| { | |
| "best_metric": 8.516840934753418, | |
| "best_model_checkpoint": "tiny_bert_rand_10_v2/checkpoint-10000", | |
| "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": 2.8956918716430664, | |
| "learning_rate": 5e-06, | |
| "loss": 12.0528, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.7897758483886719, | |
| "learning_rate": 1e-05, | |
| "loss": 10.4985, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.651282787322998, | |
| "learning_rate": 1.5e-05, | |
| "loss": 9.511, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.1039087772369385, | |
| "learning_rate": 2e-05, | |
| "loss": 9.313, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.4416009187698364, | |
| "learning_rate": 2.5e-05, | |
| "loss": 9.1977, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.8553386926651, | |
| "learning_rate": 3e-05, | |
| "loss": 9.0932, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.4036210775375366, | |
| "learning_rate": 3.5e-05, | |
| "loss": 9.0077, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.5917521715164185, | |
| "learning_rate": 4e-05, | |
| "loss": 8.9268, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.7779933214187622, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.8497, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.4821877479553223, | |
| "learning_rate": 5e-05, | |
| "loss": 8.7864, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.7116888761520386, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 8.7418, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.8798326253890991, | |
| "learning_rate": 6e-05, | |
| "loss": 8.7053, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.150996685028076, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 8.6733, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 1.4576572179794312, | |
| "learning_rate": 7e-05, | |
| "loss": 8.6438, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.110964775085449, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 8.603, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.01837158203125, | |
| "learning_rate": 8e-05, | |
| "loss": 8.5734, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.289891481399536, | |
| "learning_rate": 8.5e-05, | |
| "loss": 8.5586, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.272826671600342, | |
| "learning_rate": 9e-05, | |
| "loss": 8.5364, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 1.7595138549804688, | |
| "learning_rate": 9.5e-05, | |
| "loss": 8.517, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.2657644748687744, | |
| "learning_rate": 0.0001, | |
| "loss": 8.4324, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15014329836667686, | |
| "eval_loss": 8.516840934753418, | |
| "eval_runtime": 0.7382, | |
| "eval_samples_per_second": 648.908, | |
| "eval_steps_per_second": 6.774, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.0055155754089355, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 8.4296, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.568305015563965, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 8.4221, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.9037818908691406, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 8.4254, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 1.9787847995758057, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 8.3976, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.553103446960449, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 8.3192, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.580944299697876, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 8.3209, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.5693955421447754, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 8.3257, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 2.989201068878174, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 8.3246, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.8055412769317627, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 8.288, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.1214358806610107, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 8.2457, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.7707295417785645, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 8.2513, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.9960660934448242, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 8.2557, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.213331937789917, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 8.2587, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.097681760787964, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 8.2045, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 2.481717824935913, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 8.1824, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.7683825492858887, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 8.1951, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.1083154678344727, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 8.2034, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 1.9768694639205933, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 8.2037, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.308565855026245, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 8.1362, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 1.9111008644104004, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 8.1359, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1513510480092034, | |
| "eval_loss": 8.578495025634766, | |
| "eval_runtime": 0.6882, | |
| "eval_samples_per_second": 696.015, | |
| "eval_steps_per_second": 7.265, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.631357192993164, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 8.1384, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 2.378553628921509, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 8.1512, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 2.799654722213745, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 8.1354, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 3.0007619857788086, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 8.0683, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 3.1445086002349854, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 8.0738, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 2.6713919639587402, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 8.0859, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 2.236157178878784, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 8.0858, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 2.471778154373169, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.0462, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 3.3160719871520996, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.9911, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.7243497371673584, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.9938, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 3.617450475692749, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 8.0075, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 3.0119311809539795, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 8.0181, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 3.3908631801605225, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.93, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 3.8109664916992188, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.8902, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 3.2663817405700684, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.9052, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 3.765650749206543, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.9112, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 3.401289463043213, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.9237, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 3.8314595222473145, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.7856, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 4.359664440155029, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.7689, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 4.1663923263549805, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.791, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15407001147270336, | |
| "eval_loss": 9.017232894897461, | |
| "eval_runtime": 0.6928, | |
| "eval_samples_per_second": 691.393, | |
| "eval_steps_per_second": 7.217, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 3.787510871887207, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.8107, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 4.561687469482422, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.8127, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 4.102744102478027, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.6261, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 3.8658947944641113, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.6531, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 4.556894779205322, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.683, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 4.922505855560303, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.6933, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 4.5583696365356445, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.6311, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 5.245599269866943, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 7.5044, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 4.251413822174072, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 7.5302, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 4.954182147979736, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 7.5437, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 5.95086145401001, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 7.5642, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 5.5313310623168945, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 7.4553, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 6.156241416931152, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 7.3874, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 5.391365051269531, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 7.4066, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 5.771765232086182, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 7.4263, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 4.8253631591796875, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 7.4372, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 5.755190849304199, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 7.2846, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 5.651447772979736, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 7.2686, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 5.568587303161621, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 7.2883, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 6.008751392364502, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 7.3, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15122799147695412, | |
| "eval_loss": 9.511892318725586, | |
| "eval_runtime": 0.6891, | |
| "eval_samples_per_second": 695.131, | |
| "eval_steps_per_second": 7.256, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 6.794548511505127, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 7.3289, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 6.964717388153076, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 7.1414, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 6.341984748840332, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 7.1565, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 7.717092514038086, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 7.1761, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 6.288142204284668, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 7.1904, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 7.551562309265137, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 7.1564, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 6.559009075164795, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 7.0359, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 8.913935661315918, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 7.0632, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 7.239193916320801, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 7.0772, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 7.293575286865234, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 7.1013, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 7.253826141357422, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 7.0193, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 7.525722503662109, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.9527, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 8.178430557250977, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.9639, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 8.326142311096191, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.9762, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 6.413033485412598, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.99, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 8.123743057250977, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.9007, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 7.148409366607666, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.8722, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.879838466644287, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.8796, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 7.241456985473633, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.8861, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 7.6523895263671875, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.9042, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1517929738974402, | |
| "eval_loss": 10.092184066772461, | |
| "eval_runtime": 0.6859, | |
| "eval_samples_per_second": 698.316, | |
| "eval_steps_per_second": 7.289, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 7.539754867553711, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.7883, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 7.285797595977783, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.7968, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 8.29957389831543, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.8139, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 8.12922477722168, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.8139, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 7.652002334594727, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.7965, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 9.461636543273926, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.7156, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 6.493343830108643, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.7325, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 7.998405456542969, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.7438, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 8.43569278717041, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.7429, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 8.705036163330078, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.7146, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 8.198236465454102, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.6754, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 8.432761192321777, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.6801, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 8.555904388427734, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.6826, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 8.87362003326416, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.6981, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 6.883061408996582, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.6573, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 7.8268585205078125, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.655, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 8.802783012390137, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.6503, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 9.875340461730957, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.6436, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 6.891987323760986, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.6469, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.050247306064896e+17, | |
| "train_loss": 7.77363054584596, | |
| "train_runtime": 11901.4866, | |
| "train_samples_per_second": 480.274, | |
| "train_steps_per_second": 5.004 | |
| } | |
| ], | |
| "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 | |
| } | |