| { | |
| "best_metric": 6.37915563583374, | |
| "best_model_checkpoint": "tiny_bert_km_10_v1/checkpoint-20000", | |
| "epoch": 25.0, | |
| "eval_steps": 10000, | |
| "global_step": 59550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 2.7388250827789307, | |
| "learning_rate": 5e-06, | |
| "loss": 11.6168, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.9273589849472046, | |
| "learning_rate": 1e-05, | |
| "loss": 9.7786, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 4.180909156799316, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.4199, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 7.06673002243042, | |
| "learning_rate": 2e-05, | |
| "loss": 7.7491, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 3.5001888275146484, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.4968, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 5.8357625007629395, | |
| "learning_rate": 3e-05, | |
| "loss": 7.348, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 3.2707433700561523, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.2502, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 3.111070156097412, | |
| "learning_rate": 4e-05, | |
| "loss": 7.1679, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 4.673737049102783, | |
| "learning_rate": 4.5e-05, | |
| "loss": 7.0854, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 7.858347415924072, | |
| "learning_rate": 5e-05, | |
| "loss": 6.9865, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.2085518836975098, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.9155, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 3.2788798809051514, | |
| "learning_rate": 6e-05, | |
| "loss": 6.865, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.9550942182540894, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 6.8255, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 3.064227819442749, | |
| "learning_rate": 7e-05, | |
| "loss": 6.7831, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 4.296220779418945, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.7104, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.6907804012298584, | |
| "learning_rate": 8e-05, | |
| "loss": 6.6902, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 3.918893337249756, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.6831, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 4.212062358856201, | |
| "learning_rate": 9e-05, | |
| "loss": 6.6582, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 4.881993293762207, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.6402, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.220136880874634, | |
| "learning_rate": 0.0001, | |
| "loss": 6.5744, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.14897465149280725, | |
| "eval_loss": 6.5599045753479, | |
| "eval_runtime": 0.7174, | |
| "eval_samples_per_second": 667.662, | |
| "eval_steps_per_second": 6.969, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.8914411067962646, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 6.5648, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 1.8477387428283691, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 6.555, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.3890867233276367, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 6.5482, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.4220800399780273, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.52, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.328453540802002, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.4597, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 1.8657039403915405, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.4575, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.2525665760040283, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.4556, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.4106040000915527, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.4439, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.6294212341308594, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.4127, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 1.7568219900131226, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.3798, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 1.913711428642273, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.3688, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.5787100791931152, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.3717, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 1.3250707387924194, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.3659, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 1.4115550518035889, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.3233, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 1.4667936563491821, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.3029, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.138885021209717, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.3087, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.342480182647705, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.3095, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.0234673023223877, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.31, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.6231536865234375, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.2591, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 1.916068434715271, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.2586, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15247341395662056, | |
| "eval_loss": 6.37915563583374, | |
| "eval_runtime": 0.6861, | |
| "eval_samples_per_second": 698.143, | |
| "eval_steps_per_second": 7.288, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 1.823195457458496, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.2556, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 2.02011775970459, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.2563, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.9880938529968262, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.2474, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 1.8998876810073853, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.216, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 2.565138101577759, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.2138, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 2.2761712074279785, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.2175, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 2.7417733669281006, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.2142, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 2.4117789268493652, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.1953, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 2.096160888671875, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.1829, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 1.4140809774398804, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.1775, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 3.86072039604187, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.1777, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 2.6061575412750244, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.1788, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 1.4297683238983154, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 6.1585, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 1.1012049913406372, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.1529, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 2.554164409637451, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.1453, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 2.2018940448760986, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.1476, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 1.779039740562439, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.1455, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 1.8783884048461914, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.1282, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 1.489698886871338, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.1167, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 1.6969472169876099, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.1222, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15429386909925288, | |
| "eval_loss": 6.382654190063477, | |
| "eval_runtime": 0.6965, | |
| "eval_samples_per_second": 687.753, | |
| "eval_steps_per_second": 7.179, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 1.95393705368042, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.1174, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 2.023423671722412, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.1199, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 2.619342088699341, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.0982, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 0.87702876329422, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.0935, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 2.397125244140625, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.098, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 3.058929920196533, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.0966, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 1.2640635967254639, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.0867, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 1.8706773519515991, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.0725, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 1.2959846258163452, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.0745, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 1.155418038368225, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.0718, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 1.2469680309295654, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.0716, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 1.8946082592010498, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.0597, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 1.3731963634490967, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.0573, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 2.045217990875244, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.0529, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 4.961300373077393, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.0536, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 1.9158598184585571, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.0494, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 3.2773678302764893, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.0447, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 1.7545653581619263, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.0357, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 2.4140148162841797, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.038, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 3.643996238708496, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.0276, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15204104519457218, | |
| "eval_loss": 6.4046454429626465, | |
| "eval_runtime": 0.6753, | |
| "eval_samples_per_second": 709.271, | |
| "eval_steps_per_second": 7.404, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 2.8720858097076416, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.0361, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 1.8828935623168945, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.0217, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 1.7057238817214966, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.0186, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 1.7667478322982788, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.0177, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 2.077155351638794, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.0187, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 0.9542981386184692, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.0114, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 3.122864007949829, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.005, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 2.7688279151916504, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.0042, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 2.1192431449890137, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.0047, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 1.6212798357009888, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.0067, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 2.2240281105041504, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.0015, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 3.265800714492798, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.9882, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 1.3882496356964111, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.9911, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 2.1964356899261475, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.9941, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 1.27981698513031, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.9893, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 1.3110111951828003, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.9848, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 3.652817726135254, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.9822, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 0.8469610214233398, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.9795, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 1.5175819396972656, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.9764, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 3.077249050140381, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.9768, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1514845655648078, | |
| "eval_loss": 6.436910629272461, | |
| "eval_runtime": 0.6864, | |
| "eval_samples_per_second": 697.858, | |
| "eval_steps_per_second": 7.285, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 1.3623372316360474, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.9697, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 1.1987028121948242, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.9733, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 1.3726773262023926, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.9694, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 1.0265616178512573, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.9698, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 1.121810793876648, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.9693, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 1.8554517030715942, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.9587, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 2.199672222137451, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.9646, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 1.7775245904922485, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.9635, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 0.8349699378013611, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.9581, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 0.9259258508682251, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.9582, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 2.898632049560547, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.9559, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 0.9831527471542358, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.9566, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 1.2274006605148315, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.9499, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 1.2322320938110352, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.954, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 1.4643625020980835, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.9489, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 0.8192181587219238, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.9526, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 0.8231164216995239, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.9499, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 0.7896997928619385, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.9476, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 1.2367829084396362, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.9484, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.050247306064896e+17, | |
| "train_loss": 6.340149333707231, | |
| "train_runtime": 11680.4757, | |
| "train_samples_per_second": 489.361, | |
| "train_steps_per_second": 5.098 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.050247306064896e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |