| { | |
| "best_metric": 10.216060638427734, | |
| "best_model_checkpoint": "tiny_bert_rand_50_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": 2.5450780391693115, | |
| "learning_rate": 5e-06, | |
| "loss": 13.7115, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.7334048748016357, | |
| "learning_rate": 1e-05, | |
| "loss": 12.1301, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.4021859169006348, | |
| "learning_rate": 1.5e-05, | |
| "loss": 11.1362, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.5765750408172607, | |
| "learning_rate": 2e-05, | |
| "loss": 10.9361, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.6201908588409424, | |
| "learning_rate": 2.5e-05, | |
| "loss": 10.8203, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.630454182624817, | |
| "learning_rate": 3e-05, | |
| "loss": 10.713, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.744618535041809, | |
| "learning_rate": 3.5e-05, | |
| "loss": 10.6259, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.723514437675476, | |
| "learning_rate": 4e-05, | |
| "loss": 10.5438, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.5438804626464844, | |
| "learning_rate": 4.5e-05, | |
| "loss": 10.4702, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.6797691583633423, | |
| "learning_rate": 5e-05, | |
| "loss": 10.4047, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.7786786556243896, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.3572, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.7495299577713013, | |
| "learning_rate": 6e-05, | |
| "loss": 10.3256, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.6919372081756592, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.2989, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 1.9645551443099976, | |
| "learning_rate": 7e-05, | |
| "loss": 10.2753, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 1.7399874925613403, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.2173, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.064260482788086, | |
| "learning_rate": 8e-05, | |
| "loss": 10.1827, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.9603065252304077, | |
| "learning_rate": 8.5e-05, | |
| "loss": 10.173, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.2697181701660156, | |
| "learning_rate": 9e-05, | |
| "loss": 10.1495, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 1.9686516523361206, | |
| "learning_rate": 9.5e-05, | |
| "loss": 10.1368, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.6566529273986816, | |
| "learning_rate": 0.0001, | |
| "loss": 9.9989, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15000417373883526, | |
| "eval_loss": 10.216060638427734, | |
| "eval_runtime": 0.746, | |
| "eval_samples_per_second": 642.101, | |
| "eval_steps_per_second": 6.703, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.3557376861572266, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.9941, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.9733705520629883, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.9872, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.6935906410217285, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.9915, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.8545939922332764, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 9.9501, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.907075881958008, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 9.8025, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.890052318572998, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 9.8168, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.747251510620117, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 9.838, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 2.7311058044433594, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 9.8432, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.738133430480957, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 9.7665, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.793278932571411, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 9.6846, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.8773210048675537, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 9.7064, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.8377397060394287, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 9.7294, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 3.216353178024292, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 9.7439, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.007575750350952, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 9.6223, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.3819682598114014, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 9.5823, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.9600727558135986, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 9.6127, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.8249659538269043, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 9.6431, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.939486265182495, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 9.6554, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 3.0083022117614746, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 9.486, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.2319276332855225, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 9.4887, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15289430118690198, | |
| "eval_loss": 10.373096466064453, | |
| "eval_runtime": 0.7137, | |
| "eval_samples_per_second": 671.171, | |
| "eval_steps_per_second": 7.006, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 3.4256772994995117, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 9.5171, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 3.3967843055725098, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 9.5491, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 3.7140486240386963, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 9.5219, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 3.8211681842803955, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 9.3399, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 3.85943865776062, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 9.367, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 3.666921615600586, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 9.4054, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 4.093038082122803, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 9.4128, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 4.4062581062316895, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 9.3307, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 3.9779603481292725, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 9.1895, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 4.332785129547119, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 9.2071, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 4.290595054626465, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 9.2278, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 4.600621223449707, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 9.2559, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 4.6008992195129395, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 9.0898, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 5.450432777404785, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 8.9971, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 5.028412342071533, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 9.0267, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 4.975551128387451, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 9.0355, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 5.384194374084473, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 9.0672, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 5.409440994262695, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 8.8094, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 5.306190490722656, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 8.7776, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 5.64209508895874, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 8.8108, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15404202926938468, | |
| "eval_loss": 11.025629997253418, | |
| "eval_runtime": 0.7018, | |
| "eval_samples_per_second": 682.505, | |
| "eval_steps_per_second": 7.124, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 5.788949012756348, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 8.8334, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 6.211551666259766, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 8.8272, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 5.864919662475586, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 8.5295, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 6.621845722198486, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 8.568, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 7.034498691558838, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 8.6022, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 6.143002033233643, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 8.6212, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 6.473554611206055, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 8.5266, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 7.229272842407227, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 8.3185, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 6.569444179534912, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 8.3565, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 6.898689270019531, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 8.3952, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 7.376826763153076, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 8.4028, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 7.734011173248291, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 8.2321, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 7.37497091293335, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 8.1283, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 7.05466890335083, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 8.146, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 7.689860820770264, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 8.1809, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 7.437720775604248, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 8.2097, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 9.077725410461426, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 7.9712, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 8.8407564163208, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 7.9335, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 8.493143081665039, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 7.9638, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 8.477731704711914, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 7.9917, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15117191880677358, | |
| "eval_loss": 11.861244201660156, | |
| "eval_runtime": 0.7062, | |
| "eval_samples_per_second": 678.301, | |
| "eval_steps_per_second": 7.08, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 7.724287033081055, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 8.0123, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 9.805917739868164, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 7.7214, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 8.868997573852539, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 7.7517, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 9.577037811279297, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 7.7825, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 9.461718559265137, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 7.8145, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 9.81477165222168, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 7.7458, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 8.664056777954102, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 7.5666, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 9.285131454467773, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 7.5914, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 9.865168571472168, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 7.6308, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 9.430205345153809, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 7.6509, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 10.704974174499512, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 7.5339, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 10.935819625854492, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 7.4251, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 10.890119552612305, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 7.4528, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 10.795445442199707, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 7.4623, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 11.10928726196289, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 7.4811, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 10.465703010559082, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 7.3538, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 11.332059860229492, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 7.3001, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 9.759811401367188, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 7.3255, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 10.310357093811035, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 7.3336, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 9.50801944732666, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 7.3605, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1513724170802142, | |
| "eval_loss": 12.65954875946045, | |
| "eval_runtime": 0.7218, | |
| "eval_samples_per_second": 663.642, | |
| "eval_steps_per_second": 6.927, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 10.566963195800781, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 7.1795, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 9.286499977111816, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 7.1996, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 10.896336555480957, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 7.2202, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 10.798194885253906, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 7.2236, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 11.20406723022461, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 7.2053, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 8.228652954101562, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 7.1009, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 10.032462120056152, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 7.1028, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 11.267309188842773, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 7.1211, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 9.765129089355469, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 7.1175, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 9.495115280151367, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 7.0862, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 10.347193717956543, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 7.0205, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 8.845890045166016, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 7.0504, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 9.43069839477539, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 7.032, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 11.445097923278809, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 7.0434, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 10.131210327148438, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 7.0085, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 9.935945510864258, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.9998, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 10.405517578125, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.9961, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 9.528275489807129, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.9864, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 10.806137084960938, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.9943, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.053850510375936e+17, | |
| "train_loss": 8.781780386283355, | |
| "train_runtime": 12106.2258, | |
| "train_samples_per_second": 472.152, | |
| "train_steps_per_second": 4.919 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.053850510375936e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |