| { | |
| "best_global_step": 40000, | |
| "best_metric": 6.34233283996582, | |
| "best_model_checkpoint": "distilbert_km_10_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": 3.149277687072754, | |
| "learning_rate": 5e-06, | |
| "loss": 11.0764, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 10.052783966064453, | |
| "learning_rate": 1e-05, | |
| "loss": 8.9271, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 12.762539863586426, | |
| "learning_rate": 1.5e-05, | |
| "loss": 7.8887, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 9.455464363098145, | |
| "learning_rate": 2e-05, | |
| "loss": 7.5877, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 10.117177963256836, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.4111, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 5.13746976852417, | |
| "learning_rate": 3e-05, | |
| "loss": 7.264, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 5.037430286407471, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.1357, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 6.551547050476074, | |
| "learning_rate": 4e-05, | |
| "loss": 7.0394, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 4.965848445892334, | |
| "learning_rate": 4.5e-05, | |
| "loss": 6.9413, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 9.906851768493652, | |
| "learning_rate": 5e-05, | |
| "loss": 6.8396, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.100743293762207, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.7783, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.323625087738037, | |
| "learning_rate": 6e-05, | |
| "loss": 6.7508, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.165152072906494, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 6.7093, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 3.371307134628296, | |
| "learning_rate": 7e-05, | |
| "loss": 6.6854, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.754547595977783, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.6227, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 3.0923752784729004, | |
| "learning_rate": 8e-05, | |
| "loss": 6.6074, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.9533729553222656, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.6022, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 3.656068801879883, | |
| "learning_rate": 9e-05, | |
| "loss": 6.5778, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 3.9771640300750732, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.5609, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 4.5529093742370605, | |
| "learning_rate": 0.0001, | |
| "loss": 6.4927, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.1506441470269067, | |
| "eval_loss": 6.491031169891357, | |
| "eval_runtime": 0.8544, | |
| "eval_samples_per_second": 560.624, | |
| "eval_steps_per_second": 5.852, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.6148977279663086, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 6.4835, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 1.6763428449630737, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 6.4706, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.436123847961426, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 6.4602, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.2335009574890137, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.4301, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.339343786239624, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.3686, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 1.7537976503372192, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.3617, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.161440849304199, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.3623, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.6785426139831543, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.3539, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.707353115081787, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.3191, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 1.9640616178512573, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.282, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 1.7043923139572144, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.2776, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.7088178396224976, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.2801, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 1.3803937435150146, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.2774, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 1.2506041526794434, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.2285, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 1.847280740737915, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.2082, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 1.283583402633667, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.2143, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.1002233028411865, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.2163, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 1.969815969467163, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.2165, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.953533887863159, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.1615, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 2.5779690742492676, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.1606, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15275400544347484, | |
| "eval_loss": 6.36586332321167, | |
| "eval_runtime": 0.8473, | |
| "eval_samples_per_second": 565.309, | |
| "eval_steps_per_second": 5.901, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.233391761779785, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.1604, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 1.7626516819000244, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.1599, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 1.7346878051757812, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.1512, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 4.285078525543213, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.1127, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 2.8082990646362305, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.1129, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 1.5027289390563965, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.1127, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 3.909857749938965, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.1098, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 1.8172169923782349, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.0915, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 2.0356717109680176, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.0782, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.776805877685547, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.0715, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 3.1396679878234863, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.0709, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 1.6162575483322144, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.0715, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 1.911193609237671, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 6.0508, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 1.2606751918792725, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.0411, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 1.6268528699874878, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.0355, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 1.8731907606124878, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.0345, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 1.4184792041778564, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.0318, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 2.0361735820770264, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.0176, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 1.9261606931686401, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.0025, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 1.3379472494125366, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.0115, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15393010045610991, | |
| "eval_loss": 6.3590474128723145, | |
| "eval_runtime": 0.8436, | |
| "eval_samples_per_second": 567.777, | |
| "eval_steps_per_second": 5.927, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 1.470361590385437, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.0029, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 2.9312326908111572, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.0069, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 3.1739726066589355, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 5.9841, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 0.8212246298789978, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 5.981, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 2.419975996017456, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 5.986, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 3.096012830734253, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 5.9837, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 1.1732163429260254, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 5.9685, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 1.1891952753067017, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 5.96, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 2.274113178253174, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 5.9597, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 1.7185032367706299, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 5.9592, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 0.7538923621177673, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 5.957, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 2.0026018619537354, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 5.9454, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 2.45963454246521, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 5.9415, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 2.826068878173828, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 5.9412, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 1.6555607318878174, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 5.9408, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 2.398268222808838, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 5.9382, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 3.805772304534912, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 5.9281, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 1.98552405834198, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 5.9201, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 2.444162607192993, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 5.9244, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 1.7435388565063477, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 5.9142, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15268588090164853, | |
| "eval_loss": 6.34233283996582, | |
| "eval_runtime": 0.842, | |
| "eval_samples_per_second": 568.91, | |
| "eval_steps_per_second": 5.939, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 1.2505182027816772, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 5.9218, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 2.178734540939331, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 5.9079, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 1.8934943675994873, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 5.9035, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 0.8694837093353271, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 5.9055, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 2.6238772869110107, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 5.9041, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 0.9356751441955566, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 5.899, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 0.8301547169685364, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 5.8903, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 2.3180980682373047, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 5.8916, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 1.6560579538345337, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 5.8899, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 0.7703125476837158, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 5.8936, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 2.9054269790649414, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 5.8895, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 1.8457813262939453, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.8761, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 1.3675782680511475, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.8767, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 0.9035144448280334, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.88, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 0.7942680716514587, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.8762, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 0.887523353099823, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.8725, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 0.875932514667511, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.8685, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 0.8090535402297974, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.8635, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 2.0840654373168945, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.8623, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 2.67318058013916, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.8634, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1518210110185886, | |
| "eval_loss": 6.444693565368652, | |
| "eval_runtime": 0.8427, | |
| "eval_samples_per_second": 568.384, | |
| "eval_steps_per_second": 5.933, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 0.7596118450164795, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.8562, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 3.1897964477539062, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.8576, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 0.7993090152740479, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.8565, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 1.0190683603286743, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.8541, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 1.5026600360870361, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.8548, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 1.6342543363571167, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.8453, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 2.908343553543091, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.849, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 2.3473317623138428, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.8493, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 0.7943249344825745, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.8445, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 0.7953370809555054, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.8432, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 0.792066216468811, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.8423, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 1.0336476564407349, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.8414, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 0.8503936529159546, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.8356, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 1.2063177824020386, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.8387, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 1.4527183771133423, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.8359, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 0.7901241779327393, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.8373, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 1.3762155771255493, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.834, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 0.7735074162483215, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.8337, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 0.7774016261100769, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.8334, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.6822177215744e+17, | |
| "train_loss": 6.218850561441442, | |
| "train_runtime": 17090.279, | |
| "train_samples_per_second": 334.458, | |
| "train_steps_per_second": 3.484 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.6822177215744e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |