| { | |
| "best_metric": 9.192044258117676, | |
| "best_model_checkpoint": "distilbert_rand_20_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": 2.5388851165771484, | |
| "learning_rate": 5e-06, | |
| "loss": 12.2945, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.0317511558532715, | |
| "learning_rate": 1e-05, | |
| "loss": 10.5895, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 2.547168731689453, | |
| "learning_rate": 1.5e-05, | |
| "loss": 10.0641, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 2.388481616973877, | |
| "learning_rate": 2e-05, | |
| "loss": 9.9145, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 3.0018980503082275, | |
| "learning_rate": 2.5e-05, | |
| "loss": 9.7791, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 2.1324257850646973, | |
| "learning_rate": 3e-05, | |
| "loss": 9.6624, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.958751916885376, | |
| "learning_rate": 3.5e-05, | |
| "loss": 9.5664, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.509357452392578, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4895, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 2.2945008277893066, | |
| "learning_rate": 4.5e-05, | |
| "loss": 9.4305, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.1530516147613525, | |
| "learning_rate": 5e-05, | |
| "loss": 9.3801, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.593848705291748, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 9.3394, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.2520973682403564, | |
| "learning_rate": 6e-05, | |
| "loss": 9.3038, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.745112419128418, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 9.2716, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 3.265744924545288, | |
| "learning_rate": 7e-05, | |
| "loss": 9.243, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.6872153282165527, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 9.2024, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.8887720108032227, | |
| "learning_rate": 8e-05, | |
| "loss": 9.1763, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.497943878173828, | |
| "learning_rate": 8.5e-05, | |
| "loss": 9.1628, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.481393337249756, | |
| "learning_rate": 9e-05, | |
| "loss": 9.1404, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.962963581085205, | |
| "learning_rate": 9.5e-05, | |
| "loss": 9.1291, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.0763323307037354, | |
| "learning_rate": 0.0001, | |
| "loss": 9.0219, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.151701494198503, | |
| "eval_loss": 9.192044258117676, | |
| "eval_runtime": 0.8579, | |
| "eval_samples_per_second": 558.368, | |
| "eval_steps_per_second": 5.828, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 3.471935510635376, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.0242, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 3.224785804748535, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.0194, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.1821765899658203, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.0236, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.7196967601776123, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 8.9851, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.693763494491577, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 8.868, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 3.4376792907714844, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 8.8834, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.813124179840088, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 8.8941, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.179331064224243, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 8.8973, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.916917085647583, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 8.827, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 3.313599109649658, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 8.7613, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 3.103424549102783, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 8.7768, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.9828693866729736, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 8.7915, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 3.0750062465667725, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 8.8007, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.3732056617736816, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 8.6801, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.928379774093628, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 8.6324, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 3.493884801864624, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 8.6612, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 3.120750665664673, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 8.6772, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.9115488529205322, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 8.6885, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 3.8506040573120117, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 8.4588, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 4.189122676849365, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 8.4577, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1520244675776537, | |
| "eval_loss": 9.528057098388672, | |
| "eval_runtime": 0.8426, | |
| "eval_samples_per_second": 568.449, | |
| "eval_steps_per_second": 5.934, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 4.542478561401367, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 8.4734, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 4.224944591522217, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 8.4959, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 5.228671073913574, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 8.4571, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 5.052661418914795, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 8.1411, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 5.872274398803711, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 8.172, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 5.12518310546875, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 8.2105, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 5.485617637634277, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 8.2181, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 5.4581685066223145, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.0794, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 5.811217784881592, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.8441, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 6.305867671966553, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.8651, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 6.001807689666748, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 7.9054, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 7.388605117797852, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 7.9282, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 6.720639705657959, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.7044, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 7.121181488037109, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.5772, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 7.391552925109863, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.6082, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 7.045429706573486, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.6297, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 7.8674798011779785, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.66, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 6.961917400360107, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.3387, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 8.515259742736816, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.3089, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 7.693201065063477, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.3566, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15359431401628565, | |
| "eval_loss": 10.705141067504883, | |
| "eval_runtime": 0.844, | |
| "eval_samples_per_second": 567.552, | |
| "eval_steps_per_second": 5.924, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 8.267168045043945, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.3892, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 8.442320823669434, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.3808, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 8.887129783630371, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.0462, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 8.140633583068848, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.0862, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 8.852100372314453, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.1283, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 7.941941738128662, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.1476, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 7.973927021026611, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.0585, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 8.13227653503418, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.8528, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 8.01505184173584, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.8899, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 9.830168724060059, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.906, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 11.142940521240234, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.932, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 9.089313507080078, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.7877, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 10.703131675720215, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.6849, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 9.057072639465332, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.7096, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 8.93012809753418, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.7355, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 9.557744026184082, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.7622, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 8.993326187133789, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.5861, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 9.047318458557129, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.5706, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 8.779540061950684, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.5711, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 8.630488395690918, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.5959, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.1505270830996972, | |
| "eval_loss": 12.043688774108887, | |
| "eval_runtime": 0.8375, | |
| "eval_samples_per_second": 571.913, | |
| "eval_steps_per_second": 5.97, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 7.818565368652344, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.6065, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 6.938087463378906, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.4545, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 9.896014213562012, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.4651, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 10.248242378234863, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.4818, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 7.8425140380859375, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.4974, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 7.016345024108887, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.4735, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 6.346151828765869, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.3868, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 7.2288899421691895, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.4031, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 7.863675594329834, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.4178, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 9.660747528076172, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.4323, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 6.660861492156982, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.3739, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 7.965744495391846, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.3449, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 7.83579683303833, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.349, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 6.225755214691162, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.3539, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 9.147672653198242, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.3527, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 7.586123466491699, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.3229, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 6.96025276184082, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.3089, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.395577430725098, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.3008, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 6.091034889221191, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.2981, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 6.2691969871521, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.3072, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1513443799590658, | |
| "eval_loss": 12.816728591918945, | |
| "eval_runtime": 0.8406, | |
| "eval_samples_per_second": 569.829, | |
| "eval_steps_per_second": 5.948, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.214353561401367, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.27, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 8.669116020202637, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.2813, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 4.987499713897705, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.2685, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 7.48527717590332, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.2716, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 4.866922378540039, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.2639, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 4.931224346160889, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.2365, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 8.845723152160645, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.2401, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 7.354602813720703, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.2477, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 2.828876256942749, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.2301, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 5.047966957092285, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.2287, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 5.880908489227295, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.2133, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 5.750067234039307, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.2254, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 6.461637020111084, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.2067, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 5.363232612609863, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.2166, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.551758766174316, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.2057, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 8.446605682373047, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.2057, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 4.893612861633301, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.2076, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 6.084640026092529, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.2008, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 4.361387729644775, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.192, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.68356804521728e+17, | |
| "train_loss": 7.646758914822596, | |
| "train_runtime": 16728.5175, | |
| "train_samples_per_second": 341.69, | |
| "train_steps_per_second": 3.56 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 7.68356804521728e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |