| { | |
| "best_metric": 8.46003246307373, | |
| "best_model_checkpoint": "distilbert_rand_10_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": 2.4927899837493896, | |
| "learning_rate": 5e-06, | |
| "loss": 11.5885, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.0215930938720703, | |
| "learning_rate": 1e-05, | |
| "loss": 9.8862, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 2.182110548019409, | |
| "learning_rate": 1.5e-05, | |
| "loss": 9.3637, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 2.431750774383545, | |
| "learning_rate": 2e-05, | |
| "loss": 9.206, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 2.031982660293579, | |
| "learning_rate": 2.5e-05, | |
| "loss": 9.0721, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 2.7520811557769775, | |
| "learning_rate": 3e-05, | |
| "loss": 8.9619, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.80153751373291, | |
| "learning_rate": 3.5e-05, | |
| "loss": 8.8669, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.5652287006378174, | |
| "learning_rate": 4e-05, | |
| "loss": 8.7908, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.451920509338379, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.7325, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.2890987396240234, | |
| "learning_rate": 5e-05, | |
| "loss": 8.6828, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.507328510284424, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 8.6409, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 3.7738049030303955, | |
| "learning_rate": 6e-05, | |
| "loss": 8.6046, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.195826530456543, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 8.5716, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.414384126663208, | |
| "learning_rate": 7e-05, | |
| "loss": 8.5421, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.428041934967041, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 8.5097, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.7345211505889893, | |
| "learning_rate": 8e-05, | |
| "loss": 8.4853, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.8851404190063477, | |
| "learning_rate": 8.5e-05, | |
| "loss": 8.4699, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.1663172245025635, | |
| "learning_rate": 9e-05, | |
| "loss": 8.4463, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.784700393676758, | |
| "learning_rate": 9.5e-05, | |
| "loss": 8.43, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.117123603820801, | |
| "learning_rate": 0.0001, | |
| "loss": 8.3584, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15103369598486324, | |
| "eval_loss": 8.46003246307373, | |
| "eval_runtime": 0.8443, | |
| "eval_samples_per_second": 567.319, | |
| "eval_steps_per_second": 5.922, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 3.2652812004089355, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 8.3534, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.7463793754577637, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 8.3438, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.1724801063537598, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 8.3417, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.1871085166931152, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 8.3149, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.2624504566192627, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 8.2279, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.360450506210327, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 8.2371, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 3.333432197570801, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 8.2449, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 2.496487617492676, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 8.2435, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 3.252718687057495, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 8.1937, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 3.17020583152771, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 8.1514, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 3.7368528842926025, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 8.1531, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.477617025375366, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 8.1655, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.5955793857574463, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 8.1652, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.979802370071411, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 8.096, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.0802161693573, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 8.0685, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.9352760314941406, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 8.0833, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 3.308439254760742, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 8.0942, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.2638556957244873, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 8.0981, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 3.7358028888702393, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 7.9756, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.7296955585479736, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 7.9814, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15199640842896828, | |
| "eval_loss": 8.64383602142334, | |
| "eval_runtime": 0.8861, | |
| "eval_samples_per_second": 540.589, | |
| "eval_steps_per_second": 5.643, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.983659505844116, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 7.9862, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 3.557201623916626, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 8.0064, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 3.330355167388916, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 7.9738, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 4.105613708496094, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 7.8049, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 4.690443992614746, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 7.8261, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 3.6583445072174072, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 7.8455, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 4.262939453125, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 7.8464, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 5.28586483001709, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 7.7498, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 4.585413455963135, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.5914, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 4.890716075897217, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.5985, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 4.976234436035156, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 7.6165, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 4.914450168609619, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 7.6404, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 5.943147659301758, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.4532, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 6.186162948608398, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.3528, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 5.770488262176514, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.3841, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 5.105766296386719, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.3907, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 6.028867244720459, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.4082, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 6.533329963684082, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.1609, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 5.701409816741943, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.1313, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 6.67290735244751, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.1696, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15283879452668103, | |
| "eval_loss": 9.59500789642334, | |
| "eval_runtime": 0.828, | |
| "eval_samples_per_second": 578.528, | |
| "eval_steps_per_second": 6.039, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 6.489418029785156, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.1845, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 6.281386852264404, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.1887, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 5.988144874572754, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.9249, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 7.924295425415039, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.9612, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 6.806079387664795, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.9916, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 6.670299053192139, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.9995, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 6.499294757843018, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.9208, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 7.2796950340271, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.7604, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 7.774478435516357, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.791, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 7.771714687347412, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.8145, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 10.217869758605957, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.8265, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 6.056733131408691, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.7035, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 8.028042793273926, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.6196, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 7.373075485229492, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.6417, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 7.532264709472656, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.666, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 8.606426239013672, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.677, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 8.520004272460938, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.5156, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 8.703201293945312, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.5058, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 8.758565902709961, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.5102, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 8.765710830688477, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.5265, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15209711786475272, | |
| "eval_loss": 10.72950553894043, | |
| "eval_runtime": 0.848, | |
| "eval_samples_per_second": 564.884, | |
| "eval_steps_per_second": 5.896, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 6.648326873779297, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.5478, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 5.784792423248291, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.3871, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 9.52237606048584, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.4003, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 11.224041938781738, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.4142, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 7.79087495803833, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.4266, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 7.79288387298584, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.3984, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 8.119388580322266, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.3205, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 7.012115478515625, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.331, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 7.084027290344238, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.3442, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 8.089616775512695, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.3588, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 7.139738082885742, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.3043, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 8.824871063232422, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.269, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 7.60262393951416, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.2664, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 7.534590721130371, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.2751, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 7.9743332862854, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.2809, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 9.115961074829102, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.2493, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 7.903227806091309, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.2299, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.601058483123779, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.2262, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 4.822947025299072, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.2277, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 4.8282012939453125, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.2316, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.15156867692825302, | |
| "eval_loss": 11.646409034729004, | |
| "eval_runtime": 0.8365, | |
| "eval_samples_per_second": 572.621, | |
| "eval_steps_per_second": 5.977, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.716351509094238, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.191, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 4.851901531219482, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.2023, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 6.435737609863281, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.1946, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 9.239810943603516, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.1943, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 4.5796098709106445, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.189, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 8.577524185180664, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.1568, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 8.177299499511719, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.1611, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 5.332095146179199, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.1694, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 2.616030693054199, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.1576, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 4.092552661895752, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.1503, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 5.155515670776367, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.1419, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 4.855210304260254, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.1485, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 4.516139030456543, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.1345, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 6.820849895477295, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.1421, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.8821187019348145, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.1281, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 6.711137771606445, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.1326, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 4.263157844543457, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.1313, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 4.636920928955078, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.1238, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 4.23433256149292, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.1186, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.6822177215744e+17, | |
| "train_loss": 7.33399917000387, | |
| "train_runtime": 17090.9511, | |
| "train_samples_per_second": 334.445, | |
| "train_steps_per_second": 3.484 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 7.6822177215744e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |