| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.985936343449296, | |
| "eval_steps": 200, | |
| "global_step": 6750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.29607698001480387, | |
| "grad_norm": 2.7709997274459304, | |
| "learning_rate": 1.9407407407407407e-05, | |
| "loss": 6.1641, | |
| "mean_token_accuracy": 0.1623367673992674, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29607698001480387, | |
| "eval_loss": 4.407998561859131, | |
| "eval_mean_token_accuracy": 0.25814827533577533, | |
| "eval_runtime": 17.8969, | |
| "eval_samples_per_second": 7.152, | |
| "eval_steps_per_second": 0.894, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5921539600296077, | |
| "grad_norm": 3.4130161394563543, | |
| "learning_rate": 1.8814814814814816e-05, | |
| "loss": 3.9329, | |
| "mean_token_accuracy": 0.30167891483516485, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5921539600296077, | |
| "eval_loss": 3.607056140899658, | |
| "eval_mean_token_accuracy": 0.337335927960928, | |
| "eval_runtime": 17.8919, | |
| "eval_samples_per_second": 7.154, | |
| "eval_steps_per_second": 0.894, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8882309400444115, | |
| "grad_norm": 1.9205457771259444, | |
| "learning_rate": 1.8222222222222224e-05, | |
| "loss": 3.3827, | |
| "mean_token_accuracy": 0.35828514194139194, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8882309400444115, | |
| "eval_loss": 3.2180376052856445, | |
| "eval_mean_token_accuracy": 0.3790006868131868, | |
| "eval_runtime": 17.8941, | |
| "eval_samples_per_second": 7.153, | |
| "eval_steps_per_second": 0.894, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1835677276091783, | |
| "grad_norm": 2.219040531148233, | |
| "learning_rate": 1.7629629629629633e-05, | |
| "loss": 3.0463, | |
| "mean_token_accuracy": 0.3991521079866944, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1835677276091783, | |
| "eval_loss": 2.958108901977539, | |
| "eval_mean_token_accuracy": 0.41293498168498166, | |
| "eval_runtime": 17.8905, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4796447076239823, | |
| "grad_norm": 1.9578494903008412, | |
| "learning_rate": 1.7037037037037038e-05, | |
| "loss": 2.8108, | |
| "mean_token_accuracy": 0.4309763431013431, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4796447076239823, | |
| "eval_loss": 2.753509044647217, | |
| "eval_mean_token_accuracy": 0.4397664835164835, | |
| "eval_runtime": 17.886, | |
| "eval_samples_per_second": 7.156, | |
| "eval_steps_per_second": 0.895, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7757216876387862, | |
| "grad_norm": 1.6226934039550227, | |
| "learning_rate": 1.6444444444444444e-05, | |
| "loss": 2.6416, | |
| "mean_token_accuracy": 0.4566705586080586, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7757216876387862, | |
| "eval_loss": 2.5862350463867188, | |
| "eval_mean_token_accuracy": 0.4640758547008547, | |
| "eval_runtime": 17.9046, | |
| "eval_samples_per_second": 7.149, | |
| "eval_steps_per_second": 0.894, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.071058475203553, | |
| "grad_norm": 1.5520186396248283, | |
| "learning_rate": 1.5851851851851852e-05, | |
| "loss": 2.4744, | |
| "mean_token_accuracy": 0.48079080485095527, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.071058475203553, | |
| "eval_loss": 2.44563627243042, | |
| "eval_mean_token_accuracy": 0.4833009004884005, | |
| "eval_runtime": 17.8972, | |
| "eval_samples_per_second": 7.152, | |
| "eval_steps_per_second": 0.894, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.3671354552183566, | |
| "grad_norm": 1.5647563757923375, | |
| "learning_rate": 1.525925925925926e-05, | |
| "loss": 2.3221, | |
| "mean_token_accuracy": 0.5035665064102564, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.3671354552183566, | |
| "eval_loss": 2.3370866775512695, | |
| "eval_mean_token_accuracy": 0.49927884615384616, | |
| "eval_runtime": 17.8991, | |
| "eval_samples_per_second": 7.151, | |
| "eval_steps_per_second": 0.894, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.6632124352331608, | |
| "grad_norm": 1.5085290967116813, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 2.2258, | |
| "mean_token_accuracy": 0.5176262591575091, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.6632124352331608, | |
| "eval_loss": 2.240328550338745, | |
| "eval_mean_token_accuracy": 0.5144898504273504, | |
| "eval_runtime": 17.8923, | |
| "eval_samples_per_second": 7.154, | |
| "eval_steps_per_second": 0.894, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.9592894152479645, | |
| "grad_norm": 1.5061171138250924, | |
| "learning_rate": 1.4074074074074075e-05, | |
| "loss": 2.1464, | |
| "mean_token_accuracy": 0.5297419108669108, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.9592894152479645, | |
| "eval_loss": 2.162349224090576, | |
| "eval_mean_token_accuracy": 0.5255132020757021, | |
| "eval_runtime": 17.9046, | |
| "eval_samples_per_second": 7.149, | |
| "eval_steps_per_second": 0.894, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.254626202812731, | |
| "grad_norm": 1.5604440734793847, | |
| "learning_rate": 1.3481481481481482e-05, | |
| "loss": 2.0498, | |
| "mean_token_accuracy": 0.5452948610843348, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.254626202812731, | |
| "eval_loss": 2.100726366043091, | |
| "eval_mean_token_accuracy": 0.534930173992674, | |
| "eval_runtime": 17.8785, | |
| "eval_samples_per_second": 7.159, | |
| "eval_steps_per_second": 0.895, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.5507031828275353, | |
| "grad_norm": 1.5621392816113453, | |
| "learning_rate": 1.288888888888889e-05, | |
| "loss": 1.9897, | |
| "mean_token_accuracy": 0.554459249084249, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.5507031828275353, | |
| "eval_loss": 2.05090069770813, | |
| "eval_mean_token_accuracy": 0.5429983211233211, | |
| "eval_runtime": 17.896, | |
| "eval_samples_per_second": 7.152, | |
| "eval_steps_per_second": 0.894, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.846780162842339, | |
| "grad_norm": 1.2766839744837124, | |
| "learning_rate": 1.2296296296296298e-05, | |
| "loss": 1.9316, | |
| "mean_token_accuracy": 0.5623727106227107, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.846780162842339, | |
| "eval_loss": 1.9980659484863281, | |
| "eval_mean_token_accuracy": 0.5504884004884005, | |
| "eval_runtime": 17.8898, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.142116950407106, | |
| "grad_norm": 1.3499621558383104, | |
| "learning_rate": 1.1703703703703703e-05, | |
| "loss": 1.8843, | |
| "mean_token_accuracy": 0.5703756491350477, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.142116950407106, | |
| "eval_loss": 1.9619176387786865, | |
| "eval_mean_token_accuracy": 0.5568185286935287, | |
| "eval_runtime": 17.8884, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.438193930421909, | |
| "grad_norm": 1.2966133165796498, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.8236, | |
| "mean_token_accuracy": 0.58019971001221, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.438193930421909, | |
| "eval_loss": 1.928423523902893, | |
| "eval_mean_token_accuracy": 0.5617177960927962, | |
| "eval_runtime": 17.9022, | |
| "eval_samples_per_second": 7.15, | |
| "eval_steps_per_second": 0.894, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.734270910436713, | |
| "grad_norm": 1.3799177772818907, | |
| "learning_rate": 1.0518518518518519e-05, | |
| "loss": 1.8055, | |
| "mean_token_accuracy": 0.5831678113553114, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.734270910436713, | |
| "eval_loss": 1.8989052772521973, | |
| "eval_mean_token_accuracy": 0.5659836691086692, | |
| "eval_runtime": 17.9121, | |
| "eval_samples_per_second": 7.146, | |
| "eval_steps_per_second": 0.893, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.02960769800148, | |
| "grad_norm": 1.189701686001914, | |
| "learning_rate": 9.925925925925927e-06, | |
| "loss": 1.7795, | |
| "mean_token_accuracy": 0.5869535331001496, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.02960769800148, | |
| "eval_loss": 1.8730087280273438, | |
| "eval_mean_token_accuracy": 0.5700644841269842, | |
| "eval_runtime": 17.8877, | |
| "eval_samples_per_second": 7.156, | |
| "eval_steps_per_second": 0.894, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.325684678016284, | |
| "grad_norm": 1.2336131696453405, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 1.7181, | |
| "mean_token_accuracy": 0.5980093101343101, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.325684678016284, | |
| "eval_loss": 1.8516058921813965, | |
| "eval_mean_token_accuracy": 0.5734088827838828, | |
| "eval_runtime": 17.8893, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.6217616580310885, | |
| "grad_norm": 1.3072064080159254, | |
| "learning_rate": 8.740740740740741e-06, | |
| "loss": 1.7039, | |
| "mean_token_accuracy": 0.5995736797924298, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 5.6217616580310885, | |
| "eval_loss": 1.8317267894744873, | |
| "eval_mean_token_accuracy": 0.5765376984126984, | |
| "eval_runtime": 17.8909, | |
| "eval_samples_per_second": 7.154, | |
| "eval_steps_per_second": 0.894, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 5.917838638045892, | |
| "grad_norm": 1.1207509804360356, | |
| "learning_rate": 8.148148148148148e-06, | |
| "loss": 1.6899, | |
| "mean_token_accuracy": 0.6014800442612942, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.917838638045892, | |
| "eval_loss": 1.8143231868743896, | |
| "eval_mean_token_accuracy": 0.5789148351648352, | |
| "eval_runtime": 17.9255, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 0.893, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.213175425610658, | |
| "grad_norm": 1.1903087300751234, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 1.6474, | |
| "mean_token_accuracy": 0.6092672156581931, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.213175425610658, | |
| "eval_loss": 1.8032296895980835, | |
| "eval_mean_token_accuracy": 0.580849358974359, | |
| "eval_runtime": 17.9223, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 0.893, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.509252405625462, | |
| "grad_norm": 1.1713052935721413, | |
| "learning_rate": 6.962962962962964e-06, | |
| "loss": 1.6235, | |
| "mean_token_accuracy": 0.6131747557997558, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.509252405625462, | |
| "eval_loss": 1.7903690338134766, | |
| "eval_mean_token_accuracy": 0.5832226800976801, | |
| "eval_runtime": 17.9071, | |
| "eval_samples_per_second": 7.148, | |
| "eval_steps_per_second": 0.893, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.805329385640267, | |
| "grad_norm": 1.219773082768343, | |
| "learning_rate": 6.370370370370371e-06, | |
| "loss": 1.6199, | |
| "mean_token_accuracy": 0.6140330815018316, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 6.805329385640267, | |
| "eval_loss": 1.775641679763794, | |
| "eval_mean_token_accuracy": 0.5847260378510378, | |
| "eval_runtime": 17.9026, | |
| "eval_samples_per_second": 7.15, | |
| "eval_steps_per_second": 0.894, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.100666173205033, | |
| "grad_norm": 1.227329336420539, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 1.5968, | |
| "mean_token_accuracy": 0.6178009507896726, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.100666173205033, | |
| "eval_loss": 1.767627239227295, | |
| "eval_mean_token_accuracy": 0.5867120726495727, | |
| "eval_runtime": 17.9009, | |
| "eval_samples_per_second": 7.15, | |
| "eval_steps_per_second": 0.894, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.3967431532198376, | |
| "grad_norm": 1.190619502177119, | |
| "learning_rate": 5.185185185185185e-06, | |
| "loss": 1.564, | |
| "mean_token_accuracy": 0.6240972603785103, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.3967431532198376, | |
| "eval_loss": 1.7600514888763428, | |
| "eval_mean_token_accuracy": 0.5872462606837607, | |
| "eval_runtime": 17.8856, | |
| "eval_samples_per_second": 7.157, | |
| "eval_steps_per_second": 0.895, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.692820133234641, | |
| "grad_norm": 1.0951341674513762, | |
| "learning_rate": 4.592592592592593e-06, | |
| "loss": 1.5614, | |
| "mean_token_accuracy": 0.624267322954823, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 7.692820133234641, | |
| "eval_loss": 1.7513068914413452, | |
| "eval_mean_token_accuracy": 0.5891063797313797, | |
| "eval_runtime": 17.8894, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 7.988897113249445, | |
| "grad_norm": 1.0806878005212168, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.5575, | |
| "mean_token_accuracy": 0.6250475427350427, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 7.988897113249445, | |
| "eval_loss": 1.7434966564178467, | |
| "eval_mean_token_accuracy": 0.5898237179487179, | |
| "eval_runtime": 17.917, | |
| "eval_samples_per_second": 7.144, | |
| "eval_steps_per_second": 0.893, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 8.284233900814211, | |
| "grad_norm": 1.1517570180426586, | |
| "learning_rate": 3.4074074074074077e-06, | |
| "loss": 1.5248, | |
| "mean_token_accuracy": 0.6320280631370857, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 8.284233900814211, | |
| "eval_loss": 1.741744041442871, | |
| "eval_mean_token_accuracy": 0.5906135531135531, | |
| "eval_runtime": 17.8965, | |
| "eval_samples_per_second": 7.152, | |
| "eval_steps_per_second": 0.894, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 8.580310880829016, | |
| "grad_norm": 1.0340477842720268, | |
| "learning_rate": 2.814814814814815e-06, | |
| "loss": 1.5212, | |
| "mean_token_accuracy": 0.6322646138583639, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 8.580310880829016, | |
| "eval_loss": 1.735644817352295, | |
| "eval_mean_token_accuracy": 0.5920615842490843, | |
| "eval_runtime": 17.8899, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 8.876387860843819, | |
| "grad_norm": 1.030414959224373, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.5159, | |
| "mean_token_accuracy": 0.6326660561660562, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 8.876387860843819, | |
| "eval_loss": 1.7318824529647827, | |
| "eval_mean_token_accuracy": 0.5921760531135531, | |
| "eval_runtime": 17.9183, | |
| "eval_samples_per_second": 7.144, | |
| "eval_steps_per_second": 0.893, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.171724648408587, | |
| "grad_norm": 0.9859235701888986, | |
| "learning_rate": 1.62962962962963e-06, | |
| "loss": 1.5007, | |
| "mean_token_accuracy": 0.6362307248585444, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 9.171724648408587, | |
| "eval_loss": 1.7302496433258057, | |
| "eval_mean_token_accuracy": 0.5925709706959706, | |
| "eval_runtime": 17.9121, | |
| "eval_samples_per_second": 7.146, | |
| "eval_steps_per_second": 0.893, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 9.46780162842339, | |
| "grad_norm": 0.9796094549613502, | |
| "learning_rate": 1.0370370370370371e-06, | |
| "loss": 1.488, | |
| "mean_token_accuracy": 0.638957036019536, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 9.46780162842339, | |
| "eval_loss": 1.7277941703796387, | |
| "eval_mean_token_accuracy": 0.5929544413919414, | |
| "eval_runtime": 17.9246, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 0.893, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 9.763878608438194, | |
| "grad_norm": 0.9868368122782655, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 1.4899, | |
| "mean_token_accuracy": 0.6385683379120879, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 9.763878608438194, | |
| "eval_loss": 1.7258272171020508, | |
| "eval_mean_token_accuracy": 0.5932234432234432, | |
| "eval_runtime": 17.8902, | |
| "eval_samples_per_second": 7.155, | |
| "eval_steps_per_second": 0.894, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 9.985936343449296, | |
| "step": 6750, | |
| "total_flos": 768626978193408.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 1.3815, | |
| "train_samples_per_second": 2424.882, | |
| "train_steps_per_second": 152.008 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 768626978193408.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |