| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 5000, | |
| "global_step": 52737, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005688605722737357, | |
| "grad_norm": 2.727705717086792, | |
| "learning_rate": 0.0007984982080891974, | |
| "loss": 2.3682, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.011377211445474714, | |
| "grad_norm": 2.5978145599365234, | |
| "learning_rate": 0.0007969812465631342, | |
| "loss": 2.2098, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01706581716821207, | |
| "grad_norm": 2.488832473754883, | |
| "learning_rate": 0.0007954642850370708, | |
| "loss": 2.1041, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.02275442289094943, | |
| "grad_norm": 2.1368465423583984, | |
| "learning_rate": 0.0007939473235110074, | |
| "loss": 2.0233, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.028443028613686784, | |
| "grad_norm": 2.7280213832855225, | |
| "learning_rate": 0.0007924303619849442, | |
| "loss": 1.9639, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03413163433642414, | |
| "grad_norm": 2.5257463455200195, | |
| "learning_rate": 0.0007909134004588809, | |
| "loss": 2.0229, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0398202400591615, | |
| "grad_norm": 2.748051404953003, | |
| "learning_rate": 0.0007893964389328175, | |
| "loss": 1.9867, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04550884578189886, | |
| "grad_norm": 2.212047815322876, | |
| "learning_rate": 0.0007878794774067543, | |
| "loss": 1.9354, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.05119745150463621, | |
| "grad_norm": 2.423400640487671, | |
| "learning_rate": 0.000786362515880691, | |
| "loss": 1.9127, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.05688605722737357, | |
| "grad_norm": 2.379678726196289, | |
| "learning_rate": 0.0007848455543546277, | |
| "loss": 1.8927, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06257466295011092, | |
| "grad_norm": 2.5806541442871094, | |
| "learning_rate": 0.0007833285928285645, | |
| "loss": 1.8471, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.06826326867284828, | |
| "grad_norm": 2.4539499282836914, | |
| "learning_rate": 0.0007818116313025011, | |
| "loss": 1.8296, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.07395187439558565, | |
| "grad_norm": 2.7818546295166016, | |
| "learning_rate": 0.0007802946697764378, | |
| "loss": 1.8136, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.079640480118323, | |
| "grad_norm": 2.979959487915039, | |
| "learning_rate": 0.0007787777082503746, | |
| "loss": 1.7844, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.08532908584106036, | |
| "grad_norm": 2.3885879516601562, | |
| "learning_rate": 0.0007772607467243113, | |
| "loss": 1.7627, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09101769156379771, | |
| "grad_norm": 2.2447025775909424, | |
| "learning_rate": 0.0007757437851982479, | |
| "loss": 1.7375, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.09670629728653507, | |
| "grad_norm": 2.184580087661743, | |
| "learning_rate": 0.0007742268236721846, | |
| "loss": 1.7212, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.10239490300927243, | |
| "grad_norm": 2.2506866455078125, | |
| "learning_rate": 0.0007727098621461213, | |
| "loss": 1.7264, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.10808350873200978, | |
| "grad_norm": 2.059812068939209, | |
| "learning_rate": 0.000771192900620058, | |
| "loss": 1.7123, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.11377211445474714, | |
| "grad_norm": 2.3013007640838623, | |
| "learning_rate": 0.0007696759390939948, | |
| "loss": 1.7122, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.11946072017748449, | |
| "grad_norm": 2.7073047161102295, | |
| "learning_rate": 0.0007681741471831921, | |
| "loss": 1.6847, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.12514932590022185, | |
| "grad_norm": 2.023949384689331, | |
| "learning_rate": 0.0007666571856571288, | |
| "loss": 1.6941, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.1308379316229592, | |
| "grad_norm": 1.9444501399993896, | |
| "learning_rate": 0.0007651402241310656, | |
| "loss": 1.6682, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.13652653734569656, | |
| "grad_norm": 2.691826105117798, | |
| "learning_rate": 0.0007636232626050023, | |
| "loss": 1.6409, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.1422151430684339, | |
| "grad_norm": 2.483386993408203, | |
| "learning_rate": 0.0007621063010789389, | |
| "loss": 1.662, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.1479037487911713, | |
| "grad_norm": 2.1850545406341553, | |
| "learning_rate": 0.0007605893395528756, | |
| "loss": 1.6532, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.15359235451390865, | |
| "grad_norm": 1.989560842514038, | |
| "learning_rate": 0.0007590723780268123, | |
| "loss": 1.6231, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.159280960236646, | |
| "grad_norm": 2.1362531185150146, | |
| "learning_rate": 0.000757555416500749, | |
| "loss": 1.6019, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.16496956595938336, | |
| "grad_norm": 2.3262641429901123, | |
| "learning_rate": 0.0007560384549746857, | |
| "loss": 1.6103, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.17065817168212072, | |
| "grad_norm": 2.297419309616089, | |
| "learning_rate": 0.0007545214934486224, | |
| "loss": 1.6314, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.17634677740485807, | |
| "grad_norm": 2.1368629932403564, | |
| "learning_rate": 0.0007530045319225591, | |
| "loss": 1.5838, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.18203538312759543, | |
| "grad_norm": 2.3383195400238037, | |
| "learning_rate": 0.0007514875703964959, | |
| "loss": 1.5857, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.18772398885033278, | |
| "grad_norm": 2.149740219116211, | |
| "learning_rate": 0.0007499706088704326, | |
| "loss": 1.6016, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.19341259457307014, | |
| "grad_norm": 2.096703290939331, | |
| "learning_rate": 0.0007484536473443692, | |
| "loss": 1.5904, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.1991012002958075, | |
| "grad_norm": 2.2043957710266113, | |
| "learning_rate": 0.000746936685818306, | |
| "loss": 1.5787, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.20478980601854485, | |
| "grad_norm": 2.6369898319244385, | |
| "learning_rate": 0.0007454197242922427, | |
| "loss": 1.5539, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.2104784117412822, | |
| "grad_norm": 1.9776628017425537, | |
| "learning_rate": 0.0007439027627661794, | |
| "loss": 1.5815, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.21616701746401956, | |
| "grad_norm": 2.2001795768737793, | |
| "learning_rate": 0.0007423858012401161, | |
| "loss": 1.5583, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.22185562318675692, | |
| "grad_norm": 2.2252562046051025, | |
| "learning_rate": 0.0007408688397140527, | |
| "loss": 1.5539, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.22754422890949427, | |
| "grad_norm": 2.15871262550354, | |
| "learning_rate": 0.0007393518781879895, | |
| "loss": 1.5786, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.23323283463223163, | |
| "grad_norm": 2.026066303253174, | |
| "learning_rate": 0.0007378500862771869, | |
| "loss": 1.5452, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.23892144035496898, | |
| "grad_norm": 2.116511583328247, | |
| "learning_rate": 0.0007363331247511235, | |
| "loss": 1.5381, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.24461004607770637, | |
| "grad_norm": 1.9454152584075928, | |
| "learning_rate": 0.0007348161632250602, | |
| "loss": 1.557, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.2502986518004437, | |
| "grad_norm": 1.8668495416641235, | |
| "learning_rate": 0.000733299201698997, | |
| "loss": 1.5406, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.2559872575231811, | |
| "grad_norm": 2.0886125564575195, | |
| "learning_rate": 0.0007317822401729337, | |
| "loss": 1.519, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.2616758632459184, | |
| "grad_norm": 2.5768446922302246, | |
| "learning_rate": 0.0007302652786468704, | |
| "loss": 1.5178, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.2673644689686558, | |
| "grad_norm": 2.4169631004333496, | |
| "learning_rate": 0.000728748317120807, | |
| "loss": 1.5283, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.2730530746913931, | |
| "grad_norm": 2.7676291465759277, | |
| "learning_rate": 0.0007272313555947437, | |
| "loss": 1.5265, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.2787416804141305, | |
| "grad_norm": 1.9152452945709229, | |
| "learning_rate": 0.0007257143940686805, | |
| "loss": 1.5152, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.2844302861368678, | |
| "grad_norm": 2.56608510017395, | |
| "learning_rate": 0.0007241974325426172, | |
| "loss": 1.5084, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.2844302861368678, | |
| "eval_accuracy": 0.636444, | |
| "eval_loss": 1.469992995262146, | |
| "eval_runtime": 85.5907, | |
| "eval_samples_per_second": 2920.879, | |
| "eval_steps_per_second": 11.415, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.2901188918596052, | |
| "grad_norm": 2.2403135299682617, | |
| "learning_rate": 0.0007226804710165538, | |
| "loss": 1.5076, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.2958074975823426, | |
| "grad_norm": 2.058535099029541, | |
| "learning_rate": 0.0007211635094904906, | |
| "loss": 1.4973, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.3014961033050799, | |
| "grad_norm": 1.9374159574508667, | |
| "learning_rate": 0.0007196465479644273, | |
| "loss": 1.5055, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.3071847090278173, | |
| "grad_norm": 1.8894695043563843, | |
| "learning_rate": 0.000718129586438364, | |
| "loss": 1.4996, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.31287331475055463, | |
| "grad_norm": 2.5466501712799072, | |
| "learning_rate": 0.0007166126249123008, | |
| "loss": 1.5006, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.318561920473292, | |
| "grad_norm": 1.9721605777740479, | |
| "learning_rate": 0.0007150956633862374, | |
| "loss": 1.4932, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.32425052619602934, | |
| "grad_norm": 1.8921763896942139, | |
| "learning_rate": 0.0007135787018601741, | |
| "loss": 1.4762, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.3299391319187667, | |
| "grad_norm": 2.49052357673645, | |
| "learning_rate": 0.0007120617403341108, | |
| "loss": 1.4636, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.33562773764150405, | |
| "grad_norm": 1.8825891017913818, | |
| "learning_rate": 0.0007105447788080475, | |
| "loss": 1.4903, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.34131634336424144, | |
| "grad_norm": 1.9227776527404785, | |
| "learning_rate": 0.0007090278172819842, | |
| "loss": 1.4602, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.34700494908697876, | |
| "grad_norm": 2.173774242401123, | |
| "learning_rate": 0.0007075108557559209, | |
| "loss": 1.4581, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.35269355480971615, | |
| "grad_norm": 1.9840656518936157, | |
| "learning_rate": 0.0007059938942298576, | |
| "loss": 1.4504, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.3583821605324535, | |
| "grad_norm": 2.368171453475952, | |
| "learning_rate": 0.000704492102319055, | |
| "loss": 1.4659, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.36407076625519086, | |
| "grad_norm": 2.005125045776367, | |
| "learning_rate": 0.0007029751407929917, | |
| "loss": 1.4698, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.3697593719779282, | |
| "grad_norm": 1.8724095821380615, | |
| "learning_rate": 0.0007014581792669284, | |
| "loss": 1.4429, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.37544797770066557, | |
| "grad_norm": 1.8412431478500366, | |
| "learning_rate": 0.0006999412177408651, | |
| "loss": 1.4368, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.3811365834234029, | |
| "grad_norm": 1.9016755819320679, | |
| "learning_rate": 0.0006984242562148018, | |
| "loss": 1.4351, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.3868251891461403, | |
| "grad_norm": 1.9896953105926514, | |
| "learning_rate": 0.0006969072946887384, | |
| "loss": 1.4563, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.39251379486887766, | |
| "grad_norm": 2.3341548442840576, | |
| "learning_rate": 0.0006953903331626751, | |
| "loss": 1.4457, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.398202400591615, | |
| "grad_norm": 1.95259690284729, | |
| "learning_rate": 0.0006938733716366119, | |
| "loss": 1.4636, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.4038910063143524, | |
| "grad_norm": 1.8444461822509766, | |
| "learning_rate": 0.0006923564101105486, | |
| "loss": 1.4418, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.4095796120370897, | |
| "grad_norm": 1.9170624017715454, | |
| "learning_rate": 0.0006908394485844853, | |
| "loss": 1.4267, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.4152682177598271, | |
| "grad_norm": 1.6293827295303345, | |
| "learning_rate": 0.000689322487058422, | |
| "loss": 1.4474, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.4209568234825644, | |
| "grad_norm": 2.2202467918395996, | |
| "learning_rate": 0.0006878055255323587, | |
| "loss": 1.4166, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.4266454292053018, | |
| "grad_norm": 1.9069397449493408, | |
| "learning_rate": 0.0006862885640062954, | |
| "loss": 1.4194, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.4323340349280391, | |
| "grad_norm": 2.0205297470092773, | |
| "learning_rate": 0.0006847716024802322, | |
| "loss": 1.4328, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.4380226406507765, | |
| "grad_norm": 1.6736252307891846, | |
| "learning_rate": 0.0006832546409541689, | |
| "loss": 1.4309, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.44371124637351383, | |
| "grad_norm": 1.7010937929153442, | |
| "learning_rate": 0.0006817376794281055, | |
| "loss": 1.412, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.4493998520962512, | |
| "grad_norm": 2.748424768447876, | |
| "learning_rate": 0.0006802207179020422, | |
| "loss": 1.421, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.45508845781898855, | |
| "grad_norm": 1.908728837966919, | |
| "learning_rate": 0.0006787037563759789, | |
| "loss": 1.4172, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.46077706354172593, | |
| "grad_norm": 1.8672014474868774, | |
| "learning_rate": 0.0006771867948499157, | |
| "loss": 1.4448, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.46646566926446326, | |
| "grad_norm": 2.128519058227539, | |
| "learning_rate": 0.0006756698333238524, | |
| "loss": 1.4158, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.47215427498720064, | |
| "grad_norm": 1.7498713731765747, | |
| "learning_rate": 0.000674152871797789, | |
| "loss": 1.412, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.47784288070993797, | |
| "grad_norm": 1.7801289558410645, | |
| "learning_rate": 0.0006726510798869864, | |
| "loss": 1.4146, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.48353148643267535, | |
| "grad_norm": 1.9360538721084595, | |
| "learning_rate": 0.0006711341183609232, | |
| "loss": 1.4252, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.48922009215541273, | |
| "grad_norm": 2.3669304847717285, | |
| "learning_rate": 0.0006696171568348598, | |
| "loss": 1.4057, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.49490869787815006, | |
| "grad_norm": 1.7751379013061523, | |
| "learning_rate": 0.0006681001953087965, | |
| "loss": 1.4049, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.5005973036008874, | |
| "grad_norm": 2.5389885902404785, | |
| "learning_rate": 0.0006665832337827332, | |
| "loss": 1.3837, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.5062859093236248, | |
| "grad_norm": 2.5082690715789795, | |
| "learning_rate": 0.0006650662722566699, | |
| "loss": 1.3924, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.5119745150463622, | |
| "grad_norm": 2.011589527130127, | |
| "learning_rate": 0.0006635493107306066, | |
| "loss": 1.3956, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.5176631207690995, | |
| "grad_norm": 1.819793939590454, | |
| "learning_rate": 0.0006620323492045433, | |
| "loss": 1.4063, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.5233517264918368, | |
| "grad_norm": 2.081247568130493, | |
| "learning_rate": 0.00066051538767848, | |
| "loss": 1.4145, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.5290403322145742, | |
| "grad_norm": 2.151563882827759, | |
| "learning_rate": 0.0006589984261524168, | |
| "loss": 1.4002, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.5347289379373116, | |
| "grad_norm": 1.9170759916305542, | |
| "learning_rate": 0.0006574814646263535, | |
| "loss": 1.3863, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.5404175436600489, | |
| "grad_norm": 1.6435186862945557, | |
| "learning_rate": 0.0006559645031002901, | |
| "loss": 1.3888, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.5461061493827862, | |
| "grad_norm": 1.8130972385406494, | |
| "learning_rate": 0.0006544475415742268, | |
| "loss": 1.3904, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.5517947551055237, | |
| "grad_norm": 1.8200345039367676, | |
| "learning_rate": 0.0006529305800481636, | |
| "loss": 1.3647, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.557483360828261, | |
| "grad_norm": 1.7286423444747925, | |
| "learning_rate": 0.0006514136185221003, | |
| "loss": 1.3815, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.5631719665509983, | |
| "grad_norm": 2.345879554748535, | |
| "learning_rate": 0.000649896656996037, | |
| "loss": 1.3919, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.5688605722737357, | |
| "grad_norm": 1.8189209699630737, | |
| "learning_rate": 0.0006483796954699736, | |
| "loss": 1.3684, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.5688605722737357, | |
| "eval_accuracy": 0.667392, | |
| "eval_loss": 1.3353288173675537, | |
| "eval_runtime": 85.0475, | |
| "eval_samples_per_second": 2939.533, | |
| "eval_steps_per_second": 11.488, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.5745491779964731, | |
| "grad_norm": 1.7264429330825806, | |
| "learning_rate": 0.0006468627339439103, | |
| "loss": 1.4044, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.5802377837192104, | |
| "grad_norm": 1.8806540966033936, | |
| "learning_rate": 0.0006453457724178471, | |
| "loss": 1.3746, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.5859263894419477, | |
| "grad_norm": 1.7714815139770508, | |
| "learning_rate": 0.0006438288108917838, | |
| "loss": 1.3837, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.5916149951646852, | |
| "grad_norm": 1.713157057762146, | |
| "learning_rate": 0.0006423118493657205, | |
| "loss": 1.3939, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.5973036008874225, | |
| "grad_norm": 2.169168472290039, | |
| "learning_rate": 0.0006408100574549179, | |
| "loss": 1.3658, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.6029922066101598, | |
| "grad_norm": 1.727501630783081, | |
| "learning_rate": 0.0006392930959288546, | |
| "loss": 1.3907, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.6086808123328972, | |
| "grad_norm": 2.0120322704315186, | |
| "learning_rate": 0.0006377761344027913, | |
| "loss": 1.3757, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.6143694180556346, | |
| "grad_norm": 1.799139142036438, | |
| "learning_rate": 0.0006362591728767279, | |
| "loss": 1.3803, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.6200580237783719, | |
| "grad_norm": 1.8817808628082275, | |
| "learning_rate": 0.0006347422113506646, | |
| "loss": 1.3702, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.6257466295011093, | |
| "grad_norm": 2.1144518852233887, | |
| "learning_rate": 0.0006332252498246013, | |
| "loss": 1.3832, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.6314352352238466, | |
| "grad_norm": 2.1396071910858154, | |
| "learning_rate": 0.0006317082882985381, | |
| "loss": 1.3611, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.637123840946584, | |
| "grad_norm": 1.6794757843017578, | |
| "learning_rate": 0.0006301913267724747, | |
| "loss": 1.368, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.6428124466693214, | |
| "grad_norm": 2.268433094024658, | |
| "learning_rate": 0.0006286743652464114, | |
| "loss": 1.3498, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.6485010523920587, | |
| "grad_norm": 1.8515706062316895, | |
| "learning_rate": 0.0006271574037203482, | |
| "loss": 1.3489, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.654189658114796, | |
| "grad_norm": 2.482171058654785, | |
| "learning_rate": 0.0006256404421942849, | |
| "loss": 1.3501, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.6598782638375335, | |
| "grad_norm": 1.9485667943954468, | |
| "learning_rate": 0.0006241234806682216, | |
| "loss": 1.3483, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.6655668695602708, | |
| "grad_norm": 1.8601367473602295, | |
| "learning_rate": 0.0006226065191421583, | |
| "loss": 1.3392, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.6712554752830081, | |
| "grad_norm": 1.870851993560791, | |
| "learning_rate": 0.000621089557616095, | |
| "loss": 1.352, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.6769440810057455, | |
| "grad_norm": 1.9454014301300049, | |
| "learning_rate": 0.0006195725960900317, | |
| "loss": 1.3537, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.6826326867284829, | |
| "grad_norm": 1.9180669784545898, | |
| "learning_rate": 0.0006180556345639685, | |
| "loss": 1.3541, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.6883212924512202, | |
| "grad_norm": 1.7796809673309326, | |
| "learning_rate": 0.0006165386730379051, | |
| "loss": 1.331, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.6940098981739575, | |
| "grad_norm": 2.040998935699463, | |
| "learning_rate": 0.0006150217115118417, | |
| "loss": 1.3214, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.699698503896695, | |
| "grad_norm": 1.7188791036605835, | |
| "learning_rate": 0.0006135047499857785, | |
| "loss": 1.3577, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.7053871096194323, | |
| "grad_norm": 1.9152625799179077, | |
| "learning_rate": 0.0006119877884597152, | |
| "loss": 1.3682, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.7110757153421696, | |
| "grad_norm": 2.150810718536377, | |
| "learning_rate": 0.0006104708269336519, | |
| "loss": 1.3388, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.716764321064907, | |
| "grad_norm": 1.97470223903656, | |
| "learning_rate": 0.0006089538654075887, | |
| "loss": 1.3319, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 0.7224529267876444, | |
| "grad_norm": 1.663122296333313, | |
| "learning_rate": 0.0006074369038815253, | |
| "loss": 1.3593, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 0.7281415325103817, | |
| "grad_norm": 1.6453677415847778, | |
| "learning_rate": 0.0006059351119707227, | |
| "loss": 1.3592, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 0.733830138233119, | |
| "grad_norm": 1.6896419525146484, | |
| "learning_rate": 0.0006044181504446595, | |
| "loss": 1.3183, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 0.7395187439558564, | |
| "grad_norm": 1.7903008460998535, | |
| "learning_rate": 0.000602901188918596, | |
| "loss": 1.3373, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.7452073496785938, | |
| "grad_norm": 2.2026655673980713, | |
| "learning_rate": 0.0006013842273925327, | |
| "loss": 1.3403, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 0.7508959554013311, | |
| "grad_norm": 1.9204201698303223, | |
| "learning_rate": 0.0005998672658664695, | |
| "loss": 1.3199, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 0.7565845611240685, | |
| "grad_norm": 1.946899652481079, | |
| "learning_rate": 0.0005983503043404062, | |
| "loss": 1.3298, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 0.7622731668468058, | |
| "grad_norm": 2.019131898880005, | |
| "learning_rate": 0.0005968333428143428, | |
| "loss": 1.3449, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 0.7679617725695432, | |
| "grad_norm": 1.848008155822754, | |
| "learning_rate": 0.0005953163812882796, | |
| "loss": 1.3206, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.7736503782922806, | |
| "grad_norm": 2.373288631439209, | |
| "learning_rate": 0.0005937994197622163, | |
| "loss": 1.3564, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 0.7793389840150179, | |
| "grad_norm": 2.556985855102539, | |
| "learning_rate": 0.000592282458236153, | |
| "loss": 1.3254, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 0.7850275897377553, | |
| "grad_norm": 1.8957433700561523, | |
| "learning_rate": 0.0005907654967100898, | |
| "loss": 1.3498, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 0.7907161954604927, | |
| "grad_norm": 1.7315127849578857, | |
| "learning_rate": 0.0005892485351840264, | |
| "loss": 1.3249, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 0.79640480118323, | |
| "grad_norm": 1.973764419555664, | |
| "learning_rate": 0.0005877315736579631, | |
| "loss": 1.3305, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.8020934069059673, | |
| "grad_norm": 1.711145281791687, | |
| "learning_rate": 0.0005862146121318999, | |
| "loss": 1.3011, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 0.8077820126287047, | |
| "grad_norm": 1.8515042066574097, | |
| "learning_rate": 0.0005846976506058365, | |
| "loss": 1.3195, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 0.8134706183514421, | |
| "grad_norm": 1.6278700828552246, | |
| "learning_rate": 0.0005831806890797733, | |
| "loss": 1.3308, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 0.8191592240741794, | |
| "grad_norm": 1.444455623626709, | |
| "learning_rate": 0.0005816637275537099, | |
| "loss": 1.3119, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 0.8248478297969167, | |
| "grad_norm": 1.6277796030044556, | |
| "learning_rate": 0.0005801467660276466, | |
| "loss": 1.3227, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.8305364355196542, | |
| "grad_norm": 1.8428665399551392, | |
| "learning_rate": 0.0005786298045015834, | |
| "loss": 1.3336, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 0.8362250412423915, | |
| "grad_norm": 1.6377763748168945, | |
| "learning_rate": 0.0005771128429755201, | |
| "loss": 1.3141, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 0.8419136469651288, | |
| "grad_norm": 1.7305645942687988, | |
| "learning_rate": 0.0005755958814494568, | |
| "loss": 1.3062, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 0.8476022526878662, | |
| "grad_norm": 2.469701051712036, | |
| "learning_rate": 0.0005740940895386541, | |
| "loss": 1.3074, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 0.8532908584106036, | |
| "grad_norm": 1.952755331993103, | |
| "learning_rate": 0.0005725771280125909, | |
| "loss": 1.3568, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.8532908584106036, | |
| "eval_accuracy": 0.68038, | |
| "eval_loss": 1.2764052152633667, | |
| "eval_runtime": 82.5452, | |
| "eval_samples_per_second": 3028.643, | |
| "eval_steps_per_second": 11.836, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.8589794641333409, | |
| "grad_norm": 3.221471071243286, | |
| "learning_rate": 0.0005710601664865274, | |
| "loss": 1.3341, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 0.8646680698560782, | |
| "grad_norm": 2.2455317974090576, | |
| "learning_rate": 0.0005695432049604642, | |
| "loss": 1.3276, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 0.8703566755788157, | |
| "grad_norm": 1.8076684474945068, | |
| "learning_rate": 0.0005680262434344009, | |
| "loss": 1.2922, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 0.876045281301553, | |
| "grad_norm": 1.701774001121521, | |
| "learning_rate": 0.0005665092819083376, | |
| "loss": 1.3003, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 0.8817338870242903, | |
| "grad_norm": 1.5403673648834229, | |
| "learning_rate": 0.0005649923203822744, | |
| "loss": 1.3207, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.8874224927470277, | |
| "grad_norm": 1.9462639093399048, | |
| "learning_rate": 0.000563475358856211, | |
| "loss": 1.3098, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 0.8931110984697651, | |
| "grad_norm": 1.6688456535339355, | |
| "learning_rate": 0.0005619583973301477, | |
| "loss": 1.2993, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 0.8987997041925024, | |
| "grad_norm": 1.6060837507247925, | |
| "learning_rate": 0.0005604414358040845, | |
| "loss": 1.3145, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 0.9044883099152398, | |
| "grad_norm": 1.8593111038208008, | |
| "learning_rate": 0.0005589244742780212, | |
| "loss": 1.2836, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 0.9101769156379771, | |
| "grad_norm": 2.035261869430542, | |
| "learning_rate": 0.0005574075127519579, | |
| "loss": 1.3125, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.9158655213607145, | |
| "grad_norm": 1.6091046333312988, | |
| "learning_rate": 0.0005558905512258946, | |
| "loss": 1.2868, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 0.9215541270834519, | |
| "grad_norm": 1.656204104423523, | |
| "learning_rate": 0.0005543735896998313, | |
| "loss": 1.3075, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 0.9272427328061892, | |
| "grad_norm": 1.5555946826934814, | |
| "learning_rate": 0.0005528566281737679, | |
| "loss": 1.2963, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 0.9329313385289265, | |
| "grad_norm": 1.7379626035690308, | |
| "learning_rate": 0.0005513396666477047, | |
| "loss": 1.2905, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 0.938619944251664, | |
| "grad_norm": 1.5103166103363037, | |
| "learning_rate": 0.0005498227051216414, | |
| "loss": 1.2848, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.9443085499744013, | |
| "grad_norm": 1.5895978212356567, | |
| "learning_rate": 0.000548305743595578, | |
| "loss": 1.293, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 0.9499971556971386, | |
| "grad_norm": 1.6526978015899658, | |
| "learning_rate": 0.0005467887820695148, | |
| "loss": 1.288, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 0.9556857614198759, | |
| "grad_norm": 1.7471717596054077, | |
| "learning_rate": 0.0005452718205434515, | |
| "loss": 1.3099, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 0.9613743671426134, | |
| "grad_norm": 1.5995450019836426, | |
| "learning_rate": 0.0005437700286326488, | |
| "loss": 1.2886, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 0.9670629728653507, | |
| "grad_norm": 1.7462047338485718, | |
| "learning_rate": 0.0005422530671065856, | |
| "loss": 1.317, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.972751578588088, | |
| "grad_norm": 1.5739308595657349, | |
| "learning_rate": 0.0005407361055805223, | |
| "loss": 1.2997, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 0.9784401843108255, | |
| "grad_norm": 1.6608139276504517, | |
| "learning_rate": 0.0005392191440544589, | |
| "loss": 1.3037, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 0.9841287900335628, | |
| "grad_norm": 1.7515637874603271, | |
| "learning_rate": 0.0005377021825283956, | |
| "loss": 1.302, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 0.9898173957563001, | |
| "grad_norm": 1.572986364364624, | |
| "learning_rate": 0.0005361852210023323, | |
| "loss": 1.2945, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 0.9955060014790375, | |
| "grad_norm": 1.9207016229629517, | |
| "learning_rate": 0.000534668259476269, | |
| "loss": 1.2747, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.0011946072017748, | |
| "grad_norm": 1.9010945558547974, | |
| "learning_rate": 0.0005331512979502058, | |
| "loss": 1.263, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 1.0068832129245122, | |
| "grad_norm": 2.4259393215179443, | |
| "learning_rate": 0.0005316343364241425, | |
| "loss": 1.2741, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 1.0125718186472497, | |
| "grad_norm": 2.5002028942108154, | |
| "learning_rate": 0.0005301173748980791, | |
| "loss": 1.2686, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 1.0182604243699869, | |
| "grad_norm": 1.7075704336166382, | |
| "learning_rate": 0.0005286004133720159, | |
| "loss": 1.2661, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 1.0239490300927243, | |
| "grad_norm": 1.7390458583831787, | |
| "learning_rate": 0.0005270834518459526, | |
| "loss": 1.2698, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.0296376358154615, | |
| "grad_norm": 1.980185627937317, | |
| "learning_rate": 0.0005255664903198893, | |
| "loss": 1.2569, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 1.035326241538199, | |
| "grad_norm": 1.79970383644104, | |
| "learning_rate": 0.0005240495287938261, | |
| "loss": 1.2738, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 1.0410148472609364, | |
| "grad_norm": 1.6184749603271484, | |
| "learning_rate": 0.0005225325672677627, | |
| "loss": 1.2637, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 1.0467034529836736, | |
| "grad_norm": 2.3463358879089355, | |
| "learning_rate": 0.0005210156057416993, | |
| "loss": 1.2665, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 1.052392058706411, | |
| "grad_norm": 1.8550745248794556, | |
| "learning_rate": 0.0005194986442156361, | |
| "loss": 1.2664, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.0580806644291485, | |
| "grad_norm": 1.8582580089569092, | |
| "learning_rate": 0.0005179816826895728, | |
| "loss": 1.2442, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 1.0637692701518857, | |
| "grad_norm": 1.88007390499115, | |
| "learning_rate": 0.0005164647211635095, | |
| "loss": 1.2536, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 1.0694578758746232, | |
| "grad_norm": 1.804671287536621, | |
| "learning_rate": 0.0005149477596374462, | |
| "loss": 1.2459, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 1.0751464815973604, | |
| "grad_norm": 1.7329107522964478, | |
| "learning_rate": 0.0005134307981113829, | |
| "loss": 1.2499, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 1.0808350873200978, | |
| "grad_norm": 1.693323016166687, | |
| "learning_rate": 0.0005119290062005802, | |
| "loss": 1.25, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.0865236930428352, | |
| "grad_norm": 1.600060224533081, | |
| "learning_rate": 0.000510412044674517, | |
| "loss": 1.2515, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 1.0922122987655725, | |
| "grad_norm": 1.8084614276885986, | |
| "learning_rate": 0.0005088950831484537, | |
| "loss": 1.246, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 1.09790090448831, | |
| "grad_norm": 1.8022205829620361, | |
| "learning_rate": 0.0005073781216223904, | |
| "loss": 1.2597, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 1.1035895102110473, | |
| "grad_norm": 1.6137562990188599, | |
| "learning_rate": 0.0005058611600963271, | |
| "loss": 1.2685, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 1.1092781159337846, | |
| "grad_norm": 1.7756201028823853, | |
| "learning_rate": 0.0005043441985702637, | |
| "loss": 1.2606, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.114966721656522, | |
| "grad_norm": 1.8828805685043335, | |
| "learning_rate": 0.0005028272370442004, | |
| "loss": 1.2582, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 1.1206553273792594, | |
| "grad_norm": 1.6829185485839844, | |
| "learning_rate": 0.0005013102755181372, | |
| "loss": 1.2563, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 1.1263439331019967, | |
| "grad_norm": 1.6716195344924927, | |
| "learning_rate": 0.0004997933139920739, | |
| "loss": 1.2405, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 1.132032538824734, | |
| "grad_norm": 1.7629872560501099, | |
| "learning_rate": 0.0004982763524660106, | |
| "loss": 1.2649, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 1.1377211445474713, | |
| "grad_norm": 1.704967737197876, | |
| "learning_rate": 0.0004967593909399473, | |
| "loss": 1.226, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.1377211445474713, | |
| "eval_accuracy": 0.692396, | |
| "eval_loss": 1.2322564125061035, | |
| "eval_runtime": 82.0826, | |
| "eval_samples_per_second": 3045.712, | |
| "eval_steps_per_second": 11.903, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.1434097502702087, | |
| "grad_norm": 1.5182781219482422, | |
| "learning_rate": 0.000495242429413884, | |
| "loss": 1.2343, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 1.1490983559929462, | |
| "grad_norm": 2.637796640396118, | |
| "learning_rate": 0.0004937254678878207, | |
| "loss": 1.2506, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 1.1547869617156834, | |
| "grad_norm": 1.8955748081207275, | |
| "learning_rate": 0.0004922085063617575, | |
| "loss": 1.2625, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 1.1604755674384208, | |
| "grad_norm": 2.0370640754699707, | |
| "learning_rate": 0.0004906915448356942, | |
| "loss": 1.2551, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 1.1661641731611583, | |
| "grad_norm": 1.8047020435333252, | |
| "learning_rate": 0.0004891745833096308, | |
| "loss": 1.2489, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.1718527788838955, | |
| "grad_norm": 1.5440089702606201, | |
| "learning_rate": 0.0004876576217835675, | |
| "loss": 1.2646, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 1.177541384606633, | |
| "grad_norm": 1.5029830932617188, | |
| "learning_rate": 0.00048614066025750425, | |
| "loss": 1.2536, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 1.1832299903293704, | |
| "grad_norm": 1.4674205780029297, | |
| "learning_rate": 0.0004846236987314409, | |
| "loss": 1.2457, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 1.1889185960521076, | |
| "grad_norm": 1.5259037017822266, | |
| "learning_rate": 0.00048310673720537765, | |
| "loss": 1.2449, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 1.194607201774845, | |
| "grad_norm": 1.6339012384414673, | |
| "learning_rate": 0.0004815897756793144, | |
| "loss": 1.2163, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.2002958074975822, | |
| "grad_norm": 1.5565885305404663, | |
| "learning_rate": 0.00048007281415325106, | |
| "loss": 1.2461, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 1.2059844132203197, | |
| "grad_norm": 1.676540493965149, | |
| "learning_rate": 0.0004785558526271878, | |
| "loss": 1.2555, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 1.2116730189430571, | |
| "grad_norm": 1.6003342866897583, | |
| "learning_rate": 0.00047705406071638513, | |
| "loss": 1.2507, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 1.2173616246657943, | |
| "grad_norm": 2.2655630111694336, | |
| "learning_rate": 0.00047553709919032175, | |
| "loss": 1.2144, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 1.2230502303885318, | |
| "grad_norm": 1.695094108581543, | |
| "learning_rate": 0.0004740201376642585, | |
| "loss": 1.2415, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.2287388361112692, | |
| "grad_norm": 1.8387731313705444, | |
| "learning_rate": 0.0004725031761381952, | |
| "loss": 1.2406, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 1.2344274418340064, | |
| "grad_norm": 1.6776598691940308, | |
| "learning_rate": 0.0004709862146121319, | |
| "loss": 1.2673, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 1.2401160475567439, | |
| "grad_norm": 1.6573506593704224, | |
| "learning_rate": 0.0004694692530860686, | |
| "loss": 1.2587, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 1.2458046532794813, | |
| "grad_norm": 1.6786317825317383, | |
| "learning_rate": 0.00046795229156000535, | |
| "loss": 1.2464, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 1.2514932590022185, | |
| "grad_norm": 1.887971043586731, | |
| "learning_rate": 0.00046643533003394203, | |
| "loss": 1.2501, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.257181864724956, | |
| "grad_norm": 1.7499343156814575, | |
| "learning_rate": 0.00046491836850787876, | |
| "loss": 1.2296, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 1.2628704704476932, | |
| "grad_norm": 2.057670831680298, | |
| "learning_rate": 0.00046340140698181544, | |
| "loss": 1.2346, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 1.2685590761704306, | |
| "grad_norm": 1.7353135347366333, | |
| "learning_rate": 0.00046188444545575217, | |
| "loss": 1.2512, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 1.274247681893168, | |
| "grad_norm": 2.0662734508514404, | |
| "learning_rate": 0.0004603674839296889, | |
| "loss": 1.2493, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 1.2799362876159053, | |
| "grad_norm": 1.5519914627075195, | |
| "learning_rate": 0.0004588505224036255, | |
| "loss": 1.2404, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.2856248933386427, | |
| "grad_norm": 1.8667906522750854, | |
| "learning_rate": 0.0004573335608775622, | |
| "loss": 1.247, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 1.29131349906138, | |
| "grad_norm": 1.8621453046798706, | |
| "learning_rate": 0.00045581659935149893, | |
| "loss": 1.2428, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 1.2970021047841174, | |
| "grad_norm": 1.7203937768936157, | |
| "learning_rate": 0.00045429963782543566, | |
| "loss": 1.2273, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 1.3026907105068548, | |
| "grad_norm": 1.7497667074203491, | |
| "learning_rate": 0.00045278267629937234, | |
| "loss": 1.2458, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 1.3083793162295922, | |
| "grad_norm": 2.057507276535034, | |
| "learning_rate": 0.0004512657147733091, | |
| "loss": 1.2325, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.3140679219523295, | |
| "grad_norm": 1.4594337940216064, | |
| "learning_rate": 0.0004497487532472458, | |
| "loss": 1.2319, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 1.319756527675067, | |
| "grad_norm": 2.1696736812591553, | |
| "learning_rate": 0.0004482317917211825, | |
| "loss": 1.234, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 1.3254451333978041, | |
| "grad_norm": 1.8165019750595093, | |
| "learning_rate": 0.0004467148301951192, | |
| "loss": 1.2256, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 1.3311337391205416, | |
| "grad_norm": 1.5531728267669678, | |
| "learning_rate": 0.00044519786866905594, | |
| "loss": 1.2518, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 1.336822344843279, | |
| "grad_norm": 1.4592831134796143, | |
| "learning_rate": 0.0004436809071429926, | |
| "loss": 1.2192, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.3425109505660162, | |
| "grad_norm": 1.74478280544281, | |
| "learning_rate": 0.00044216394561692935, | |
| "loss": 1.2427, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 1.3481995562887537, | |
| "grad_norm": 1.8685113191604614, | |
| "learning_rate": 0.000440646984090866, | |
| "loss": 1.2581, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 1.3538881620114909, | |
| "grad_norm": 1.7366535663604736, | |
| "learning_rate": 0.0004391300225648027, | |
| "loss": 1.2471, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 1.3595767677342283, | |
| "grad_norm": 1.6585444211959839, | |
| "learning_rate": 0.0004376130610387394, | |
| "loss": 1.2198, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 1.3652653734569657, | |
| "grad_norm": 1.9299806356430054, | |
| "learning_rate": 0.0004360960995126761, | |
| "loss": 1.2314, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.3709539791797032, | |
| "grad_norm": 1.8172481060028076, | |
| "learning_rate": 0.00043457913798661285, | |
| "loss": 1.2098, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 1.3766425849024404, | |
| "grad_norm": 1.5579493045806885, | |
| "learning_rate": 0.0004330621764605495, | |
| "loss": 1.2043, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 1.3823311906251778, | |
| "grad_norm": 1.8178203105926514, | |
| "learning_rate": 0.00043154521493448625, | |
| "loss": 1.2235, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 1.388019796347915, | |
| "grad_norm": 1.676126480102539, | |
| "learning_rate": 0.000430028253408423, | |
| "loss": 1.2388, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 1.3937084020706525, | |
| "grad_norm": 1.863893985748291, | |
| "learning_rate": 0.00042851129188235966, | |
| "loss": 1.2206, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.39939700779339, | |
| "grad_norm": 1.5618318319320679, | |
| "learning_rate": 0.0004269943303562964, | |
| "loss": 1.2359, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 1.4050856135161272, | |
| "grad_norm": 1.3972681760787964, | |
| "learning_rate": 0.0004254773688302331, | |
| "loss": 1.2152, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 1.4107742192388646, | |
| "grad_norm": 1.584274411201477, | |
| "learning_rate": 0.00042396040730416975, | |
| "loss": 1.211, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 1.4164628249616018, | |
| "grad_norm": 1.7282276153564453, | |
| "learning_rate": 0.0004224434457781064, | |
| "loss": 1.2034, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 1.4221514306843392, | |
| "grad_norm": 2.2420654296875, | |
| "learning_rate": 0.00042092648425204316, | |
| "loss": 1.2125, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.4221514306843392, | |
| "eval_accuracy": 0.703072, | |
| "eval_loss": 1.185011863708496, | |
| "eval_runtime": 81.7158, | |
| "eval_samples_per_second": 3059.385, | |
| "eval_steps_per_second": 11.956, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.4278400364070767, | |
| "grad_norm": 1.5998648405075073, | |
| "learning_rate": 0.0004194095227259799, | |
| "loss": 1.2145, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 1.433528642129814, | |
| "grad_norm": 1.8546173572540283, | |
| "learning_rate": 0.00041789256119991656, | |
| "loss": 1.2293, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 1.4392172478525513, | |
| "grad_norm": 1.6815022230148315, | |
| "learning_rate": 0.00041639076928911395, | |
| "loss": 1.2165, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 1.4449058535752888, | |
| "grad_norm": 1.5567988157272339, | |
| "learning_rate": 0.0004148889773783113, | |
| "loss": 1.2231, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 1.450594459298026, | |
| "grad_norm": 1.9424728155136108, | |
| "learning_rate": 0.000413372015852248, | |
| "loss": 1.2302, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.4562830650207634, | |
| "grad_norm": 1.9052510261535645, | |
| "learning_rate": 0.00041185505432618464, | |
| "loss": 1.2174, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 1.4619716707435009, | |
| "grad_norm": 1.470513939857483, | |
| "learning_rate": 0.0004103380928001213, | |
| "loss": 1.2167, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 1.467660276466238, | |
| "grad_norm": 1.5082899332046509, | |
| "learning_rate": 0.00040882113127405805, | |
| "loss": 1.2098, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 1.4733488821889755, | |
| "grad_norm": 1.7309447526931763, | |
| "learning_rate": 0.0004073041697479948, | |
| "loss": 1.2111, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 1.4790374879117127, | |
| "grad_norm": 1.7894470691680908, | |
| "learning_rate": 0.00040578720822193146, | |
| "loss": 1.2237, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.4847260936344502, | |
| "grad_norm": 1.712557077407837, | |
| "learning_rate": 0.0004042702466958682, | |
| "loss": 1.2314, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 1.4904146993571876, | |
| "grad_norm": 1.794565200805664, | |
| "learning_rate": 0.0004027532851698049, | |
| "loss": 1.2072, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 1.4961033050799248, | |
| "grad_norm": 1.655179500579834, | |
| "learning_rate": 0.0004012363236437416, | |
| "loss": 1.218, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 1.5017919108026623, | |
| "grad_norm": 1.8749343156814575, | |
| "learning_rate": 0.00039971936211767833, | |
| "loss": 1.2037, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 1.5074805165253995, | |
| "grad_norm": 1.5337320566177368, | |
| "learning_rate": 0.000398202400591615, | |
| "loss": 1.2179, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.513169122248137, | |
| "grad_norm": 1.5731686353683472, | |
| "learning_rate": 0.0003966854390655517, | |
| "loss": 1.2037, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 1.5188577279708744, | |
| "grad_norm": 1.5700329542160034, | |
| "learning_rate": 0.0003951684775394884, | |
| "loss": 1.2189, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 1.5245463336936118, | |
| "grad_norm": 1.9315118789672852, | |
| "learning_rate": 0.00039365151601342515, | |
| "loss": 1.2314, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 1.530234939416349, | |
| "grad_norm": 1.6017844676971436, | |
| "learning_rate": 0.0003921345544873618, | |
| "loss": 1.211, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 1.5359235451390862, | |
| "grad_norm": 1.586595058441162, | |
| "learning_rate": 0.0003906175929612985, | |
| "loss": 1.2079, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.5416121508618237, | |
| "grad_norm": 1.8215593099594116, | |
| "learning_rate": 0.00038910063143523523, | |
| "loss": 1.2022, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 1.5473007565845611, | |
| "grad_norm": 1.7390124797821045, | |
| "learning_rate": 0.00038758366990917196, | |
| "loss": 1.2143, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 1.5529893623072986, | |
| "grad_norm": 1.792608618736267, | |
| "learning_rate": 0.00038606670838310864, | |
| "loss": 1.2104, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 1.558677968030036, | |
| "grad_norm": 1.802167296409607, | |
| "learning_rate": 0.00038454974685704537, | |
| "loss": 1.1924, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 1.5643665737527732, | |
| "grad_norm": 1.7943332195281982, | |
| "learning_rate": 0.0003830327853309821, | |
| "loss": 1.2096, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.5700551794755104, | |
| "grad_norm": 1.745893120765686, | |
| "learning_rate": 0.0003815158238049187, | |
| "loss": 1.1941, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 1.5757437851982479, | |
| "grad_norm": 1.6740118265151978, | |
| "learning_rate": 0.00037999886227885546, | |
| "loss": 1.2355, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 1.5814323909209853, | |
| "grad_norm": 1.681840419769287, | |
| "learning_rate": 0.0003784970703680528, | |
| "loss": 1.2034, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 1.5871209966437227, | |
| "grad_norm": 1.6897751092910767, | |
| "learning_rate": 0.0003769801088419895, | |
| "loss": 1.2169, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 1.59280960236646, | |
| "grad_norm": 1.686784267425537, | |
| "learning_rate": 0.0003754631473159262, | |
| "loss": 1.2093, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.5984982080891972, | |
| "grad_norm": 1.6020421981811523, | |
| "learning_rate": 0.00037394618578986293, | |
| "loss": 1.2131, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 1.6041868138119346, | |
| "grad_norm": 1.478246808052063, | |
| "learning_rate": 0.0003724292242637996, | |
| "loss": 1.2048, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 1.609875419534672, | |
| "grad_norm": 1.4912410974502563, | |
| "learning_rate": 0.00037091226273773634, | |
| "loss": 1.18, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 1.6155640252574095, | |
| "grad_norm": 1.6362539529800415, | |
| "learning_rate": 0.00036939530121167307, | |
| "loss": 1.2022, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 1.6212526309801467, | |
| "grad_norm": 1.5238479375839233, | |
| "learning_rate": 0.00036787833968560975, | |
| "loss": 1.2105, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.6269412367028842, | |
| "grad_norm": 1.6359635591506958, | |
| "learning_rate": 0.0003663613781595464, | |
| "loss": 1.1833, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 1.6326298424256214, | |
| "grad_norm": 1.6206257343292236, | |
| "learning_rate": 0.00036484441663348316, | |
| "loss": 1.1945, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 1.6383184481483588, | |
| "grad_norm": 1.7032015323638916, | |
| "learning_rate": 0.00036332745510741983, | |
| "loss": 1.2065, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 1.6440070538710962, | |
| "grad_norm": 1.7177228927612305, | |
| "learning_rate": 0.00036181049358135657, | |
| "loss": 1.2035, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 1.6496956595938337, | |
| "grad_norm": 1.5967752933502197, | |
| "learning_rate": 0.0003602935320552933, | |
| "loss": 1.2036, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.655384265316571, | |
| "grad_norm": 1.6632803678512573, | |
| "learning_rate": 0.00035877657052923, | |
| "loss": 1.226, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 1.6610728710393081, | |
| "grad_norm": 1.5134357213974, | |
| "learning_rate": 0.00035725960900316665, | |
| "loss": 1.1947, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 1.6667614767620456, | |
| "grad_norm": 1.5506322383880615, | |
| "learning_rate": 0.0003557426474771034, | |
| "loss": 1.1963, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 1.672450082484783, | |
| "grad_norm": 1.4821183681488037, | |
| "learning_rate": 0.0003542256859510401, | |
| "loss": 1.2039, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 1.6781386882075204, | |
| "grad_norm": 2.278379440307617, | |
| "learning_rate": 0.0003527087244249768, | |
| "loss": 1.205, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.6838272939302577, | |
| "grad_norm": 1.5077921152114868, | |
| "learning_rate": 0.0003511917628989135, | |
| "loss": 1.1984, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 1.689515899652995, | |
| "grad_norm": 1.629607915878296, | |
| "learning_rate": 0.0003496748013728502, | |
| "loss": 1.2023, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 1.6952045053757323, | |
| "grad_norm": 1.5007668733596802, | |
| "learning_rate": 0.0003481578398467869, | |
| "loss": 1.1907, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 1.7008931110984697, | |
| "grad_norm": 1.7543882131576538, | |
| "learning_rate": 0.0003466408783207236, | |
| "loss": 1.1949, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 1.7065817168212072, | |
| "grad_norm": 1.6254594326019287, | |
| "learning_rate": 0.00034512391679466034, | |
| "loss": 1.1912, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.7065817168212072, | |
| "eval_accuracy": 0.709248, | |
| "eval_loss": 1.1566522121429443, | |
| "eval_runtime": 79.5399, | |
| "eval_samples_per_second": 3143.077, | |
| "eval_steps_per_second": 12.283, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.7122703225439446, | |
| "grad_norm": 1.873049020767212, | |
| "learning_rate": 0.000343606955268597, | |
| "loss": 1.2063, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 1.7179589282666818, | |
| "grad_norm": 1.5862141847610474, | |
| "learning_rate": 0.00034208999374253375, | |
| "loss": 1.1926, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 1.723647533989419, | |
| "grad_norm": 1.9915696382522583, | |
| "learning_rate": 0.0003405730322164704, | |
| "loss": 1.1952, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 1.7293361397121565, | |
| "grad_norm": 1.856048822402954, | |
| "learning_rate": 0.0003390560706904071, | |
| "loss": 1.1953, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 1.735024745434894, | |
| "grad_norm": 1.6758267879486084, | |
| "learning_rate": 0.00033753910916434383, | |
| "loss": 1.1906, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.7407133511576314, | |
| "grad_norm": 1.8683140277862549, | |
| "learning_rate": 0.00033602214763828056, | |
| "loss": 1.2025, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 1.7464019568803686, | |
| "grad_norm": 1.452721118927002, | |
| "learning_rate": 0.00033450518611221724, | |
| "loss": 1.1866, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 1.752090562603106, | |
| "grad_norm": 1.5711089372634888, | |
| "learning_rate": 0.0003329882245861539, | |
| "loss": 1.1856, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 1.7577791683258432, | |
| "grad_norm": 2.0584185123443604, | |
| "learning_rate": 0.00033147126306009065, | |
| "loss": 1.1873, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 1.7634677740485807, | |
| "grad_norm": 1.5743275880813599, | |
| "learning_rate": 0.0003299543015340274, | |
| "loss": 1.1988, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.7691563797713181, | |
| "grad_norm": 1.5788936614990234, | |
| "learning_rate": 0.00032843734000796406, | |
| "loss": 1.1932, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 1.7748449854940556, | |
| "grad_norm": 1.6406651735305786, | |
| "learning_rate": 0.0003269203784819008, | |
| "loss": 1.1876, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 1.7805335912167928, | |
| "grad_norm": 1.6410019397735596, | |
| "learning_rate": 0.00032540341695583747, | |
| "loss": 1.1859, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 1.78622219693953, | |
| "grad_norm": 1.548140287399292, | |
| "learning_rate": 0.00032388645542977414, | |
| "loss": 1.2032, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 1.7919108026622674, | |
| "grad_norm": 1.9242947101593018, | |
| "learning_rate": 0.0003223694939037109, | |
| "loss": 1.199, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.7975994083850049, | |
| "grad_norm": 2.0189428329467773, | |
| "learning_rate": 0.0003208525323776476, | |
| "loss": 1.1832, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 1.8032880141077423, | |
| "grad_norm": 1.740432620048523, | |
| "learning_rate": 0.0003193355708515843, | |
| "loss": 1.183, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 1.8089766198304795, | |
| "grad_norm": 1.743503451347351, | |
| "learning_rate": 0.0003178337789407816, | |
| "loss": 1.1991, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 1.8146652255532167, | |
| "grad_norm": 1.7166736125946045, | |
| "learning_rate": 0.00031631681741471835, | |
| "loss": 1.2031, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 1.8203538312759542, | |
| "grad_norm": 1.626386046409607, | |
| "learning_rate": 0.000314799855888655, | |
| "loss": 1.1987, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.8260424369986916, | |
| "grad_norm": 1.5402131080627441, | |
| "learning_rate": 0.00031328289436259176, | |
| "loss": 1.172, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 1.831731042721429, | |
| "grad_norm": 1.6522256135940552, | |
| "learning_rate": 0.0003117659328365285, | |
| "loss": 1.179, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 1.8374196484441665, | |
| "grad_norm": 1.482009768486023, | |
| "learning_rate": 0.00031024897131046517, | |
| "loss": 1.1903, | |
| "step": 32300 | |
| }, | |
| { | |
| "epoch": 1.8431082541669037, | |
| "grad_norm": 1.6417380571365356, | |
| "learning_rate": 0.00030873200978440184, | |
| "loss": 1.2024, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 1.848796859889641, | |
| "grad_norm": 1.532333493232727, | |
| "learning_rate": 0.0003072150482583386, | |
| "loss": 1.1843, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.8544854656123784, | |
| "grad_norm": 2.004293441772461, | |
| "learning_rate": 0.00030569808673227525, | |
| "loss": 1.192, | |
| "step": 32600 | |
| }, | |
| { | |
| "epoch": 1.8601740713351158, | |
| "grad_norm": 1.7226125001907349, | |
| "learning_rate": 0.000304181125206212, | |
| "loss": 1.1902, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 1.8658626770578532, | |
| "grad_norm": 1.7714165449142456, | |
| "learning_rate": 0.0003026641636801487, | |
| "loss": 1.1908, | |
| "step": 32800 | |
| }, | |
| { | |
| "epoch": 1.8715512827805905, | |
| "grad_norm": 1.5100337266921997, | |
| "learning_rate": 0.00030114720215408534, | |
| "loss": 1.1735, | |
| "step": 32900 | |
| }, | |
| { | |
| "epoch": 1.8772398885033277, | |
| "grad_norm": 1.6792744398117065, | |
| "learning_rate": 0.00029963024062802207, | |
| "loss": 1.191, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.8829284942260651, | |
| "grad_norm": 1.705554723739624, | |
| "learning_rate": 0.0002981132791019588, | |
| "loss": 1.1878, | |
| "step": 33100 | |
| }, | |
| { | |
| "epoch": 1.8886170999488026, | |
| "grad_norm": 1.4528917074203491, | |
| "learning_rate": 0.0002965963175758955, | |
| "loss": 1.1685, | |
| "step": 33200 | |
| }, | |
| { | |
| "epoch": 1.89430570567154, | |
| "grad_norm": 1.7752711772918701, | |
| "learning_rate": 0.0002950793560498322, | |
| "loss": 1.1743, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 1.8999943113942772, | |
| "grad_norm": 1.762074589729309, | |
| "learning_rate": 0.00029356239452376894, | |
| "loss": 1.1775, | |
| "step": 33400 | |
| }, | |
| { | |
| "epoch": 1.9056829171170147, | |
| "grad_norm": 1.6388828754425049, | |
| "learning_rate": 0.0002920454329977056, | |
| "loss": 1.1762, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.9113715228397519, | |
| "grad_norm": 1.5171791315078735, | |
| "learning_rate": 0.0002905284714716423, | |
| "loss": 1.1649, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 1.9170601285624893, | |
| "grad_norm": 1.6547460556030273, | |
| "learning_rate": 0.000289011509945579, | |
| "loss": 1.1904, | |
| "step": 33700 | |
| }, | |
| { | |
| "epoch": 1.9227487342852267, | |
| "grad_norm": 1.705083966255188, | |
| "learning_rate": 0.00028750971803477636, | |
| "loss": 1.1667, | |
| "step": 33800 | |
| }, | |
| { | |
| "epoch": 1.9284373400079642, | |
| "grad_norm": 1.731803059577942, | |
| "learning_rate": 0.00028599275650871304, | |
| "loss": 1.1788, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 1.9341259457307014, | |
| "grad_norm": 2.056766986846924, | |
| "learning_rate": 0.00028447579498264977, | |
| "loss": 1.1878, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.9398145514534386, | |
| "grad_norm": 1.8016914129257202, | |
| "learning_rate": 0.00028295883345658644, | |
| "loss": 1.1632, | |
| "step": 34100 | |
| }, | |
| { | |
| "epoch": 1.945503157176176, | |
| "grad_norm": 1.7706475257873535, | |
| "learning_rate": 0.0002814418719305232, | |
| "loss": 1.1658, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 1.9511917628989135, | |
| "grad_norm": 1.8184970617294312, | |
| "learning_rate": 0.0002799249104044599, | |
| "loss": 1.1666, | |
| "step": 34300 | |
| }, | |
| { | |
| "epoch": 1.956880368621651, | |
| "grad_norm": 1.6529743671417236, | |
| "learning_rate": 0.0002784079488783966, | |
| "loss": 1.1846, | |
| "step": 34400 | |
| }, | |
| { | |
| "epoch": 1.9625689743443882, | |
| "grad_norm": 1.5860931873321533, | |
| "learning_rate": 0.00027689098735233326, | |
| "loss": 1.1917, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.9682575800671256, | |
| "grad_norm": 1.672756552696228, | |
| "learning_rate": 0.00027537402582627, | |
| "loss": 1.1654, | |
| "step": 34600 | |
| }, | |
| { | |
| "epoch": 1.9739461857898628, | |
| "grad_norm": 1.7606583833694458, | |
| "learning_rate": 0.0002738570643002067, | |
| "loss": 1.176, | |
| "step": 34700 | |
| }, | |
| { | |
| "epoch": 1.9796347915126002, | |
| "grad_norm": 1.912277340888977, | |
| "learning_rate": 0.0002723401027741434, | |
| "loss": 1.1695, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 1.9853233972353377, | |
| "grad_norm": 1.7096484899520874, | |
| "learning_rate": 0.00027082314124808013, | |
| "loss": 1.1669, | |
| "step": 34900 | |
| }, | |
| { | |
| "epoch": 1.9910120029580751, | |
| "grad_norm": 1.7793241739273071, | |
| "learning_rate": 0.0002693061797220168, | |
| "loss": 1.1902, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.9910120029580751, | |
| "eval_accuracy": 0.71648, | |
| "eval_loss": 1.1297262907028198, | |
| "eval_runtime": 79.5966, | |
| "eval_samples_per_second": 3140.839, | |
| "eval_steps_per_second": 12.274, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.9967006086808123, | |
| "grad_norm": 1.4913907051086426, | |
| "learning_rate": 0.0002677892181959535, | |
| "loss": 1.1607, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 2.0023892144035496, | |
| "grad_norm": 1.639985203742981, | |
| "learning_rate": 0.0002662874262851509, | |
| "loss": 1.1727, | |
| "step": 35200 | |
| }, | |
| { | |
| "epoch": 2.008077820126287, | |
| "grad_norm": 1.6419970989227295, | |
| "learning_rate": 0.00026477046475908755, | |
| "loss": 1.1392, | |
| "step": 35300 | |
| }, | |
| { | |
| "epoch": 2.0137664258490244, | |
| "grad_norm": 1.8132672309875488, | |
| "learning_rate": 0.00026325350323302423, | |
| "loss": 1.1503, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 2.019455031571762, | |
| "grad_norm": 1.4656819105148315, | |
| "learning_rate": 0.00026173654170696096, | |
| "loss": 1.1565, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 2.0251436372944993, | |
| "grad_norm": 1.3595716953277588, | |
| "learning_rate": 0.0002602195801808977, | |
| "loss": 1.1526, | |
| "step": 35600 | |
| }, | |
| { | |
| "epoch": 2.0308322430172363, | |
| "grad_norm": 1.6904360055923462, | |
| "learning_rate": 0.00025870261865483437, | |
| "loss": 1.1448, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 2.0365208487399737, | |
| "grad_norm": 1.7240209579467773, | |
| "learning_rate": 0.0002571856571287711, | |
| "loss": 1.1424, | |
| "step": 35800 | |
| }, | |
| { | |
| "epoch": 2.042209454462711, | |
| "grad_norm": 1.5376731157302856, | |
| "learning_rate": 0.00025566869560270783, | |
| "loss": 1.143, | |
| "step": 35900 | |
| }, | |
| { | |
| "epoch": 2.0478980601854486, | |
| "grad_norm": 1.893202781677246, | |
| "learning_rate": 0.00025415173407664445, | |
| "loss": 1.1519, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.053586665908186, | |
| "grad_norm": 1.9057211875915527, | |
| "learning_rate": 0.0002526347725505812, | |
| "loss": 1.1375, | |
| "step": 36100 | |
| }, | |
| { | |
| "epoch": 2.059275271630923, | |
| "grad_norm": 1.7818187475204468, | |
| "learning_rate": 0.0002511178110245179, | |
| "loss": 1.1424, | |
| "step": 36200 | |
| }, | |
| { | |
| "epoch": 2.0649638773536605, | |
| "grad_norm": 1.825323462486267, | |
| "learning_rate": 0.0002496008494984546, | |
| "loss": 1.1196, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 2.070652483076398, | |
| "grad_norm": 2.0049736499786377, | |
| "learning_rate": 0.0002480838879723913, | |
| "loss": 1.1317, | |
| "step": 36400 | |
| }, | |
| { | |
| "epoch": 2.0763410887991354, | |
| "grad_norm": 1.599846363067627, | |
| "learning_rate": 0.000246566926446328, | |
| "loss": 1.1573, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 2.082029694521873, | |
| "grad_norm": 1.5434855222702026, | |
| "learning_rate": 0.00024504996492026473, | |
| "loss": 1.1493, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 2.0877183002446102, | |
| "grad_norm": 1.6306787729263306, | |
| "learning_rate": 0.0002435330033942014, | |
| "loss": 1.1564, | |
| "step": 36700 | |
| }, | |
| { | |
| "epoch": 2.0934069059673472, | |
| "grad_norm": 1.6914353370666504, | |
| "learning_rate": 0.00024201604186813814, | |
| "loss": 1.1395, | |
| "step": 36800 | |
| }, | |
| { | |
| "epoch": 2.0990955116900847, | |
| "grad_norm": 1.6444432735443115, | |
| "learning_rate": 0.00024049908034207485, | |
| "loss": 1.1615, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 2.104784117412822, | |
| "grad_norm": 1.821244239807129, | |
| "learning_rate": 0.00023898211881601155, | |
| "loss": 1.1429, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.1104727231355596, | |
| "grad_norm": 1.6050491333007812, | |
| "learning_rate": 0.00023746515728994823, | |
| "loss": 1.1376, | |
| "step": 37100 | |
| }, | |
| { | |
| "epoch": 2.116161328858297, | |
| "grad_norm": 1.7375249862670898, | |
| "learning_rate": 0.00023594819576388493, | |
| "loss": 1.1253, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 2.121849934581034, | |
| "grad_norm": 2.0717177391052246, | |
| "learning_rate": 0.00023443123423782166, | |
| "loss": 1.1527, | |
| "step": 37300 | |
| }, | |
| { | |
| "epoch": 2.1275385403037714, | |
| "grad_norm": 1.43324875831604, | |
| "learning_rate": 0.00023291427271175837, | |
| "loss": 1.1475, | |
| "step": 37400 | |
| }, | |
| { | |
| "epoch": 2.133227146026509, | |
| "grad_norm": 1.448669195175171, | |
| "learning_rate": 0.00023139731118569507, | |
| "loss": 1.1133, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 2.1389157517492463, | |
| "grad_norm": 1.521912932395935, | |
| "learning_rate": 0.00022988034965963178, | |
| "loss": 1.1292, | |
| "step": 37600 | |
| }, | |
| { | |
| "epoch": 2.1446043574719837, | |
| "grad_norm": 1.6070728302001953, | |
| "learning_rate": 0.00022836338813356845, | |
| "loss": 1.1384, | |
| "step": 37700 | |
| }, | |
| { | |
| "epoch": 2.1502929631947207, | |
| "grad_norm": 1.3853884935379028, | |
| "learning_rate": 0.00022684642660750516, | |
| "loss": 1.1344, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 2.155981568917458, | |
| "grad_norm": 1.569415807723999, | |
| "learning_rate": 0.0002253294650814419, | |
| "loss": 1.1572, | |
| "step": 37900 | |
| }, | |
| { | |
| "epoch": 2.1616701746401956, | |
| "grad_norm": 1.544966220855713, | |
| "learning_rate": 0.0002238125035553786, | |
| "loss": 1.1378, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.167358780362933, | |
| "grad_norm": 1.6090420484542847, | |
| "learning_rate": 0.0002222955420293153, | |
| "loss": 1.1331, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 2.1730473860856705, | |
| "grad_norm": 1.542605996131897, | |
| "learning_rate": 0.00022077858050325197, | |
| "loss": 1.1302, | |
| "step": 38200 | |
| }, | |
| { | |
| "epoch": 2.178735991808408, | |
| "grad_norm": 1.744084119796753, | |
| "learning_rate": 0.00021926161897718868, | |
| "loss": 1.1305, | |
| "step": 38300 | |
| }, | |
| { | |
| "epoch": 2.184424597531145, | |
| "grad_norm": 1.630118489265442, | |
| "learning_rate": 0.0002177446574511254, | |
| "loss": 1.1294, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 2.1901132032538824, | |
| "grad_norm": 1.6920104026794434, | |
| "learning_rate": 0.0002162276959250621, | |
| "loss": 1.1337, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 2.19580180897662, | |
| "grad_norm": 1.654189944267273, | |
| "learning_rate": 0.00021471073439899882, | |
| "loss": 1.1182, | |
| "step": 38600 | |
| }, | |
| { | |
| "epoch": 2.2014904146993572, | |
| "grad_norm": 1.8575996160507202, | |
| "learning_rate": 0.00021319377287293555, | |
| "loss": 1.1261, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 2.2071790204220947, | |
| "grad_norm": 1.5796535015106201, | |
| "learning_rate": 0.0002116768113468722, | |
| "loss": 1.1389, | |
| "step": 38800 | |
| }, | |
| { | |
| "epoch": 2.212867626144832, | |
| "grad_norm": 1.6893657445907593, | |
| "learning_rate": 0.00021015984982080893, | |
| "loss": 1.122, | |
| "step": 38900 | |
| }, | |
| { | |
| "epoch": 2.218556231867569, | |
| "grad_norm": 1.5983092784881592, | |
| "learning_rate": 0.00020864288829474563, | |
| "loss": 1.1487, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.2242448375903066, | |
| "grad_norm": 1.632049798965454, | |
| "learning_rate": 0.00020712592676868234, | |
| "loss": 1.1476, | |
| "step": 39100 | |
| }, | |
| { | |
| "epoch": 2.229933443313044, | |
| "grad_norm": 2.039854049682617, | |
| "learning_rate": 0.00020562413485787965, | |
| "loss": 1.1443, | |
| "step": 39200 | |
| }, | |
| { | |
| "epoch": 2.2356220490357814, | |
| "grad_norm": 1.5673627853393555, | |
| "learning_rate": 0.00020410717333181638, | |
| "loss": 1.1259, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 2.241310654758519, | |
| "grad_norm": 1.6900497674942017, | |
| "learning_rate": 0.00020259021180575308, | |
| "loss": 1.1356, | |
| "step": 39400 | |
| }, | |
| { | |
| "epoch": 2.246999260481256, | |
| "grad_norm": 1.8306878805160522, | |
| "learning_rate": 0.00020107325027968979, | |
| "loss": 1.1349, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 2.2526878662039933, | |
| "grad_norm": 1.620490550994873, | |
| "learning_rate": 0.0001995562887536265, | |
| "loss": 1.1417, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 2.2583764719267307, | |
| "grad_norm": 1.828751802444458, | |
| "learning_rate": 0.0001980393272275632, | |
| "loss": 1.1302, | |
| "step": 39700 | |
| }, | |
| { | |
| "epoch": 2.264065077649468, | |
| "grad_norm": 1.4963942766189575, | |
| "learning_rate": 0.0001965223657014999, | |
| "loss": 1.152, | |
| "step": 39800 | |
| }, | |
| { | |
| "epoch": 2.2697536833722056, | |
| "grad_norm": 2.081669807434082, | |
| "learning_rate": 0.0001950054041754366, | |
| "loss": 1.1385, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 2.2754422890949426, | |
| "grad_norm": 1.6873656511306763, | |
| "learning_rate": 0.0001934884426493733, | |
| "loss": 1.131, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.2754422890949426, | |
| "eval_accuracy": 0.721316, | |
| "eval_loss": 1.1105972528457642, | |
| "eval_runtime": 80.6985, | |
| "eval_samples_per_second": 3097.949, | |
| "eval_steps_per_second": 12.107, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.28113089481768, | |
| "grad_norm": 1.5599457025527954, | |
| "learning_rate": 0.00019197148112331004, | |
| "loss": 1.1525, | |
| "step": 40100 | |
| }, | |
| { | |
| "epoch": 2.2868195005404175, | |
| "grad_norm": 1.816628098487854, | |
| "learning_rate": 0.00019045451959724672, | |
| "loss": 1.1295, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 2.292508106263155, | |
| "grad_norm": 1.5481749773025513, | |
| "learning_rate": 0.00018893755807118342, | |
| "loss": 1.124, | |
| "step": 40300 | |
| }, | |
| { | |
| "epoch": 2.2981967119858924, | |
| "grad_norm": 1.632873296737671, | |
| "learning_rate": 0.00018742059654512015, | |
| "loss": 1.1217, | |
| "step": 40400 | |
| }, | |
| { | |
| "epoch": 2.3038853177086294, | |
| "grad_norm": 1.4403363466262817, | |
| "learning_rate": 0.00018590363501905683, | |
| "loss": 1.1315, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 2.309573923431367, | |
| "grad_norm": 1.6744205951690674, | |
| "learning_rate": 0.00018438667349299353, | |
| "loss": 1.1473, | |
| "step": 40600 | |
| }, | |
| { | |
| "epoch": 2.3152625291541042, | |
| "grad_norm": 1.5021002292633057, | |
| "learning_rate": 0.00018286971196693026, | |
| "loss": 1.1127, | |
| "step": 40700 | |
| }, | |
| { | |
| "epoch": 2.3209511348768417, | |
| "grad_norm": 1.689931869506836, | |
| "learning_rate": 0.00018135275044086694, | |
| "loss": 1.1394, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 2.326639740599579, | |
| "grad_norm": 2.1370577812194824, | |
| "learning_rate": 0.00017983578891480367, | |
| "loss": 1.148, | |
| "step": 40900 | |
| }, | |
| { | |
| "epoch": 2.3323283463223166, | |
| "grad_norm": 1.9048566818237305, | |
| "learning_rate": 0.00017831882738874038, | |
| "loss": 1.1181, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 2.338016952045054, | |
| "grad_norm": 1.8328748941421509, | |
| "learning_rate": 0.00017680186586267705, | |
| "loss": 1.1302, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 2.343705557767791, | |
| "grad_norm": 1.7709869146347046, | |
| "learning_rate": 0.0001753000739518744, | |
| "loss": 1.1369, | |
| "step": 41200 | |
| }, | |
| { | |
| "epoch": 2.3493941634905284, | |
| "grad_norm": 1.6296570301055908, | |
| "learning_rate": 0.00017378311242581112, | |
| "loss": 1.1302, | |
| "step": 41300 | |
| }, | |
| { | |
| "epoch": 2.355082769213266, | |
| "grad_norm": 1.6044236421585083, | |
| "learning_rate": 0.0001722661508997478, | |
| "loss": 1.1313, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 2.3607713749360033, | |
| "grad_norm": 1.4571659564971924, | |
| "learning_rate": 0.00017074918937368453, | |
| "loss": 1.1249, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 2.3664599806587407, | |
| "grad_norm": 1.7237457036972046, | |
| "learning_rate": 0.00016923222784762123, | |
| "loss": 1.1312, | |
| "step": 41600 | |
| }, | |
| { | |
| "epoch": 2.3721485863814777, | |
| "grad_norm": 1.552881121635437, | |
| "learning_rate": 0.0001677152663215579, | |
| "loss": 1.1282, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 2.377837192104215, | |
| "grad_norm": 1.6091784238815308, | |
| "learning_rate": 0.00016619830479549464, | |
| "loss": 1.1236, | |
| "step": 41800 | |
| }, | |
| { | |
| "epoch": 2.3835257978269526, | |
| "grad_norm": 1.8620885610580444, | |
| "learning_rate": 0.00016468134326943134, | |
| "loss": 1.1469, | |
| "step": 41900 | |
| }, | |
| { | |
| "epoch": 2.38921440354969, | |
| "grad_norm": 1.717551827430725, | |
| "learning_rate": 0.00016316438174336802, | |
| "loss": 1.121, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.3949030092724275, | |
| "grad_norm": 1.6212184429168701, | |
| "learning_rate": 0.00016164742021730475, | |
| "loss": 1.0997, | |
| "step": 42100 | |
| }, | |
| { | |
| "epoch": 2.4005916149951645, | |
| "grad_norm": 1.3878498077392578, | |
| "learning_rate": 0.00016013045869124146, | |
| "loss": 1.1362, | |
| "step": 42200 | |
| }, | |
| { | |
| "epoch": 2.406280220717902, | |
| "grad_norm": 1.6336196660995483, | |
| "learning_rate": 0.00015861349716517816, | |
| "loss": 1.1256, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 2.4119688264406394, | |
| "grad_norm": 1.7155201435089111, | |
| "learning_rate": 0.00015709653563911486, | |
| "loss": 1.1133, | |
| "step": 42400 | |
| }, | |
| { | |
| "epoch": 2.417657432163377, | |
| "grad_norm": 1.7675564289093018, | |
| "learning_rate": 0.00015557957411305157, | |
| "loss": 1.1416, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 2.4233460378861142, | |
| "grad_norm": 1.676527976989746, | |
| "learning_rate": 0.00015406261258698827, | |
| "loss": 1.1378, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 2.4290346436088512, | |
| "grad_norm": 1.6293052434921265, | |
| "learning_rate": 0.00015254565106092498, | |
| "loss": 1.1177, | |
| "step": 42700 | |
| }, | |
| { | |
| "epoch": 2.4347232493315887, | |
| "grad_norm": 1.5264780521392822, | |
| "learning_rate": 0.00015102868953486168, | |
| "loss": 1.1063, | |
| "step": 42800 | |
| }, | |
| { | |
| "epoch": 2.440411855054326, | |
| "grad_norm": 1.6453486680984497, | |
| "learning_rate": 0.00014951172800879839, | |
| "loss": 1.1375, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 2.4461004607770636, | |
| "grad_norm": 1.692336082458496, | |
| "learning_rate": 0.0001479947664827351, | |
| "loss": 1.1004, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 2.451789066499801, | |
| "grad_norm": 1.868812084197998, | |
| "learning_rate": 0.0001464778049566718, | |
| "loss": 1.1288, | |
| "step": 43100 | |
| }, | |
| { | |
| "epoch": 2.4574776722225384, | |
| "grad_norm": 1.7713991403579712, | |
| "learning_rate": 0.0001449608434306085, | |
| "loss": 1.1229, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 2.4631662779452754, | |
| "grad_norm": 1.6394290924072266, | |
| "learning_rate": 0.00014345905151980583, | |
| "loss": 1.0968, | |
| "step": 43300 | |
| }, | |
| { | |
| "epoch": 2.468854883668013, | |
| "grad_norm": 1.7240723371505737, | |
| "learning_rate": 0.00014194208999374254, | |
| "loss": 1.1151, | |
| "step": 43400 | |
| }, | |
| { | |
| "epoch": 2.4745434893907503, | |
| "grad_norm": 1.9284464120864868, | |
| "learning_rate": 0.00014042512846767924, | |
| "loss": 1.1302, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 2.4802320951134877, | |
| "grad_norm": 1.6855792999267578, | |
| "learning_rate": 0.00013890816694161595, | |
| "loss": 1.1163, | |
| "step": 43600 | |
| }, | |
| { | |
| "epoch": 2.485920700836225, | |
| "grad_norm": 1.8182587623596191, | |
| "learning_rate": 0.00013739120541555265, | |
| "loss": 1.1172, | |
| "step": 43700 | |
| }, | |
| { | |
| "epoch": 2.4916093065589626, | |
| "grad_norm": 1.5971157550811768, | |
| "learning_rate": 0.00013587424388948935, | |
| "loss": 1.1071, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 2.4972979122816996, | |
| "grad_norm": 1.7139756679534912, | |
| "learning_rate": 0.00013435728236342606, | |
| "loss": 1.1239, | |
| "step": 43900 | |
| }, | |
| { | |
| "epoch": 2.502986518004437, | |
| "grad_norm": 1.7199363708496094, | |
| "learning_rate": 0.00013284032083736276, | |
| "loss": 1.1444, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.5086751237271745, | |
| "grad_norm": 1.7295994758605957, | |
| "learning_rate": 0.00013132335931129947, | |
| "loss": 1.122, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 2.514363729449912, | |
| "grad_norm": 1.9433492422103882, | |
| "learning_rate": 0.00012980639778523617, | |
| "loss": 1.1209, | |
| "step": 44200 | |
| }, | |
| { | |
| "epoch": 2.5200523351726494, | |
| "grad_norm": 1.5811411142349243, | |
| "learning_rate": 0.0001282894362591729, | |
| "loss": 1.1084, | |
| "step": 44300 | |
| }, | |
| { | |
| "epoch": 2.5257409408953864, | |
| "grad_norm": 1.5232020616531372, | |
| "learning_rate": 0.00012677247473310958, | |
| "loss": 1.1372, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 2.531429546618124, | |
| "grad_norm": 2.6212551593780518, | |
| "learning_rate": 0.00012525551320704628, | |
| "loss": 1.1246, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 2.5371181523408612, | |
| "grad_norm": 1.4962718486785889, | |
| "learning_rate": 0.00012373855168098301, | |
| "loss": 1.1386, | |
| "step": 44600 | |
| }, | |
| { | |
| "epoch": 2.5428067580635987, | |
| "grad_norm": 1.7713087797164917, | |
| "learning_rate": 0.0001222215901549197, | |
| "loss": 1.1314, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 2.548495363786336, | |
| "grad_norm": 1.5493218898773193, | |
| "learning_rate": 0.00012070462862885641, | |
| "loss": 1.1204, | |
| "step": 44800 | |
| }, | |
| { | |
| "epoch": 2.554183969509073, | |
| "grad_norm": 1.6126313209533691, | |
| "learning_rate": 0.00011918766710279313, | |
| "loss": 1.1283, | |
| "step": 44900 | |
| }, | |
| { | |
| "epoch": 2.5598725752318106, | |
| "grad_norm": 1.5327433347702026, | |
| "learning_rate": 0.00011767070557672982, | |
| "loss": 1.124, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.5598725752318106, | |
| "eval_accuracy": 0.725824, | |
| "eval_loss": 1.0916061401367188, | |
| "eval_runtime": 80.2149, | |
| "eval_samples_per_second": 3116.626, | |
| "eval_steps_per_second": 12.18, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.565561180954548, | |
| "grad_norm": 1.5026576519012451, | |
| "learning_rate": 0.00011615374405066652, | |
| "loss": 1.1197, | |
| "step": 45100 | |
| }, | |
| { | |
| "epoch": 2.5712497866772854, | |
| "grad_norm": 1.6989002227783203, | |
| "learning_rate": 0.00011463678252460321, | |
| "loss": 1.1247, | |
| "step": 45200 | |
| }, | |
| { | |
| "epoch": 2.576938392400023, | |
| "grad_norm": 1.5901920795440674, | |
| "learning_rate": 0.00011313499061380057, | |
| "loss": 1.1352, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 2.58262699812276, | |
| "grad_norm": 1.4382330179214478, | |
| "learning_rate": 0.00011161802908773727, | |
| "loss": 1.1093, | |
| "step": 45400 | |
| }, | |
| { | |
| "epoch": 2.5883156038454973, | |
| "grad_norm": 1.8520530462265015, | |
| "learning_rate": 0.00011010106756167397, | |
| "loss": 1.1081, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 2.5940042095682347, | |
| "grad_norm": 1.8772435188293457, | |
| "learning_rate": 0.00010858410603561066, | |
| "loss": 1.1157, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 2.599692815290972, | |
| "grad_norm": 1.6013365983963013, | |
| "learning_rate": 0.00010706714450954738, | |
| "loss": 1.1463, | |
| "step": 45700 | |
| }, | |
| { | |
| "epoch": 2.6053814210137096, | |
| "grad_norm": 1.582515835762024, | |
| "learning_rate": 0.0001055501829834841, | |
| "loss": 1.1135, | |
| "step": 45800 | |
| }, | |
| { | |
| "epoch": 2.611070026736447, | |
| "grad_norm": 1.3782535791397095, | |
| "learning_rate": 0.00010403322145742079, | |
| "loss": 1.1101, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 2.6167586324591845, | |
| "grad_norm": 1.465584397315979, | |
| "learning_rate": 0.00010251625993135749, | |
| "loss": 1.1354, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 2.6224472381819215, | |
| "grad_norm": 1.4038536548614502, | |
| "learning_rate": 0.00010099929840529421, | |
| "loss": 1.1078, | |
| "step": 46100 | |
| }, | |
| { | |
| "epoch": 2.628135843904659, | |
| "grad_norm": 1.9926286935806274, | |
| "learning_rate": 9.948233687923091e-05, | |
| "loss": 1.1044, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 2.6338244496273964, | |
| "grad_norm": 1.6215740442276, | |
| "learning_rate": 9.796537535316762e-05, | |
| "loss": 1.1188, | |
| "step": 46300 | |
| }, | |
| { | |
| "epoch": 2.639513055350134, | |
| "grad_norm": 1.5623165369033813, | |
| "learning_rate": 9.644841382710431e-05, | |
| "loss": 1.1155, | |
| "step": 46400 | |
| }, | |
| { | |
| "epoch": 2.6452016610728712, | |
| "grad_norm": 1.491926670074463, | |
| "learning_rate": 9.493145230104102e-05, | |
| "loss": 1.1211, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 2.6508902667956082, | |
| "grad_norm": 1.7084381580352783, | |
| "learning_rate": 9.341449077497773e-05, | |
| "loss": 1.1091, | |
| "step": 46600 | |
| }, | |
| { | |
| "epoch": 2.6565788725183457, | |
| "grad_norm": 1.5060371160507202, | |
| "learning_rate": 9.189752924891443e-05, | |
| "loss": 1.1198, | |
| "step": 46700 | |
| }, | |
| { | |
| "epoch": 2.662267478241083, | |
| "grad_norm": 1.7321504354476929, | |
| "learning_rate": 9.038056772285112e-05, | |
| "loss": 1.1157, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 2.6679560839638206, | |
| "grad_norm": 1.559877634048462, | |
| "learning_rate": 8.886360619678784e-05, | |
| "loss": 1.1035, | |
| "step": 46900 | |
| }, | |
| { | |
| "epoch": 2.673644689686558, | |
| "grad_norm": 1.8588401079177856, | |
| "learning_rate": 8.734664467072455e-05, | |
| "loss": 1.1288, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 2.679333295409295, | |
| "grad_norm": 1.751246452331543, | |
| "learning_rate": 8.582968314466125e-05, | |
| "loss": 1.1206, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 2.6850219011320324, | |
| "grad_norm": 1.7309458255767822, | |
| "learning_rate": 8.431272161859795e-05, | |
| "loss": 1.1089, | |
| "step": 47200 | |
| }, | |
| { | |
| "epoch": 2.69071050685477, | |
| "grad_norm": 1.8057925701141357, | |
| "learning_rate": 8.281092970779529e-05, | |
| "loss": 1.1244, | |
| "step": 47300 | |
| }, | |
| { | |
| "epoch": 2.6963991125775073, | |
| "grad_norm": 1.7594059705734253, | |
| "learning_rate": 8.1293968181732e-05, | |
| "loss": 1.1188, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 2.7020877183002447, | |
| "grad_norm": 1.686438798904419, | |
| "learning_rate": 7.97770066556687e-05, | |
| "loss": 1.1068, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 2.7077763240229817, | |
| "grad_norm": 1.6962246894836426, | |
| "learning_rate": 7.82600451296054e-05, | |
| "loss": 1.1043, | |
| "step": 47600 | |
| }, | |
| { | |
| "epoch": 2.713464929745719, | |
| "grad_norm": 1.5946807861328125, | |
| "learning_rate": 7.67430836035421e-05, | |
| "loss": 1.1109, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 2.7191535354684566, | |
| "grad_norm": 1.4834094047546387, | |
| "learning_rate": 7.522612207747881e-05, | |
| "loss": 1.114, | |
| "step": 47800 | |
| }, | |
| { | |
| "epoch": 2.724842141191194, | |
| "grad_norm": 1.763058066368103, | |
| "learning_rate": 7.370916055141553e-05, | |
| "loss": 1.1091, | |
| "step": 47900 | |
| }, | |
| { | |
| "epoch": 2.7305307469139315, | |
| "grad_norm": 1.9240601062774658, | |
| "learning_rate": 7.219219902535223e-05, | |
| "loss": 1.0936, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.7362193526366685, | |
| "grad_norm": 1.4768198728561401, | |
| "learning_rate": 7.067523749928892e-05, | |
| "loss": 1.1158, | |
| "step": 48100 | |
| }, | |
| { | |
| "epoch": 2.7419079583594064, | |
| "grad_norm": 1.9692409038543701, | |
| "learning_rate": 6.915827597322563e-05, | |
| "loss": 1.1201, | |
| "step": 48200 | |
| }, | |
| { | |
| "epoch": 2.7475965640821434, | |
| "grad_norm": 1.636785864830017, | |
| "learning_rate": 6.764131444716234e-05, | |
| "loss": 1.1092, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 2.753285169804881, | |
| "grad_norm": 1.5599926710128784, | |
| "learning_rate": 6.612435292109905e-05, | |
| "loss": 1.0932, | |
| "step": 48400 | |
| }, | |
| { | |
| "epoch": 2.7589737755276182, | |
| "grad_norm": 1.695862054824829, | |
| "learning_rate": 6.460739139503574e-05, | |
| "loss": 1.1227, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 2.7646623812503557, | |
| "grad_norm": 1.8806819915771484, | |
| "learning_rate": 6.309042986897246e-05, | |
| "loss": 1.1049, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 2.770350986973093, | |
| "grad_norm": 1.814792513847351, | |
| "learning_rate": 6.157346834290916e-05, | |
| "loss": 1.1149, | |
| "step": 48700 | |
| }, | |
| { | |
| "epoch": 2.77603959269583, | |
| "grad_norm": 2.068614959716797, | |
| "learning_rate": 6.005650681684586e-05, | |
| "loss": 1.1181, | |
| "step": 48800 | |
| }, | |
| { | |
| "epoch": 2.7817281984185676, | |
| "grad_norm": 1.5576444864273071, | |
| "learning_rate": 5.853954529078256e-05, | |
| "loss": 1.1223, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 2.787416804141305, | |
| "grad_norm": 1.8175384998321533, | |
| "learning_rate": 5.7022583764719273e-05, | |
| "loss": 1.1113, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 2.7931054098640424, | |
| "grad_norm": 1.570915937423706, | |
| "learning_rate": 5.550562223865598e-05, | |
| "loss": 1.1123, | |
| "step": 49100 | |
| }, | |
| { | |
| "epoch": 2.79879401558678, | |
| "grad_norm": 1.9663364887237549, | |
| "learning_rate": 5.3988660712592675e-05, | |
| "loss": 1.1065, | |
| "step": 49200 | |
| }, | |
| { | |
| "epoch": 2.804482621309517, | |
| "grad_norm": 2.2906079292297363, | |
| "learning_rate": 5.248686880179001e-05, | |
| "loss": 1.0993, | |
| "step": 49300 | |
| }, | |
| { | |
| "epoch": 2.8101712270322543, | |
| "grad_norm": 1.566801905632019, | |
| "learning_rate": 5.096990727572673e-05, | |
| "loss": 1.0964, | |
| "step": 49400 | |
| }, | |
| { | |
| "epoch": 2.8158598327549917, | |
| "grad_norm": 1.7769867181777954, | |
| "learning_rate": 4.9452945749663425e-05, | |
| "loss": 1.0978, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 2.821548438477729, | |
| "grad_norm": 1.9856287240982056, | |
| "learning_rate": 4.7935984223600136e-05, | |
| "loss": 1.0875, | |
| "step": 49600 | |
| }, | |
| { | |
| "epoch": 2.8272370442004666, | |
| "grad_norm": 1.7836079597473145, | |
| "learning_rate": 4.6419022697536834e-05, | |
| "loss": 1.1056, | |
| "step": 49700 | |
| }, | |
| { | |
| "epoch": 2.8329256499232036, | |
| "grad_norm": 1.9246402978897095, | |
| "learning_rate": 4.4902061171473545e-05, | |
| "loss": 1.1074, | |
| "step": 49800 | |
| }, | |
| { | |
| "epoch": 2.838614255645941, | |
| "grad_norm": 1.3988184928894043, | |
| "learning_rate": 4.338509964541024e-05, | |
| "loss": 1.1206, | |
| "step": 49900 | |
| }, | |
| { | |
| "epoch": 2.8443028613686785, | |
| "grad_norm": 1.7193849086761475, | |
| "learning_rate": 4.186813811934695e-05, | |
| "loss": 1.1245, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.8443028613686785, | |
| "eval_accuracy": 0.729988, | |
| "eval_loss": 1.0782374143600464, | |
| "eval_runtime": 82.4205, | |
| "eval_samples_per_second": 3033.226, | |
| "eval_steps_per_second": 11.854, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.849991467091416, | |
| "grad_norm": 1.7059062719345093, | |
| "learning_rate": 4.035117659328366e-05, | |
| "loss": 1.1009, | |
| "step": 50100 | |
| }, | |
| { | |
| "epoch": 2.8556800728141534, | |
| "grad_norm": 1.4554681777954102, | |
| "learning_rate": 3.883421506722036e-05, | |
| "loss": 1.1108, | |
| "step": 50200 | |
| }, | |
| { | |
| "epoch": 2.8613686785368904, | |
| "grad_norm": 1.7067590951919556, | |
| "learning_rate": 3.7317253541157065e-05, | |
| "loss": 1.0956, | |
| "step": 50300 | |
| }, | |
| { | |
| "epoch": 2.867057284259628, | |
| "grad_norm": 1.7176940441131592, | |
| "learning_rate": 3.580029201509377e-05, | |
| "loss": 1.0841, | |
| "step": 50400 | |
| }, | |
| { | |
| "epoch": 2.8727458899823652, | |
| "grad_norm": 1.7251313924789429, | |
| "learning_rate": 3.4283330489030474e-05, | |
| "loss": 1.0908, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 2.8784344957051027, | |
| "grad_norm": 1.668372631072998, | |
| "learning_rate": 3.276636896296718e-05, | |
| "loss": 1.0954, | |
| "step": 50600 | |
| }, | |
| { | |
| "epoch": 2.88412310142784, | |
| "grad_norm": 1.889109492301941, | |
| "learning_rate": 3.124940743690388e-05, | |
| "loss": 1.1096, | |
| "step": 50700 | |
| }, | |
| { | |
| "epoch": 2.8898117071505776, | |
| "grad_norm": 1.509391188621521, | |
| "learning_rate": 2.973244591084059e-05, | |
| "loss": 1.09, | |
| "step": 50800 | |
| }, | |
| { | |
| "epoch": 2.895500312873315, | |
| "grad_norm": 2.024489402770996, | |
| "learning_rate": 2.821548438477729e-05, | |
| "loss": 1.0993, | |
| "step": 50900 | |
| }, | |
| { | |
| "epoch": 2.901188918596052, | |
| "grad_norm": 2.007756471633911, | |
| "learning_rate": 2.6698522858713998e-05, | |
| "loss": 1.1029, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 2.9068775243187894, | |
| "grad_norm": 1.5296841859817505, | |
| "learning_rate": 2.51815613326507e-05, | |
| "loss": 1.095, | |
| "step": 51100 | |
| }, | |
| { | |
| "epoch": 2.912566130041527, | |
| "grad_norm": 1.6109613180160522, | |
| "learning_rate": 2.3664599806587406e-05, | |
| "loss": 1.1044, | |
| "step": 51200 | |
| }, | |
| { | |
| "epoch": 2.9182547357642643, | |
| "grad_norm": 1.8067957162857056, | |
| "learning_rate": 2.216280789578474e-05, | |
| "loss": 1.1074, | |
| "step": 51300 | |
| }, | |
| { | |
| "epoch": 2.9239433414870017, | |
| "grad_norm": 1.6997472047805786, | |
| "learning_rate": 2.064584636972145e-05, | |
| "loss": 1.1036, | |
| "step": 51400 | |
| }, | |
| { | |
| "epoch": 2.9296319472097387, | |
| "grad_norm": 1.6443182229995728, | |
| "learning_rate": 1.9128884843658153e-05, | |
| "loss": 1.1258, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 2.935320552932476, | |
| "grad_norm": 1.868161916732788, | |
| "learning_rate": 1.7611923317594857e-05, | |
| "loss": 1.115, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 2.9410091586552136, | |
| "grad_norm": 1.5620206594467163, | |
| "learning_rate": 1.609496179153156e-05, | |
| "loss": 1.1075, | |
| "step": 51700 | |
| }, | |
| { | |
| "epoch": 2.946697764377951, | |
| "grad_norm": 1.6326332092285156, | |
| "learning_rate": 1.4578000265468267e-05, | |
| "loss": 1.1113, | |
| "step": 51800 | |
| }, | |
| { | |
| "epoch": 2.9523863701006885, | |
| "grad_norm": 1.805126428604126, | |
| "learning_rate": 1.3061038739404973e-05, | |
| "loss": 1.0937, | |
| "step": 51900 | |
| }, | |
| { | |
| "epoch": 2.9580749758234255, | |
| "grad_norm": 1.5693707466125488, | |
| "learning_rate": 1.1544077213341677e-05, | |
| "loss": 1.1053, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.963763581546163, | |
| "grad_norm": 1.4851309061050415, | |
| "learning_rate": 1.0027115687278382e-05, | |
| "loss": 1.1101, | |
| "step": 52100 | |
| }, | |
| { | |
| "epoch": 2.9694521872689004, | |
| "grad_norm": 1.8946778774261475, | |
| "learning_rate": 8.510154161215086e-06, | |
| "loss": 1.0934, | |
| "step": 52200 | |
| }, | |
| { | |
| "epoch": 2.975140792991638, | |
| "grad_norm": 1.5624499320983887, | |
| "learning_rate": 6.993192635151791e-06, | |
| "loss": 1.0994, | |
| "step": 52300 | |
| }, | |
| { | |
| "epoch": 2.9808293987143752, | |
| "grad_norm": 1.5662641525268555, | |
| "learning_rate": 5.476231109088497e-06, | |
| "loss": 1.1058, | |
| "step": 52400 | |
| }, | |
| { | |
| "epoch": 2.9865180044371122, | |
| "grad_norm": 1.514809250831604, | |
| "learning_rate": 3.959269583025201e-06, | |
| "loss": 1.0967, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 2.9922066101598497, | |
| "grad_norm": 1.8442556858062744, | |
| "learning_rate": 2.4423080569619053e-06, | |
| "loss": 1.1041, | |
| "step": 52600 | |
| }, | |
| { | |
| "epoch": 2.997895215882587, | |
| "grad_norm": 1.7445664405822754, | |
| "learning_rate": 9.253465308986101e-07, | |
| "loss": 1.0786, | |
| "step": 52700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 52737, | |
| "total_flos": 1.3116020904e+17, | |
| "train_loss": 1.269093582550002, | |
| "train_runtime": 7252.1306, | |
| "train_samples_per_second": 1861.522, | |
| "train_steps_per_second": 7.272 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 52737, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3116020904e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |