| { | |
| "best_metric": 7.685359477996826, | |
| "best_model_checkpoint": "distilbert_rand_5_v1/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": 2.039281129837036, | |
| "learning_rate": 5e-06, | |
| "loss": 10.9029, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.9287338256835938, | |
| "learning_rate": 1e-05, | |
| "loss": 9.1841, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.4802101850509644, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.6564, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.7293734550476074, | |
| "learning_rate": 2e-05, | |
| "loss": 8.4911, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.3555989265441895, | |
| "learning_rate": 2.5e-05, | |
| "loss": 8.3596, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.6210575103759766, | |
| "learning_rate": 3e-05, | |
| "loss": 8.2513, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.6368871927261353, | |
| "learning_rate": 3.5e-05, | |
| "loss": 8.1571, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.1069653034210205, | |
| "learning_rate": 4e-05, | |
| "loss": 8.0875, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.0065685510635376, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.0322, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.7704250812530518, | |
| "learning_rate": 5e-05, | |
| "loss": 7.9847, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.5488605499267578, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 7.9438, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.5934131145477295, | |
| "learning_rate": 6e-05, | |
| "loss": 7.9086, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.879814863204956, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 7.8748, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.822333335876465, | |
| "learning_rate": 7e-05, | |
| "loss": 7.8457, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 3.0464205741882324, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 7.8174, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.6054975986480713, | |
| "learning_rate": 8e-05, | |
| "loss": 7.7947, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.4609038829803467, | |
| "learning_rate": 8.5e-05, | |
| "loss": 7.7786, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 1.484981656074524, | |
| "learning_rate": 9e-05, | |
| "loss": 7.7554, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 3.5411243438720703, | |
| "learning_rate": 9.5e-05, | |
| "loss": 7.7371, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.6537575721740723, | |
| "learning_rate": 0.0001, | |
| "loss": 7.6959, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15081109658031666, | |
| "eval_loss": 7.685359477996826, | |
| "eval_runtime": 0.8566, | |
| "eval_samples_per_second": 559.2, | |
| "eval_steps_per_second": 5.837, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 1.9814363718032837, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 7.6852, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 3.880098819732666, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 7.6711, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.329343557357788, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 7.6665, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.457648754119873, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 7.6415, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.081408977508545, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 7.5901, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 3.657313108444214, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 7.5862, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 1.2980384826660156, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 7.589, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 2.0493595600128174, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 7.5806, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.8665246963500977, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 7.5497, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.7017197608947754, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 7.5217, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.2749807834625244, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 7.5175, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.383842945098877, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 7.5202, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.1351611614227295, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 7.519, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.5209946632385254, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 7.4714, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 2.8465027809143066, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 7.4515, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.774460792541504, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 7.4603, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.9616355895996094, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 7.466, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.9673783779144287, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 7.4637, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.6345326900482178, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 7.3823, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 4.010097026824951, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 7.388, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15171581694211397, | |
| "eval_loss": 7.723523139953613, | |
| "eval_runtime": 0.8022, | |
| "eval_samples_per_second": 597.109, | |
| "eval_steps_per_second": 6.233, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.2331223487854004, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 7.3888, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 2.360450029373169, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 7.3995, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 4.477081775665283, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 7.3814, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 3.3145029544830322, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 7.2736, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 3.1475942134857178, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 7.2858, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 3.703411340713501, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 7.2946, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 4.02745246887207, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 7.2991, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 4.219743728637695, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 7.2287, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 4.181122303009033, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.1246, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 3.9847166538238525, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.1326, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 4.292736053466797, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 7.1524, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 3.4725711345672607, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 7.1593, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 4.945437431335449, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.0236, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 5.220856666564941, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.9468, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 4.902286529541016, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.9697, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 5.443831443786621, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.9735, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 4.451772689819336, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.991, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 4.548737049102783, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.8034, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 5.091429710388184, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.7826, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 6.505558013916016, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.7994, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15412597587934074, | |
| "eval_loss": 8.394573211669922, | |
| "eval_runtime": 0.8473, | |
| "eval_samples_per_second": 565.317, | |
| "eval_steps_per_second": 5.901, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 5.261996746063232, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.8172, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 5.574769496917725, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.8096, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 7.227789402008057, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.6236, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 5.911694049835205, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.6439, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 7.990551471710205, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.6701, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 5.785084247589111, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.6689, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 5.379305362701416, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.6165, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 7.4263200759887695, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.5049, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 5.981423377990723, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.5283, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 5.744656085968018, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.5396, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 6.555784225463867, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.5556, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 7.307421684265137, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.4596, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 6.634768009185791, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.408, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 6.6910719871521, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.4188, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 5.394948959350586, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.4425, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 8.720141410827637, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.4468, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 6.355972766876221, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.3413, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 7.163991451263428, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.3292, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 9.228317260742188, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.3301, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 6.461520671844482, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.3424, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15201300885948188, | |
| "eval_loss": 9.337831497192383, | |
| "eval_runtime": 0.8129, | |
| "eval_samples_per_second": 589.218, | |
| "eval_steps_per_second": 6.151, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 4.4183783531188965, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.355, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 8.304726600646973, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.2505, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 7.90907096862793, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.2587, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 5.80265998840332, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.2695, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 7.4769511222839355, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.2729, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 6.589922904968262, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.2562, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 6.942967891693115, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.197, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 5.921877384185791, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.2027, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 7.463561058044434, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.2125, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 6.808112621307373, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.2247, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 8.286489486694336, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.1849, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 7.345949172973633, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.1541, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 6.1166672706604, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.1592, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 6.870145797729492, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.1667, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 6.141603469848633, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.1655, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 3.998727798461914, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.1416, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 5.184545993804932, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.1285, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 5.28961181640625, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.1254, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 4.611685276031494, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.1256, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 4.233712196350098, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.1304, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1515126026859562, | |
| "eval_loss": 10.069047927856445, | |
| "eval_runtime": 0.7886, | |
| "eval_samples_per_second": 607.401, | |
| "eval_steps_per_second": 6.34, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 6.673228740692139, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.095, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 3.821599006652832, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.1036, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 4.245255470275879, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.0997, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 6.308642864227295, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.1006, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 5.6796698570251465, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.0939, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 9.24228286743164, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.0681, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 5.414360046386719, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.0705, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 8.038673400878906, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.0794, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 5.303706169128418, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.0686, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 4.699199676513672, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.0621, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 2.3373281955718994, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.0539, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 5.043691158294678, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.0586, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 5.2183003425598145, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.0496, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 3.504364252090454, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.0515, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 6.812923431396484, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.0429, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 4.761589527130127, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.0466, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 5.1148295402526855, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.0421, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 3.334099292755127, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.0408, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 3.380460023880005, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.0352, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.68154255975296e+17, | |
| "train_loss": 6.949937666757681, | |
| "train_runtime": 16605.5225, | |
| "train_samples_per_second": 344.221, | |
| "train_steps_per_second": 3.586 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 7.68154255975296e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |