| { | |
| "best_metric": 6.156010150909424, | |
| "best_model_checkpoint": "distilbert_km_5_v1/checkpoint-40000", | |
| "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": 4.1612019538879395, | |
| "learning_rate": 5e-06, | |
| "loss": 10.8691, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 6.82954216003418, | |
| "learning_rate": 1e-05, | |
| "loss": 8.7487, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 4.282559394836426, | |
| "learning_rate": 1.5e-05, | |
| "loss": 7.702, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 7.227361679077148, | |
| "learning_rate": 2e-05, | |
| "loss": 7.4143, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 5.336795330047607, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.2642, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 5.528139591217041, | |
| "learning_rate": 3e-05, | |
| "loss": 7.1155, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.545119524002075, | |
| "learning_rate": 3.5e-05, | |
| "loss": 6.9876, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 3.4349753856658936, | |
| "learning_rate": 4e-05, | |
| "loss": 6.8787, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 3.931924819946289, | |
| "learning_rate": 4.5e-05, | |
| "loss": 6.7918, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 4.56055212020874, | |
| "learning_rate": 5e-05, | |
| "loss": 6.7054, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 3.6271419525146484, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.6404, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.8791913986206055, | |
| "learning_rate": 6e-05, | |
| "loss": 6.6166, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 4.116630554199219, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 6.581, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.3085134029388428, | |
| "learning_rate": 7e-05, | |
| "loss": 6.5549, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 4.115169048309326, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.4992, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.6390581130981445, | |
| "learning_rate": 8e-05, | |
| "loss": 6.4834, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.2927147150039673, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.4726, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.167675495147705, | |
| "learning_rate": 9e-05, | |
| "loss": 6.4557, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 1.181800365447998, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.4393, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 1.7572091817855835, | |
| "learning_rate": 0.0001, | |
| "loss": 6.3842, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.1512006455382732, | |
| "eval_loss": 6.3986101150512695, | |
| "eval_runtime": 0.8622, | |
| "eval_samples_per_second": 555.554, | |
| "eval_steps_per_second": 5.799, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 4.3407416343688965, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 6.3702, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 1.3264970779418945, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 6.35, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 1.8027888536453247, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 6.3464, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 1.1646184921264648, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.3202, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.508687734603882, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.2773, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 1.8887676000595093, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.2656, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 1.8515440225601196, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.2627, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.072016477584839, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.253, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 1.2400668859481812, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.2249, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.7712485790252686, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.2009, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.6809465885162354, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.1976, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.774334192276001, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.1888, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 0.7683253288269043, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.1836, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.3543953895568848, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.1536, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 0.9818401336669922, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.1376, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 1.9639158248901367, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.1384, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.449981451034546, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.1326, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.7381153106689453, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.1294, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 1.7951849699020386, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.0951, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.5764541625976562, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.092, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15272594629478942, | |
| "eval_loss": 6.29288911819458, | |
| "eval_runtime": 0.8373, | |
| "eval_samples_per_second": 572.044, | |
| "eval_steps_per_second": 5.971, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 1.5810127258300781, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.0866, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 1.2175865173339844, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.0867, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.5673795938491821, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.0745, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 1.137192726135254, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.0526, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 1.5745199918746948, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.0481, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 1.9315625429153442, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.0462, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 2.34018874168396, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.039, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 1.927125096321106, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.0236, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 1.6729779243469238, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.0184, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.282742977142334, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.0077, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 2.206977367401123, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.0073, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 1.3579425811767578, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.008, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 1.0642881393432617, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 5.9908, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 3.1353816986083984, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 5.9835, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 1.7503684759140015, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 5.9761, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 1.294704794883728, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 5.9729, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 1.7032661437988281, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 5.9715, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 1.8815373182296753, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 5.9607, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 1.7230100631713867, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 5.9519, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 1.5406538248062134, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 5.9558, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15412597587934074, | |
| "eval_loss": 6.241944313049316, | |
| "eval_runtime": 0.8313, | |
| "eval_samples_per_second": 576.23, | |
| "eval_steps_per_second": 6.015, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 1.0840150117874146, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 5.9474, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 3.1964223384857178, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 5.9478, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 0.9289094805717468, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 5.9339, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 2.1706831455230713, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 5.9279, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 1.7513415813446045, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 5.9322, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 0.8191587328910828, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 5.9275, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 1.4774624109268188, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 5.9204, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 1.1845673322677612, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 5.9113, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 1.7402995824813843, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 5.9093, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 1.0856926441192627, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 5.9062, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 0.9018107652664185, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 5.907, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 0.8164610266685486, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 5.8968, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 1.8844313621520996, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 5.8954, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 0.8575370907783508, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 5.8887, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 1.7213451862335205, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 5.8932, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 1.2610878944396973, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 5.8888, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 1.959447979927063, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 5.8824, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 0.7246803045272827, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 5.8749, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 1.3790050745010376, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 5.8792, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 2.10178279876709, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 5.869, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.1524896265560166, | |
| "eval_loss": 6.156010150909424, | |
| "eval_runtime": 0.8057, | |
| "eval_samples_per_second": 594.537, | |
| "eval_steps_per_second": 6.206, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 0.7790839672088623, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 5.8727, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 1.1335006952285767, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 5.8619, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 0.7718275785446167, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 5.8582, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 0.7538554668426514, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 5.8604, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 0.7654014825820923, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 5.8582, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 2.793358564376831, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 5.8547, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 2.503340482711792, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 5.8462, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 0.8580361604690552, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 5.8478, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 2.124924898147583, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 5.8469, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 0.751645028591156, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 5.8483, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 0.8310075998306274, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 5.8457, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 1.0850948095321655, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.8328, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 0.805169403553009, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.835, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 0.741425633430481, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.8372, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 0.7873342633247375, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.8326, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 0.7500522136688232, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.828, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 1.0403913259506226, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.8267, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 1.131286382675171, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.8225, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 2.4620168209075928, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.822, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 0.7604252099990845, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.823, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.15260605041074382, | |
| "eval_loss": 6.245114803314209, | |
| "eval_runtime": 0.8126, | |
| "eval_samples_per_second": 589.435, | |
| "eval_steps_per_second": 6.153, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 0.7925912737846375, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.8153, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 2.449723243713379, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.8188, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 0.774408757686615, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.8167, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 0.7786117792129517, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.8146, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 0.7834874987602234, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.8137, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 0.772318959236145, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.8033, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 4.681053638458252, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.808, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 0.7849641442298889, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.8104, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 0.7843820452690125, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.8071, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 1.6776659488677979, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.8038, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 0.7917454838752747, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.8029, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 0.7981336712837219, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.8019, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 0.76115882396698, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.7972, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 2.4491937160491943, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.8004, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 1.4324352741241455, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.7974, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 0.7930957674980164, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.7983, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 0.7850443720817566, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.7951, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 0.763573944568634, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.796, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 0.7744625210762024, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.7949, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.68154255975296e+17, | |
| "train_loss": 6.146615118864301, | |
| "train_runtime": 16607.646, | |
| "train_samples_per_second": 344.177, | |
| "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 | |
| } | |