824, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 70.89, | |
| "eval_loss": 0.7611370086669922, | |
| "eval_runtime": 11.0844, | |
| "eval_samples_per_second": 40.598, | |
| "eval_steps_per_second": 2.075, | |
| "eval_wer": 0.6252684075638983, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 72.15, | |
| "grad_norm": 3.0879805088043213, | |
| "learning_rate": 6.208157524613221e-06, | |
| "loss": 0.1769, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 72.15, | |
| "eval_loss": 0.771263837814331, | |
| "eval_runtime": 11.226, | |
| "eval_samples_per_second": 40.085, | |
| "eval_steps_per_second": 2.049, | |
| "eval_wer": 0.5927131675555863, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 73.42, | |
| "grad_norm": 2.1844921112060547, | |
| "learning_rate": 5.926863572433193e-06, | |
| "loss": 0.1843, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 73.42, | |
| "eval_loss": 0.7719753980636597, | |
| "eval_runtime": 11.1675, | |
| "eval_samples_per_second": 40.295, | |
| "eval_steps_per_second": 2.06, | |
| "eval_wer": 0.5955530927478008, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 74.68, | |
| "grad_norm": 2.3081653118133545, | |
| "learning_rate": 5.645569620253165e-06, | |
| "loss": 0.1709, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 74.68, | |
| "eval_loss": 0.7804738879203796, | |
| "eval_runtime": 11.2408, | |
| "eval_samples_per_second": 40.033, | |
| "eval_steps_per_second": 2.046, | |
| "eval_wer": 0.6258225393087207, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 75.95, | |
| "grad_norm": 3.2150704860687256, | |
| "learning_rate": 5.364275668073137e-06, | |
| "loss": 0.1691, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 75.95, | |
| "eval_loss": 0.7865281105041504, | |
| "eval_runtime": 11.2778, | |
| "eval_samples_per_second": 39.901, | |
| "eval_steps_per_second": 2.039, | |
| "eval_wer": 0.6281775992242156, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 77.22, | |
| "grad_norm": 1.8031377792358398, | |
| "learning_rate": 5.082981715893108e-06, | |
| "loss": 0.1701, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 77.22, | |
| "eval_loss": 0.7807941436767578, | |
| "eval_runtime": 11.2568, | |
| "eval_samples_per_second": 39.976, | |
| "eval_steps_per_second": 2.043, | |
| "eval_wer": 0.6218050841587588, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 78.48, | |
| "grad_norm": 1.8435957431793213, | |
| "learning_rate": 4.8016877637130805e-06, | |
| "loss": 0.1735, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 78.48, | |
| "eval_loss": 0.7789934873580933, | |
| "eval_runtime": 11.2304, | |
| "eval_samples_per_second": 40.07, | |
| "eval_steps_per_second": 2.048, | |
| "eval_wer": 0.5965920897693426, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 79.75, | |
| "grad_norm": 1.9381072521209717, | |
| "learning_rate": 4.520393811533053e-06, | |
| "loss": 0.1746, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 79.75, | |
| "eval_loss": 0.7949352264404297, | |
| "eval_runtime": 11.4313, | |
| "eval_samples_per_second": 39.366, | |
| "eval_steps_per_second": 2.012, | |
| "eval_wer": 0.6430698898663157, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 81.01, | |
| "grad_norm": 1.8376802206039429, | |
| "learning_rate": 4.239099859353024e-06, | |
| "loss": 0.1745, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 81.01, | |
| "eval_loss": 0.8125633001327515, | |
| "eval_runtime": 11.456, | |
| "eval_samples_per_second": 39.281, | |
| "eval_steps_per_second": 2.008, | |
| "eval_wer": 0.6284546650966267, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 82.28, | |
| "grad_norm": 2.187868118286133, | |
| "learning_rate": 3.957805907172996e-06, | |
| "loss": 0.1605, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 82.28, | |
| "eval_loss": 0.8113065361976624, | |
| "eval_runtime": 11.3144, | |
| "eval_samples_per_second": 39.772, | |
| "eval_steps_per_second": 2.033, | |
| "eval_wer": 0.6194500242432638, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 83.54, | |
| "grad_norm": 2.5882511138916016, | |
| "learning_rate": 3.676511954992968e-06, | |
| "loss": 0.1579, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 83.54, | |
| "eval_loss": 0.7976768612861633, | |
| "eval_runtime": 11.4022, | |
| "eval_samples_per_second": 39.466, | |
| "eval_steps_per_second": 2.017, | |
| "eval_wer": 0.6155018355614047, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 84.81, | |
| "grad_norm": 2.416449546813965, | |
| "learning_rate": 3.39521800281294e-06, | |
| "loss": 0.1704, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 84.81, | |
| "eval_loss": 0.8016535043716431, | |
| "eval_runtime": 11.3366, | |
| "eval_samples_per_second": 39.694, | |
| "eval_steps_per_second": 2.029, | |
| "eval_wer": 0.6139779732631433, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 86.08, | |
| "grad_norm": 1.5159286260604858, | |
| "learning_rate": 3.1139240506329116e-06, | |
| "loss": 0.1659, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 86.08, | |
| "eval_loss": 0.8146914839744568, | |
| "eval_runtime": 11.289, | |
| "eval_samples_per_second": 39.862, | |
| "eval_steps_per_second": 2.037, | |
| "eval_wer": 0.6279005333518044, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 87.34, | |
| "grad_norm": 2.9167511463165283, | |
| "learning_rate": 2.832630098452884e-06, | |
| "loss": 0.166, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 87.34, | |
| "eval_loss": 0.8088270425796509, | |
| "eval_runtime": 11.28, | |
| "eval_samples_per_second": 39.893, | |
| "eval_steps_per_second": 2.039, | |
| "eval_wer": 0.6350349795663919, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 88.61, | |
| "grad_norm": 2.3707916736602783, | |
| "learning_rate": 2.5513361462728552e-06, | |
| "loss": 0.1539, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 88.61, | |
| "eval_loss": 0.8052927255630493, | |
| "eval_runtime": 11.3196, | |
| "eval_samples_per_second": 39.754, | |
| "eval_steps_per_second": 2.032, | |
| "eval_wer": 0.616402299646741, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 89.87, | |
| "grad_norm": 2.315516471862793, | |
| "learning_rate": 2.270042194092827e-06, | |
| "loss": 0.1589, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 89.87, | |
| "eval_loss": 0.8188755512237549, | |
| "eval_runtime": 11.3039, | |
| "eval_samples_per_second": 39.809, | |
| "eval_steps_per_second": 2.035, | |
| "eval_wer": 0.6357276442474198, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 91.14, | |
| "grad_norm": 1.863142967224121, | |
| "learning_rate": 1.9887482419127992e-06, | |
| "loss": 0.1559, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 91.14, | |
| "eval_loss": 0.8152031898498535, | |
| "eval_runtime": 11.2054, | |
| "eval_samples_per_second": 40.159, | |
| "eval_steps_per_second": 2.053, | |
| "eval_wer": 0.6258225393087207, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 92.41, | |
| "grad_norm": 3.034898519515991, | |
| "learning_rate": 1.7074542897327708e-06, | |
| "loss": 0.1564, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 92.41, | |
| "eval_loss": 0.8190972208976746, | |
| "eval_runtime": 11.3427, | |
| "eval_samples_per_second": 39.673, | |
| "eval_steps_per_second": 2.028, | |
| "eval_wer": 0.6245064764147676, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 93.67, | |
| "grad_norm": 1.673194408416748, | |
| "learning_rate": 1.4261603375527428e-06, | |
| "loss": 0.158, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 93.67, | |
| "eval_loss": 0.8254526853561401, | |
| "eval_runtime": 11.282, | |
| "eval_samples_per_second": 39.887, | |
| "eval_steps_per_second": 2.039, | |
| "eval_wer": 0.6333033178638221, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 94.94, | |
| "grad_norm": 1.6554739475250244, | |
| "learning_rate": 1.1448663853727146e-06, | |
| "loss": 0.1595, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 94.94, | |
| "eval_loss": 0.8184179663658142, | |
| "eval_runtime": 11.2214, | |
| "eval_samples_per_second": 40.102, | |
| "eval_steps_per_second": 2.05, | |
| "eval_wer": 0.6205582877329086, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 96.2, | |
| "grad_norm": 2.310234785079956, | |
| "learning_rate": 8.635724331926865e-07, | |
| "loss": 0.1638, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 96.2, | |
| "eval_loss": 0.8229891061782837, | |
| "eval_runtime": 11.1115, | |
| "eval_samples_per_second": 40.499, | |
| "eval_steps_per_second": 2.07, | |
| "eval_wer": 0.6364203089284477, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 97.47, | |
| "grad_norm": 1.6313074827194214, | |
| "learning_rate": 5.822784810126583e-07, | |
| "loss": 0.1629, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 97.47, | |
| "eval_loss": 0.8244702219963074, | |
| "eval_runtime": 11.1778, | |
| "eval_samples_per_second": 40.258, | |
| "eval_steps_per_second": 2.058, | |
| "eval_wer": 0.6312253238207384, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 98.73, | |
| "grad_norm": 2.2342050075531006, | |
| "learning_rate": 3.009845288326301e-07, | |
| "loss": 0.1531, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 98.73, | |
| "eval_loss": 0.8226235508918762, | |
| "eval_runtime": 11.0989, | |
| "eval_samples_per_second": 40.545, | |
| "eval_steps_per_second": 2.072, | |
| "eval_wer": 0.626723003394057, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 3.3648197650909424, | |
| "learning_rate": 1.9690576652601972e-08, | |
| "loss": 0.1572, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 0.8258158564567566, | |
| "eval_runtime": 11.1124, | |
| "eval_samples_per_second": 40.495, | |
| "eval_steps_per_second": 2.07, | |
| "eval_wer": 0.6288702639052435, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 7900, | |
| "total_flos": 3.6740184088961606e+19, | |
| "train_loss": 0.6056561310683625, | |
| "train_runtime": 25369.8434, | |
| "train_samples_per_second": 12.404, | |
| "train_steps_per_second": 0.311 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 7900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 100, | |
| "total_flos": 3.6740184088961606e+19, | |
| "train_batch_size": 20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |