| { | |
| "best_metric": 0.7337000966072083, | |
| "best_model_checkpoint": "/home/ubuntu/utah/babylm-24/src/evaluation/results/finetune/DebertaV2-Base-10M_babylm-A/mnli/checkpoint-147264", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 245440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010185788787483703, | |
| "grad_norm": 3.972818613052368, | |
| "learning_rate": 2.9938885267275102e-05, | |
| "loss": 1.1058, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.020371577574967405, | |
| "grad_norm": 3.13022780418396, | |
| "learning_rate": 2.9877770534550196e-05, | |
| "loss": 1.0388, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.03055736636245111, | |
| "grad_norm": 5.194879055023193, | |
| "learning_rate": 2.9816655801825297e-05, | |
| "loss": 0.999, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.04074315514993481, | |
| "grad_norm": 6.858974933624268, | |
| "learning_rate": 2.975554106910039e-05, | |
| "loss": 0.9637, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.050928943937418515, | |
| "grad_norm": 7.074331760406494, | |
| "learning_rate": 2.9694426336375492e-05, | |
| "loss": 0.9697, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.06111473272490222, | |
| "grad_norm": 5.386518478393555, | |
| "learning_rate": 2.9633311603650586e-05, | |
| "loss": 0.9472, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.07130052151238592, | |
| "grad_norm": 5.582085609436035, | |
| "learning_rate": 2.9572196870925687e-05, | |
| "loss": 0.9273, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.08148631029986962, | |
| "grad_norm": 10.966410636901855, | |
| "learning_rate": 2.951108213820078e-05, | |
| "loss": 0.9316, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.09167209908735333, | |
| "grad_norm": 7.840672492980957, | |
| "learning_rate": 2.9449967405475882e-05, | |
| "loss": 0.9237, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.10185788787483703, | |
| "grad_norm": 7.442521095275879, | |
| "learning_rate": 2.9388852672750976e-05, | |
| "loss": 0.918, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.11204367666232073, | |
| "grad_norm": 8.379261016845703, | |
| "learning_rate": 2.9327737940026077e-05, | |
| "loss": 0.8798, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.12222946544980444, | |
| "grad_norm": 6.880769729614258, | |
| "learning_rate": 2.926662320730117e-05, | |
| "loss": 0.9045, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.13241525423728814, | |
| "grad_norm": 5.965817451477051, | |
| "learning_rate": 2.9205508474576272e-05, | |
| "loss": 0.8941, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.14260104302477183, | |
| "grad_norm": 9.67165756225586, | |
| "learning_rate": 2.914439374185137e-05, | |
| "loss": 0.878, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.15278683181225555, | |
| "grad_norm": 6.4934892654418945, | |
| "learning_rate": 2.9083279009126467e-05, | |
| "loss": 0.8723, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.16297262059973924, | |
| "grad_norm": 11.987072944641113, | |
| "learning_rate": 2.9022164276401565e-05, | |
| "loss": 0.8913, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.17315840938722293, | |
| "grad_norm": 9.218280792236328, | |
| "learning_rate": 2.8961049543676662e-05, | |
| "loss": 0.868, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.18334419817470665, | |
| "grad_norm": 6.697582244873047, | |
| "learning_rate": 2.8899934810951763e-05, | |
| "loss": 0.8618, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.19352998696219034, | |
| "grad_norm": 8.035455703735352, | |
| "learning_rate": 2.8838820078226858e-05, | |
| "loss": 0.8641, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.20371577574967406, | |
| "grad_norm": 6.890414237976074, | |
| "learning_rate": 2.877770534550196e-05, | |
| "loss": 0.8574, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.21390156453715775, | |
| "grad_norm": 9.001678466796875, | |
| "learning_rate": 2.8716590612777053e-05, | |
| "loss": 0.8547, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.22408735332464147, | |
| "grad_norm": 7.271191120147705, | |
| "learning_rate": 2.8655475880052154e-05, | |
| "loss": 0.8332, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.23427314211212516, | |
| "grad_norm": 9.849822998046875, | |
| "learning_rate": 2.8594361147327248e-05, | |
| "loss": 0.8447, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.24445893089960888, | |
| "grad_norm": 15.07829761505127, | |
| "learning_rate": 2.853324641460235e-05, | |
| "loss": 0.8251, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.25464471968709257, | |
| "grad_norm": 6.886602878570557, | |
| "learning_rate": 2.8472131681877446e-05, | |
| "loss": 0.8444, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.2648305084745763, | |
| "grad_norm": 9.94825267791748, | |
| "learning_rate": 2.8411016949152544e-05, | |
| "loss": 0.8504, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.27501629726205995, | |
| "grad_norm": 8.0740966796875, | |
| "learning_rate": 2.834990221642764e-05, | |
| "loss": 0.8182, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.28520208604954367, | |
| "grad_norm": 10.770129203796387, | |
| "learning_rate": 2.828878748370274e-05, | |
| "loss": 0.8373, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.2953878748370274, | |
| "grad_norm": 9.214078903198242, | |
| "learning_rate": 2.8227672750977836e-05, | |
| "loss": 0.8089, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.3055736636245111, | |
| "grad_norm": 14.231884956359863, | |
| "learning_rate": 2.8166558018252934e-05, | |
| "loss": 0.8337, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.31575945241199477, | |
| "grad_norm": 9.379185676574707, | |
| "learning_rate": 2.810544328552803e-05, | |
| "loss": 0.8315, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.3259452411994785, | |
| "grad_norm": 8.370843887329102, | |
| "learning_rate": 2.804432855280313e-05, | |
| "loss": 0.808, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.3361310299869622, | |
| "grad_norm": 7.201444149017334, | |
| "learning_rate": 2.7983213820078227e-05, | |
| "loss": 0.808, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.34631681877444587, | |
| "grad_norm": 10.696725845336914, | |
| "learning_rate": 2.7922099087353324e-05, | |
| "loss": 0.813, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.3565026075619296, | |
| "grad_norm": 6.316204071044922, | |
| "learning_rate": 2.7860984354628422e-05, | |
| "loss": 0.8213, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.3666883963494133, | |
| "grad_norm": 14.780086517333984, | |
| "learning_rate": 2.779986962190352e-05, | |
| "loss": 0.8165, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.376874185136897, | |
| "grad_norm": 11.34945011138916, | |
| "learning_rate": 2.773875488917862e-05, | |
| "loss": 0.7943, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.3870599739243807, | |
| "grad_norm": 7.136452674865723, | |
| "learning_rate": 2.7677640156453718e-05, | |
| "loss": 0.8017, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.3972457627118644, | |
| "grad_norm": 6.209742069244385, | |
| "learning_rate": 2.7616525423728815e-05, | |
| "loss": 0.7941, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.4074315514993481, | |
| "grad_norm": 8.923270225524902, | |
| "learning_rate": 2.7555410691003913e-05, | |
| "loss": 0.7826, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.41761734028683184, | |
| "grad_norm": 6.371203899383545, | |
| "learning_rate": 2.749429595827901e-05, | |
| "loss": 0.7938, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.4278031290743155, | |
| "grad_norm": 8.679354667663574, | |
| "learning_rate": 2.7433181225554108e-05, | |
| "loss": 0.7936, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.4379889178617992, | |
| "grad_norm": 5.962934970855713, | |
| "learning_rate": 2.7372066492829205e-05, | |
| "loss": 0.7924, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.44817470664928294, | |
| "grad_norm": 11.835200309753418, | |
| "learning_rate": 2.7310951760104303e-05, | |
| "loss": 0.7894, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.4583604954367666, | |
| "grad_norm": 9.62806510925293, | |
| "learning_rate": 2.72498370273794e-05, | |
| "loss": 0.7963, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.4685462842242503, | |
| "grad_norm": 7.6390509605407715, | |
| "learning_rate": 2.7188722294654498e-05, | |
| "loss": 0.783, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.47873207301173404, | |
| "grad_norm": 5.826275825500488, | |
| "learning_rate": 2.7127607561929596e-05, | |
| "loss": 0.7859, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.48891786179921776, | |
| "grad_norm": 5.199416637420654, | |
| "learning_rate": 2.7066492829204693e-05, | |
| "loss": 0.7989, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.4991036505867014, | |
| "grad_norm": 12.38355541229248, | |
| "learning_rate": 2.7005378096479794e-05, | |
| "loss": 0.7864, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.5092894393741851, | |
| "grad_norm": 7.700937271118164, | |
| "learning_rate": 2.694426336375489e-05, | |
| "loss": 0.7747, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.5194752281616688, | |
| "grad_norm": 10.8720703125, | |
| "learning_rate": 2.688314863102999e-05, | |
| "loss": 0.7974, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.5296610169491526, | |
| "grad_norm": 6.260867595672607, | |
| "learning_rate": 2.6822033898305083e-05, | |
| "loss": 0.7914, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.5398468057366362, | |
| "grad_norm": 11.113471984863281, | |
| "learning_rate": 2.6760919165580184e-05, | |
| "loss": 0.782, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.5500325945241199, | |
| "grad_norm": 7.497796058654785, | |
| "learning_rate": 2.6699804432855282e-05, | |
| "loss": 0.7642, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.5602183833116037, | |
| "grad_norm": 6.1969170570373535, | |
| "learning_rate": 2.663868970013038e-05, | |
| "loss": 0.7693, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.5704041720990873, | |
| "grad_norm": 9.965928077697754, | |
| "learning_rate": 2.6577574967405477e-05, | |
| "loss": 0.7681, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.5805899608865711, | |
| "grad_norm": 5.052292346954346, | |
| "learning_rate": 2.6516460234680575e-05, | |
| "loss": 0.7664, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.5907757496740548, | |
| "grad_norm": 6.7289276123046875, | |
| "learning_rate": 2.6455345501955672e-05, | |
| "loss": 0.7445, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.6009615384615384, | |
| "grad_norm": 8.478078842163086, | |
| "learning_rate": 2.639423076923077e-05, | |
| "loss": 0.7859, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.6111473272490222, | |
| "grad_norm": 6.2336106300354, | |
| "learning_rate": 2.6333116036505867e-05, | |
| "loss": 0.7598, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.6213331160365059, | |
| "grad_norm": 11.82295036315918, | |
| "learning_rate": 2.6272001303780965e-05, | |
| "loss": 0.7507, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.6315189048239895, | |
| "grad_norm": 8.545003890991211, | |
| "learning_rate": 2.6210886571056066e-05, | |
| "loss": 0.7769, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.6417046936114733, | |
| "grad_norm": 9.849065780639648, | |
| "learning_rate": 2.614977183833116e-05, | |
| "loss": 0.7781, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.651890482398957, | |
| "grad_norm": 7.641623020172119, | |
| "learning_rate": 2.608865710560626e-05, | |
| "loss": 0.7789, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.6620762711864406, | |
| "grad_norm": 16.397785186767578, | |
| "learning_rate": 2.6027542372881355e-05, | |
| "loss": 0.7542, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.6722620599739244, | |
| "grad_norm": 11.119671821594238, | |
| "learning_rate": 2.5966427640156456e-05, | |
| "loss": 0.7716, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.6824478487614081, | |
| "grad_norm": 12.466675758361816, | |
| "learning_rate": 2.590531290743155e-05, | |
| "loss": 0.7743, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.6926336375488917, | |
| "grad_norm": 8.151625633239746, | |
| "learning_rate": 2.584419817470665e-05, | |
| "loss": 0.7581, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.7028194263363755, | |
| "grad_norm": 7.743143558502197, | |
| "learning_rate": 2.5783083441981745e-05, | |
| "loss": 0.7599, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.7130052151238592, | |
| "grad_norm": 5.249680995941162, | |
| "learning_rate": 2.5721968709256846e-05, | |
| "loss": 0.7483, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.7231910039113429, | |
| "grad_norm": 6.738178730010986, | |
| "learning_rate": 2.566085397653194e-05, | |
| "loss": 0.7781, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.7333767926988266, | |
| "grad_norm": 6.7923102378845215, | |
| "learning_rate": 2.559973924380704e-05, | |
| "loss": 0.7464, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.7435625814863103, | |
| "grad_norm": 15.575695991516113, | |
| "learning_rate": 2.553862451108214e-05, | |
| "loss": 0.7571, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.753748370273794, | |
| "grad_norm": 13.128719329833984, | |
| "learning_rate": 2.5477509778357236e-05, | |
| "loss": 0.7625, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.7639341590612777, | |
| "grad_norm": 4.938621997833252, | |
| "learning_rate": 2.5416395045632337e-05, | |
| "loss": 0.7477, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.7741199478487614, | |
| "grad_norm": 8.46985912322998, | |
| "learning_rate": 2.535528031290743e-05, | |
| "loss": 0.7419, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.7843057366362451, | |
| "grad_norm": 18.881208419799805, | |
| "learning_rate": 2.5294165580182532e-05, | |
| "loss": 0.7403, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.7944915254237288, | |
| "grad_norm": 9.986156463623047, | |
| "learning_rate": 2.5233050847457626e-05, | |
| "loss": 0.7508, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.8046773142112125, | |
| "grad_norm": 11.625401496887207, | |
| "learning_rate": 2.5171936114732727e-05, | |
| "loss": 0.7426, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.8148631029986962, | |
| "grad_norm": 9.926522254943848, | |
| "learning_rate": 2.511082138200782e-05, | |
| "loss": 0.7365, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.8250488917861799, | |
| "grad_norm": 8.104277610778809, | |
| "learning_rate": 2.5049706649282922e-05, | |
| "loss": 0.7532, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.8352346805736637, | |
| "grad_norm": 5.898177623748779, | |
| "learning_rate": 2.4988591916558017e-05, | |
| "loss": 0.7657, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.8454204693611473, | |
| "grad_norm": 6.856070041656494, | |
| "learning_rate": 2.4927477183833118e-05, | |
| "loss": 0.7438, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.855606258148631, | |
| "grad_norm": 11.806628227233887, | |
| "learning_rate": 2.4866362451108212e-05, | |
| "loss": 0.7348, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.8657920469361148, | |
| "grad_norm": 12.08478832244873, | |
| "learning_rate": 2.4805247718383313e-05, | |
| "loss": 0.7277, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.8759778357235984, | |
| "grad_norm": 5.707648277282715, | |
| "learning_rate": 2.474413298565841e-05, | |
| "loss": 0.7458, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.8861636245110821, | |
| "grad_norm": 5.36006498336792, | |
| "learning_rate": 2.4683018252933508e-05, | |
| "loss": 0.7446, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.8963494132985659, | |
| "grad_norm": 6.603504180908203, | |
| "learning_rate": 2.4621903520208605e-05, | |
| "loss": 0.7278, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.9065352020860495, | |
| "grad_norm": 7.789953231811523, | |
| "learning_rate": 2.4560788787483703e-05, | |
| "loss": 0.7375, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.9167209908735332, | |
| "grad_norm": 11.424116134643555, | |
| "learning_rate": 2.4499674054758804e-05, | |
| "loss": 0.7297, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.926906779661017, | |
| "grad_norm": 6.7079620361328125, | |
| "learning_rate": 2.4438559322033898e-05, | |
| "loss": 0.7411, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.9370925684485006, | |
| "grad_norm": 11.931846618652344, | |
| "learning_rate": 2.4377444589309e-05, | |
| "loss": 0.749, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.9472783572359843, | |
| "grad_norm": 8.87076473236084, | |
| "learning_rate": 2.4316329856584093e-05, | |
| "loss": 0.7312, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.9574641460234681, | |
| "grad_norm": 12.787999153137207, | |
| "learning_rate": 2.4255215123859194e-05, | |
| "loss": 0.7244, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.9676499348109517, | |
| "grad_norm": 6.917387962341309, | |
| "learning_rate": 2.4194100391134288e-05, | |
| "loss": 0.733, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.9778357235984355, | |
| "grad_norm": 7.564793109893799, | |
| "learning_rate": 2.413298565840939e-05, | |
| "loss": 0.7253, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.9880215123859192, | |
| "grad_norm": 7.665174961090088, | |
| "learning_rate": 2.4071870925684483e-05, | |
| "loss": 0.7191, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.9982073011734028, | |
| "grad_norm": 9.235818862915039, | |
| "learning_rate": 2.4010756192959584e-05, | |
| "loss": 0.7288, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6841890811920166, | |
| "eval_loss": 0.7356013059616089, | |
| "eval_runtime": 8.0037, | |
| "eval_samples_per_second": 613.213, | |
| "eval_steps_per_second": 76.714, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 1.0083930899608866, | |
| "grad_norm": 9.232441902160645, | |
| "learning_rate": 2.3949641460234682e-05, | |
| "loss": 0.666, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 1.0185788787483703, | |
| "grad_norm": 5.098124980926514, | |
| "learning_rate": 2.388852672750978e-05, | |
| "loss": 0.6762, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 1.028764667535854, | |
| "grad_norm": 6.507536888122559, | |
| "learning_rate": 2.3827411994784877e-05, | |
| "loss": 0.6509, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 1.0389504563233376, | |
| "grad_norm": 12.516643524169922, | |
| "learning_rate": 2.3766297262059974e-05, | |
| "loss": 0.6569, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.0491362451108215, | |
| "grad_norm": 11.644853591918945, | |
| "learning_rate": 2.3705182529335072e-05, | |
| "loss": 0.6498, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 1.0593220338983051, | |
| "grad_norm": 15.303723335266113, | |
| "learning_rate": 2.364406779661017e-05, | |
| "loss": 0.6565, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 1.0695078226857888, | |
| "grad_norm": 11.212811470031738, | |
| "learning_rate": 2.3582953063885267e-05, | |
| "loss": 0.6492, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 1.0796936114732725, | |
| "grad_norm": 7.695071220397949, | |
| "learning_rate": 2.3521838331160365e-05, | |
| "loss": 0.6618, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 1.0898794002607561, | |
| "grad_norm": 8.234328269958496, | |
| "learning_rate": 2.3460723598435462e-05, | |
| "loss": 0.6709, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 1.1000651890482398, | |
| "grad_norm": 12.574545860290527, | |
| "learning_rate": 2.339960886571056e-05, | |
| "loss": 0.6518, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.1102509778357237, | |
| "grad_norm": 5.914109706878662, | |
| "learning_rate": 2.333849413298566e-05, | |
| "loss": 0.6494, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 1.1204367666232073, | |
| "grad_norm": 11.49263858795166, | |
| "learning_rate": 2.3277379400260758e-05, | |
| "loss": 0.6723, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 1.130622555410691, | |
| "grad_norm": 11.228019714355469, | |
| "learning_rate": 2.3216264667535856e-05, | |
| "loss": 0.6328, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 1.1408083441981747, | |
| "grad_norm": 14.478355407714844, | |
| "learning_rate": 2.3155149934810953e-05, | |
| "loss": 0.667, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 1.1509941329856583, | |
| "grad_norm": 18.817468643188477, | |
| "learning_rate": 2.309403520208605e-05, | |
| "loss": 0.6674, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 1.161179921773142, | |
| "grad_norm": 12.483678817749023, | |
| "learning_rate": 2.303292046936115e-05, | |
| "loss": 0.6677, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 1.1713657105606259, | |
| "grad_norm": 7.133495807647705, | |
| "learning_rate": 2.2971805736636246e-05, | |
| "loss": 0.642, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 1.1815514993481095, | |
| "grad_norm": 7.099282741546631, | |
| "learning_rate": 2.2910691003911343e-05, | |
| "loss": 0.664, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 1.1917372881355932, | |
| "grad_norm": 9.188867568969727, | |
| "learning_rate": 2.284957627118644e-05, | |
| "loss": 0.6703, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 1.2019230769230769, | |
| "grad_norm": 5.470687389373779, | |
| "learning_rate": 2.278846153846154e-05, | |
| "loss": 0.6575, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 1.2121088657105605, | |
| "grad_norm": 8.013532638549805, | |
| "learning_rate": 2.2727346805736636e-05, | |
| "loss": 0.6465, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 1.2222946544980444, | |
| "grad_norm": 8.846539497375488, | |
| "learning_rate": 2.2666232073011734e-05, | |
| "loss": 0.6682, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 1.232480443285528, | |
| "grad_norm": 11.2225341796875, | |
| "learning_rate": 2.260511734028683e-05, | |
| "loss": 0.653, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 1.2426662320730117, | |
| "grad_norm": 12.363032341003418, | |
| "learning_rate": 2.254400260756193e-05, | |
| "loss": 0.6807, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 1.2528520208604954, | |
| "grad_norm": 8.01014232635498, | |
| "learning_rate": 2.248288787483703e-05, | |
| "loss": 0.6625, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 1.263037809647979, | |
| "grad_norm": 20.960845947265625, | |
| "learning_rate": 2.2421773142112124e-05, | |
| "loss": 0.6524, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 1.2732235984354627, | |
| "grad_norm": 13.620718002319336, | |
| "learning_rate": 2.2360658409387225e-05, | |
| "loss": 0.6625, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 1.2834093872229466, | |
| "grad_norm": 18.55965232849121, | |
| "learning_rate": 2.2299543676662322e-05, | |
| "loss": 0.6434, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 1.2935951760104303, | |
| "grad_norm": 12.919173240661621, | |
| "learning_rate": 2.223842894393742e-05, | |
| "loss": 0.652, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 1.303780964797914, | |
| "grad_norm": 19.561386108398438, | |
| "learning_rate": 2.2177314211212517e-05, | |
| "loss": 0.6498, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 1.3139667535853976, | |
| "grad_norm": 10.529995918273926, | |
| "learning_rate": 2.2116199478487615e-05, | |
| "loss": 0.666, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 1.3241525423728815, | |
| "grad_norm": 13.268050193786621, | |
| "learning_rate": 2.2055084745762713e-05, | |
| "loss": 0.6499, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 1.3343383311603652, | |
| "grad_norm": 6.3072285652160645, | |
| "learning_rate": 2.199397001303781e-05, | |
| "loss": 0.6798, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 1.3445241199478488, | |
| "grad_norm": 5.304599761962891, | |
| "learning_rate": 2.1932855280312908e-05, | |
| "loss": 0.6559, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 1.3547099087353325, | |
| "grad_norm": 13.698792457580566, | |
| "learning_rate": 2.1871740547588005e-05, | |
| "loss": 0.6441, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 1.3648956975228161, | |
| "grad_norm": 8.951632499694824, | |
| "learning_rate": 2.1810625814863103e-05, | |
| "loss": 0.6564, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 1.3750814863102998, | |
| "grad_norm": 5.84067440032959, | |
| "learning_rate": 2.17495110821382e-05, | |
| "loss": 0.6686, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 1.3852672750977835, | |
| "grad_norm": 6.442080497741699, | |
| "learning_rate": 2.16883963494133e-05, | |
| "loss": 0.6517, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 1.3954530638852674, | |
| "grad_norm": 5.514212608337402, | |
| "learning_rate": 2.1627281616688395e-05, | |
| "loss": 0.6664, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 1.405638852672751, | |
| "grad_norm": 14.150158882141113, | |
| "learning_rate": 2.1566166883963496e-05, | |
| "loss": 0.6561, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 1.4158246414602347, | |
| "grad_norm": 17.196884155273438, | |
| "learning_rate": 2.150505215123859e-05, | |
| "loss": 0.6513, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 1.4260104302477183, | |
| "grad_norm": 6.181870937347412, | |
| "learning_rate": 2.144393741851369e-05, | |
| "loss": 0.65, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 1.436196219035202, | |
| "grad_norm": 14.865707397460938, | |
| "learning_rate": 2.1382822685788786e-05, | |
| "loss": 0.6596, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 1.4463820078226859, | |
| "grad_norm": 15.808574676513672, | |
| "learning_rate": 2.1321707953063886e-05, | |
| "loss": 0.6559, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 1.4565677966101696, | |
| "grad_norm": 16.76003074645996, | |
| "learning_rate": 2.1260593220338984e-05, | |
| "loss": 0.6467, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 1.4667535853976532, | |
| "grad_norm": 21.466825485229492, | |
| "learning_rate": 2.119947848761408e-05, | |
| "loss": 0.6579, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 1.4769393741851369, | |
| "grad_norm": 18.81052589416504, | |
| "learning_rate": 2.113836375488918e-05, | |
| "loss": 0.6699, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 1.4871251629726205, | |
| "grad_norm": 10.334957122802734, | |
| "learning_rate": 2.1077249022164277e-05, | |
| "loss": 0.655, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 1.4973109517601042, | |
| "grad_norm": 14.836852073669434, | |
| "learning_rate": 2.1016134289439378e-05, | |
| "loss": 0.66, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 1.5074967405475879, | |
| "grad_norm": 21.439233779907227, | |
| "learning_rate": 2.0955019556714472e-05, | |
| "loss": 0.6486, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 1.5176825293350718, | |
| "grad_norm": 21.09993553161621, | |
| "learning_rate": 2.0893904823989573e-05, | |
| "loss": 0.6432, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 1.5278683181225554, | |
| "grad_norm": 8.926955223083496, | |
| "learning_rate": 2.0832790091264667e-05, | |
| "loss": 0.6418, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 1.538054106910039, | |
| "grad_norm": 3.551163911819458, | |
| "learning_rate": 2.0771675358539768e-05, | |
| "loss": 0.6418, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 1.548239895697523, | |
| "grad_norm": 10.46932315826416, | |
| "learning_rate": 2.0710560625814862e-05, | |
| "loss": 0.657, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 1.5584256844850066, | |
| "grad_norm": 10.022995948791504, | |
| "learning_rate": 2.0649445893089963e-05, | |
| "loss": 0.6564, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 1.5686114732724903, | |
| "grad_norm": 15.299884796142578, | |
| "learning_rate": 2.0588331160365057e-05, | |
| "loss": 0.6535, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 1.578797262059974, | |
| "grad_norm": 11.149444580078125, | |
| "learning_rate": 2.0527216427640158e-05, | |
| "loss": 0.6409, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 1.5889830508474576, | |
| "grad_norm": 4.011179447174072, | |
| "learning_rate": 2.0466101694915252e-05, | |
| "loss": 0.6299, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 1.5991688396349413, | |
| "grad_norm": 13.744772911071777, | |
| "learning_rate": 2.0404986962190353e-05, | |
| "loss": 0.6511, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 1.609354628422425, | |
| "grad_norm": 11.129698753356934, | |
| "learning_rate": 2.0343872229465447e-05, | |
| "loss": 0.6349, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 1.6195404172099086, | |
| "grad_norm": 16.331953048706055, | |
| "learning_rate": 2.0282757496740548e-05, | |
| "loss": 0.6517, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 1.6297262059973925, | |
| "grad_norm": 12.816121101379395, | |
| "learning_rate": 2.0221642764015646e-05, | |
| "loss": 0.6409, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 1.6399119947848761, | |
| "grad_norm": 13.881726264953613, | |
| "learning_rate": 2.0160528031290743e-05, | |
| "loss": 0.6606, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 1.6500977835723598, | |
| "grad_norm": 14.215683937072754, | |
| "learning_rate": 2.0099413298565844e-05, | |
| "loss": 0.657, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 1.6602835723598437, | |
| "grad_norm": 19.557342529296875, | |
| "learning_rate": 2.003829856584094e-05, | |
| "loss": 0.6451, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 1.6704693611473274, | |
| "grad_norm": 17.83816909790039, | |
| "learning_rate": 1.997718383311604e-05, | |
| "loss": 0.6397, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 1.680655149934811, | |
| "grad_norm": 8.789198875427246, | |
| "learning_rate": 1.9916069100391133e-05, | |
| "loss": 0.6514, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 1.6908409387222947, | |
| "grad_norm": 11.361543655395508, | |
| "learning_rate": 1.9854954367666234e-05, | |
| "loss": 0.6466, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 1.7010267275097783, | |
| "grad_norm": 15.630517959594727, | |
| "learning_rate": 1.979383963494133e-05, | |
| "loss": 0.6316, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 1.711212516297262, | |
| "grad_norm": 10.55432415008545, | |
| "learning_rate": 1.973272490221643e-05, | |
| "loss": 0.6533, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 1.7213983050847457, | |
| "grad_norm": 8.428609848022461, | |
| "learning_rate": 1.9671610169491524e-05, | |
| "loss": 0.6478, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 1.7315840938722293, | |
| "grad_norm": 4.304576873779297, | |
| "learning_rate": 1.9610495436766625e-05, | |
| "loss": 0.6449, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 1.7417698826597132, | |
| "grad_norm": 21.92937469482422, | |
| "learning_rate": 1.9549380704041722e-05, | |
| "loss": 0.6425, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 1.7519556714471969, | |
| "grad_norm": 19.813392639160156, | |
| "learning_rate": 1.948826597131682e-05, | |
| "loss": 0.6238, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 1.7621414602346805, | |
| "grad_norm": 6.831646919250488, | |
| "learning_rate": 1.9427151238591917e-05, | |
| "loss": 0.6553, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 1.7723272490221644, | |
| "grad_norm": 11.058158874511719, | |
| "learning_rate": 1.9366036505867015e-05, | |
| "loss": 0.6528, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 1.782513037809648, | |
| "grad_norm": 7.01440954208374, | |
| "learning_rate": 1.9304921773142112e-05, | |
| "loss": 0.6506, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 1.7926988265971318, | |
| "grad_norm": 4.963765621185303, | |
| "learning_rate": 1.924380704041721e-05, | |
| "loss": 0.6412, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 1.8028846153846154, | |
| "grad_norm": 38.63767623901367, | |
| "learning_rate": 1.9182692307692307e-05, | |
| "loss": 0.6629, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 1.813070404172099, | |
| "grad_norm": 11.423843383789062, | |
| "learning_rate": 1.9121577574967405e-05, | |
| "loss": 0.673, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 1.8232561929595827, | |
| "grad_norm": 11.73025894165039, | |
| "learning_rate": 1.9060462842242506e-05, | |
| "loss": 0.625, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 1.8334419817470664, | |
| "grad_norm": 11.493837356567383, | |
| "learning_rate": 1.89993481095176e-05, | |
| "loss": 0.6658, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 1.84362777053455, | |
| "grad_norm": 12.196702003479004, | |
| "learning_rate": 1.89382333767927e-05, | |
| "loss": 0.6451, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 1.8538135593220337, | |
| "grad_norm": 9.690689086914062, | |
| "learning_rate": 1.8877118644067795e-05, | |
| "loss": 0.6282, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 1.8639993481095176, | |
| "grad_norm": 13.679101943969727, | |
| "learning_rate": 1.8816003911342896e-05, | |
| "loss": 0.6328, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 1.8741851368970013, | |
| "grad_norm": 14.264899253845215, | |
| "learning_rate": 1.8754889178617994e-05, | |
| "loss": 0.6355, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 1.8843709256844852, | |
| "grad_norm": 9.039682388305664, | |
| "learning_rate": 1.869377444589309e-05, | |
| "loss": 0.6471, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 1.8945567144719688, | |
| "grad_norm": 9.861899375915527, | |
| "learning_rate": 1.863265971316819e-05, | |
| "loss": 0.6549, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 1.9047425032594525, | |
| "grad_norm": 14.948457717895508, | |
| "learning_rate": 1.8571544980443286e-05, | |
| "loss": 0.6436, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 1.9149282920469362, | |
| "grad_norm": 15.613487243652344, | |
| "learning_rate": 1.8510430247718384e-05, | |
| "loss": 0.6355, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 1.9251140808344198, | |
| "grad_norm": 14.105792045593262, | |
| "learning_rate": 1.844931551499348e-05, | |
| "loss": 0.6345, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 1.9352998696219035, | |
| "grad_norm": 7.004807472229004, | |
| "learning_rate": 1.838820078226858e-05, | |
| "loss": 0.6456, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 1.9454856584093871, | |
| "grad_norm": 10.506879806518555, | |
| "learning_rate": 1.8327086049543677e-05, | |
| "loss": 0.6303, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 1.9556714471968708, | |
| "grad_norm": 11.595491409301758, | |
| "learning_rate": 1.8265971316818774e-05, | |
| "loss": 0.6509, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 1.9658572359843545, | |
| "grad_norm": 11.29542064666748, | |
| "learning_rate": 1.820485658409387e-05, | |
| "loss": 0.6273, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 1.9760430247718384, | |
| "grad_norm": 4.132607460021973, | |
| "learning_rate": 1.814374185136897e-05, | |
| "loss": 0.643, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 1.986228813559322, | |
| "grad_norm": 9.745738983154297, | |
| "learning_rate": 1.8082627118644067e-05, | |
| "loss": 0.6656, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 1.996414602346806, | |
| "grad_norm": 4.289452075958252, | |
| "learning_rate": 1.8021512385919164e-05, | |
| "loss": 0.6367, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7218826413154602, | |
| "eval_loss": 0.6872764825820923, | |
| "eval_runtime": 7.6175, | |
| "eval_samples_per_second": 644.307, | |
| "eval_steps_per_second": 80.604, | |
| "step": 98176 | |
| }, | |
| { | |
| "epoch": 2.0066003911342896, | |
| "grad_norm": 19.260536193847656, | |
| "learning_rate": 1.7960397653194265e-05, | |
| "loss": 0.5568, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 2.0167861799217732, | |
| "grad_norm": 8.347018241882324, | |
| "learning_rate": 1.7899282920469363e-05, | |
| "loss": 0.5344, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 2.026971968709257, | |
| "grad_norm": 12.238682746887207, | |
| "learning_rate": 1.783816818774446e-05, | |
| "loss": 0.5685, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 2.0371577574967406, | |
| "grad_norm": 18.42786979675293, | |
| "learning_rate": 1.7777053455019558e-05, | |
| "loss": 0.5539, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 2.047343546284224, | |
| "grad_norm": 25.627717971801758, | |
| "learning_rate": 1.7715938722294655e-05, | |
| "loss": 0.5393, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 2.057529335071708, | |
| "grad_norm": 15.942806243896484, | |
| "learning_rate": 1.7654823989569753e-05, | |
| "loss": 0.5191, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 2.0677151238591915, | |
| "grad_norm": 7.591663360595703, | |
| "learning_rate": 1.759370925684485e-05, | |
| "loss": 0.5428, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 2.077900912646675, | |
| "grad_norm": 7.183356285095215, | |
| "learning_rate": 1.7532594524119948e-05, | |
| "loss": 0.5475, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 2.088086701434159, | |
| "grad_norm": 19.916288375854492, | |
| "learning_rate": 1.7471479791395046e-05, | |
| "loss": 0.5586, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 2.098272490221643, | |
| "grad_norm": 9.07480239868164, | |
| "learning_rate": 1.7410365058670143e-05, | |
| "loss": 0.5564, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 2.1084582790091266, | |
| "grad_norm": 7.647058010101318, | |
| "learning_rate": 1.734925032594524e-05, | |
| "loss": 0.543, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "grad_norm": 8.871644973754883, | |
| "learning_rate": 1.728813559322034e-05, | |
| "loss": 0.5623, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 2.128829856584094, | |
| "grad_norm": 9.516389846801758, | |
| "learning_rate": 1.7227020860495436e-05, | |
| "loss": 0.5464, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 2.1390156453715776, | |
| "grad_norm": 13.605746269226074, | |
| "learning_rate": 1.7165906127770537e-05, | |
| "loss": 0.5425, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 2.1492014341590613, | |
| "grad_norm": 14.016572952270508, | |
| "learning_rate": 1.710479139504563e-05, | |
| "loss": 0.5496, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 2.159387222946545, | |
| "grad_norm": 14.719120979309082, | |
| "learning_rate": 1.7043676662320732e-05, | |
| "loss": 0.5591, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 2.1695730117340286, | |
| "grad_norm": 14.922130584716797, | |
| "learning_rate": 1.6982561929595826e-05, | |
| "loss": 0.5462, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 2.1797588005215123, | |
| "grad_norm": 10.74787425994873, | |
| "learning_rate": 1.6921447196870927e-05, | |
| "loss": 0.5699, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 2.189944589308996, | |
| "grad_norm": 7.978420257568359, | |
| "learning_rate": 1.6860332464146024e-05, | |
| "loss": 0.5449, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 2.2001303780964796, | |
| "grad_norm": 15.361347198486328, | |
| "learning_rate": 1.6799217731421122e-05, | |
| "loss": 0.5443, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 2.2103161668839633, | |
| "grad_norm": 6.552661418914795, | |
| "learning_rate": 1.673810299869622e-05, | |
| "loss": 0.5663, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 2.2205019556714474, | |
| "grad_norm": 6.750521659851074, | |
| "learning_rate": 1.6676988265971317e-05, | |
| "loss": 0.5488, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 2.230687744458931, | |
| "grad_norm": 11.149799346923828, | |
| "learning_rate": 1.6615873533246415e-05, | |
| "loss": 0.5412, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 2.2408735332464147, | |
| "grad_norm": 19.362638473510742, | |
| "learning_rate": 1.6554758800521512e-05, | |
| "loss": 0.5505, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 2.2510593220338984, | |
| "grad_norm": 11.263615608215332, | |
| "learning_rate": 1.6493644067796613e-05, | |
| "loss": 0.5591, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 2.261245110821382, | |
| "grad_norm": 8.90456771850586, | |
| "learning_rate": 1.6432529335071707e-05, | |
| "loss": 0.5491, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 2.2714308996088657, | |
| "grad_norm": 11.962569236755371, | |
| "learning_rate": 1.6371414602346808e-05, | |
| "loss": 0.556, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 2.2816166883963493, | |
| "grad_norm": 8.397544860839844, | |
| "learning_rate": 1.6310299869621902e-05, | |
| "loss": 0.5561, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 2.291802477183833, | |
| "grad_norm": 16.376155853271484, | |
| "learning_rate": 1.6249185136897003e-05, | |
| "loss": 0.5435, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 2.3019882659713167, | |
| "grad_norm": 9.14609432220459, | |
| "learning_rate": 1.6188070404172097e-05, | |
| "loss": 0.5457, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 2.3121740547588003, | |
| "grad_norm": 26.53936767578125, | |
| "learning_rate": 1.61269556714472e-05, | |
| "loss": 0.5295, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 2.322359843546284, | |
| "grad_norm": 17.222721099853516, | |
| "learning_rate": 1.6065840938722293e-05, | |
| "loss": 0.5177, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 2.332545632333768, | |
| "grad_norm": 9.585100173950195, | |
| "learning_rate": 1.6004726205997394e-05, | |
| "loss": 0.5419, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 2.3427314211212518, | |
| "grad_norm": 9.392489433288574, | |
| "learning_rate": 1.5943611473272488e-05, | |
| "loss": 0.5296, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 2.3529172099087354, | |
| "grad_norm": 15.146162986755371, | |
| "learning_rate": 1.588249674054759e-05, | |
| "loss": 0.5518, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 2.363102998696219, | |
| "grad_norm": 15.631641387939453, | |
| "learning_rate": 1.582138200782269e-05, | |
| "loss": 0.5584, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 2.3732887874837028, | |
| "grad_norm": 17.875492095947266, | |
| "learning_rate": 1.5760267275097784e-05, | |
| "loss": 0.5601, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 2.3834745762711864, | |
| "grad_norm": 7.7471184730529785, | |
| "learning_rate": 1.5699152542372885e-05, | |
| "loss": 0.549, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 2.39366036505867, | |
| "grad_norm": 14.712841987609863, | |
| "learning_rate": 1.563803780964798e-05, | |
| "loss": 0.5293, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 2.4038461538461537, | |
| "grad_norm": 9.229011535644531, | |
| "learning_rate": 1.557692307692308e-05, | |
| "loss": 0.5483, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 2.4140319426336374, | |
| "grad_norm": 11.47548770904541, | |
| "learning_rate": 1.5515808344198174e-05, | |
| "loss": 0.5529, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 2.424217731421121, | |
| "grad_norm": 29.980873107910156, | |
| "learning_rate": 1.5454693611473275e-05, | |
| "loss": 0.5489, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 2.4344035202086047, | |
| "grad_norm": 10.478185653686523, | |
| "learning_rate": 1.539357887874837e-05, | |
| "loss": 0.5595, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 2.444589308996089, | |
| "grad_norm": 12.978096008300781, | |
| "learning_rate": 1.533246414602347e-05, | |
| "loss": 0.5612, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 2.4547750977835725, | |
| "grad_norm": 17.834806442260742, | |
| "learning_rate": 1.5271349413298564e-05, | |
| "loss": 0.5412, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 2.464960886571056, | |
| "grad_norm": 25.001754760742188, | |
| "learning_rate": 1.5210234680573665e-05, | |
| "loss": 0.5673, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 2.47514667535854, | |
| "grad_norm": 27.953767776489258, | |
| "learning_rate": 1.5149119947848761e-05, | |
| "loss": 0.5497, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 2.4853324641460235, | |
| "grad_norm": 9.370855331420898, | |
| "learning_rate": 1.508800521512386e-05, | |
| "loss": 0.5344, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 2.495518252933507, | |
| "grad_norm": 18.109821319580078, | |
| "learning_rate": 1.5026890482398956e-05, | |
| "loss": 0.5572, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 2.505704041720991, | |
| "grad_norm": 20.435270309448242, | |
| "learning_rate": 1.4965775749674055e-05, | |
| "loss": 0.5631, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 2.5158898305084745, | |
| "grad_norm": 10.656272888183594, | |
| "learning_rate": 1.4904661016949153e-05, | |
| "loss": 0.5344, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 2.526075619295958, | |
| "grad_norm": 18.796552658081055, | |
| "learning_rate": 1.484354628422425e-05, | |
| "loss": 0.5464, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 2.5362614080834422, | |
| "grad_norm": 9.630306243896484, | |
| "learning_rate": 1.4782431551499348e-05, | |
| "loss": 0.547, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 2.5464471968709255, | |
| "grad_norm": 6.3599467277526855, | |
| "learning_rate": 1.4721316818774445e-05, | |
| "loss": 0.5577, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 2.5566329856584096, | |
| "grad_norm": 8.967144966125488, | |
| "learning_rate": 1.4660202086049545e-05, | |
| "loss": 0.5524, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 2.5668187744458932, | |
| "grad_norm": 16.690263748168945, | |
| "learning_rate": 1.4599087353324642e-05, | |
| "loss": 0.5546, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 2.577004563233377, | |
| "grad_norm": 17.713790893554688, | |
| "learning_rate": 1.453797262059974e-05, | |
| "loss": 0.5686, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 2.5871903520208606, | |
| "grad_norm": 7.970760822296143, | |
| "learning_rate": 1.4476857887874837e-05, | |
| "loss": 0.5273, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 2.5973761408083442, | |
| "grad_norm": 8.869507789611816, | |
| "learning_rate": 1.4415743155149935e-05, | |
| "loss": 0.5472, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 2.607561929595828, | |
| "grad_norm": 17.247589111328125, | |
| "learning_rate": 1.4354628422425032e-05, | |
| "loss": 0.5444, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 2.6177477183833116, | |
| "grad_norm": 16.415138244628906, | |
| "learning_rate": 1.429351368970013e-05, | |
| "loss": 0.5445, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 2.627933507170795, | |
| "grad_norm": 10.408729553222656, | |
| "learning_rate": 1.4232398956975227e-05, | |
| "loss": 0.5667, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 2.638119295958279, | |
| "grad_norm": 53.77888488769531, | |
| "learning_rate": 1.4171284224250327e-05, | |
| "loss": 0.547, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 2.648305084745763, | |
| "grad_norm": 12.244620323181152, | |
| "learning_rate": 1.4110169491525424e-05, | |
| "loss": 0.5428, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 2.658490873533246, | |
| "grad_norm": 11.80124568939209, | |
| "learning_rate": 1.4049054758800522e-05, | |
| "loss": 0.5565, | |
| "step": 130500 | |
| }, | |
| { | |
| "epoch": 2.6686766623207303, | |
| "grad_norm": 13.250449180603027, | |
| "learning_rate": 1.398794002607562e-05, | |
| "loss": 0.5423, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 2.678862451108214, | |
| "grad_norm": 15.81285572052002, | |
| "learning_rate": 1.3926825293350719e-05, | |
| "loss": 0.5621, | |
| "step": 131500 | |
| }, | |
| { | |
| "epoch": 2.6890482398956976, | |
| "grad_norm": 22.584035873413086, | |
| "learning_rate": 1.3865710560625816e-05, | |
| "loss": 0.5521, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 2.6992340286831813, | |
| "grad_norm": 23.865680694580078, | |
| "learning_rate": 1.3804595827900914e-05, | |
| "loss": 0.551, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 2.709419817470665, | |
| "grad_norm": 18.28876304626465, | |
| "learning_rate": 1.3743481095176011e-05, | |
| "loss": 0.5451, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 2.7196056062581486, | |
| "grad_norm": 16.23969078063965, | |
| "learning_rate": 1.3682366362451109e-05, | |
| "loss": 0.5392, | |
| "step": 133500 | |
| }, | |
| { | |
| "epoch": 2.7297913950456323, | |
| "grad_norm": 14.673959732055664, | |
| "learning_rate": 1.3621251629726206e-05, | |
| "loss": 0.5451, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 2.739977183833116, | |
| "grad_norm": 8.019514083862305, | |
| "learning_rate": 1.3560136897001304e-05, | |
| "loss": 0.5568, | |
| "step": 134500 | |
| }, | |
| { | |
| "epoch": 2.7501629726205996, | |
| "grad_norm": 13.495898246765137, | |
| "learning_rate": 1.3499022164276401e-05, | |
| "loss": 0.5734, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 2.7603487614080837, | |
| "grad_norm": 7.548976421356201, | |
| "learning_rate": 1.3437907431551499e-05, | |
| "loss": 0.557, | |
| "step": 135500 | |
| }, | |
| { | |
| "epoch": 2.770534550195567, | |
| "grad_norm": 20.049760818481445, | |
| "learning_rate": 1.3376792698826597e-05, | |
| "loss": 0.5658, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 2.780720338983051, | |
| "grad_norm": 9.346122741699219, | |
| "learning_rate": 1.3315677966101694e-05, | |
| "loss": 0.535, | |
| "step": 136500 | |
| }, | |
| { | |
| "epoch": 2.7909061277705347, | |
| "grad_norm": 15.080660820007324, | |
| "learning_rate": 1.3254563233376792e-05, | |
| "loss": 0.5538, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 2.8010919165580184, | |
| "grad_norm": 11.485374450683594, | |
| "learning_rate": 1.3193448500651891e-05, | |
| "loss": 0.5419, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 2.811277705345502, | |
| "grad_norm": 12.089446067810059, | |
| "learning_rate": 1.3132333767926988e-05, | |
| "loss": 0.5651, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 2.8214634941329857, | |
| "grad_norm": 5.466490268707275, | |
| "learning_rate": 1.3071219035202088e-05, | |
| "loss": 0.5217, | |
| "step": 138500 | |
| }, | |
| { | |
| "epoch": 2.8316492829204694, | |
| "grad_norm": 12.89148235321045, | |
| "learning_rate": 1.3010104302477185e-05, | |
| "loss": 0.5477, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 2.841835071707953, | |
| "grad_norm": 5.610709190368652, | |
| "learning_rate": 1.2948989569752283e-05, | |
| "loss": 0.5377, | |
| "step": 139500 | |
| }, | |
| { | |
| "epoch": 2.8520208604954367, | |
| "grad_norm": 26.2186222076416, | |
| "learning_rate": 1.288787483702738e-05, | |
| "loss": 0.5281, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 2.8622066492829203, | |
| "grad_norm": 2.1066462993621826, | |
| "learning_rate": 1.2826760104302478e-05, | |
| "loss": 0.5387, | |
| "step": 140500 | |
| }, | |
| { | |
| "epoch": 2.872392438070404, | |
| "grad_norm": 13.6646728515625, | |
| "learning_rate": 1.2765645371577575e-05, | |
| "loss": 0.5416, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 2.8825782268578877, | |
| "grad_norm": 14.357284545898438, | |
| "learning_rate": 1.2704530638852673e-05, | |
| "loss": 0.5504, | |
| "step": 141500 | |
| }, | |
| { | |
| "epoch": 2.8927640156453718, | |
| "grad_norm": 14.08674144744873, | |
| "learning_rate": 1.264341590612777e-05, | |
| "loss": 0.5691, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 2.9029498044328554, | |
| "grad_norm": 20.460561752319336, | |
| "learning_rate": 1.2582301173402868e-05, | |
| "loss": 0.5624, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 2.913135593220339, | |
| "grad_norm": 12.731225967407227, | |
| "learning_rate": 1.2521186440677966e-05, | |
| "loss": 0.5439, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 2.9233213820078228, | |
| "grad_norm": 3.781522274017334, | |
| "learning_rate": 1.2460071707953065e-05, | |
| "loss": 0.5348, | |
| "step": 143500 | |
| }, | |
| { | |
| "epoch": 2.9335071707953064, | |
| "grad_norm": 14.624204635620117, | |
| "learning_rate": 1.2398956975228162e-05, | |
| "loss": 0.5589, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 2.94369295958279, | |
| "grad_norm": 18.551095962524414, | |
| "learning_rate": 1.233784224250326e-05, | |
| "loss": 0.532, | |
| "step": 144500 | |
| }, | |
| { | |
| "epoch": 2.9538787483702738, | |
| "grad_norm": 17.094831466674805, | |
| "learning_rate": 1.2276727509778357e-05, | |
| "loss": 0.558, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 2.9640645371577574, | |
| "grad_norm": 12.306907653808594, | |
| "learning_rate": 1.2215612777053455e-05, | |
| "loss": 0.542, | |
| "step": 145500 | |
| }, | |
| { | |
| "epoch": 2.974250325945241, | |
| "grad_norm": 12.134025573730469, | |
| "learning_rate": 1.2154498044328553e-05, | |
| "loss": 0.5429, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 2.9844361147327247, | |
| "grad_norm": 22.295795440673828, | |
| "learning_rate": 1.209338331160365e-05, | |
| "loss": 0.5632, | |
| "step": 146500 | |
| }, | |
| { | |
| "epoch": 2.9946219035202084, | |
| "grad_norm": 17.749858856201172, | |
| "learning_rate": 1.2032268578878748e-05, | |
| "loss": 0.5402, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7337000966072083, | |
| "eval_loss": 0.7322831153869629, | |
| "eval_runtime": 8.3496, | |
| "eval_samples_per_second": 587.813, | |
| "eval_steps_per_second": 73.536, | |
| "step": 147264 | |
| }, | |
| { | |
| "epoch": 3.0048076923076925, | |
| "grad_norm": 22.746931076049805, | |
| "learning_rate": 1.1971153846153847e-05, | |
| "loss": 0.5024, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 3.014993481095176, | |
| "grad_norm": 19.812463760375977, | |
| "learning_rate": 1.1910039113428944e-05, | |
| "loss": 0.455, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 3.02517926988266, | |
| "grad_norm": 8.608039855957031, | |
| "learning_rate": 1.1848924380704042e-05, | |
| "loss": 0.4529, | |
| "step": 148500 | |
| }, | |
| { | |
| "epoch": 3.0353650586701435, | |
| "grad_norm": 21.67936897277832, | |
| "learning_rate": 1.178780964797914e-05, | |
| "loss": 0.467, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 3.045550847457627, | |
| "grad_norm": 23.777921676635742, | |
| "learning_rate": 1.1726694915254239e-05, | |
| "loss": 0.4399, | |
| "step": 149500 | |
| }, | |
| { | |
| "epoch": 3.055736636245111, | |
| "grad_norm": 5.8297600746154785, | |
| "learning_rate": 1.1665580182529336e-05, | |
| "loss": 0.4605, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 3.0659224250325945, | |
| "grad_norm": 1.6593589782714844, | |
| "learning_rate": 1.1604465449804434e-05, | |
| "loss": 0.4668, | |
| "step": 150500 | |
| }, | |
| { | |
| "epoch": 3.076108213820078, | |
| "grad_norm": 41.140750885009766, | |
| "learning_rate": 1.1543350717079531e-05, | |
| "loss": 0.4558, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 3.086294002607562, | |
| "grad_norm": 14.093750953674316, | |
| "learning_rate": 1.1482235984354629e-05, | |
| "loss": 0.4645, | |
| "step": 151500 | |
| }, | |
| { | |
| "epoch": 3.0964797913950455, | |
| "grad_norm": 25.050161361694336, | |
| "learning_rate": 1.1421121251629727e-05, | |
| "loss": 0.4549, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 3.106665580182529, | |
| "grad_norm": 12.942317962646484, | |
| "learning_rate": 1.1360006518904824e-05, | |
| "loss": 0.4675, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 3.1168513689700132, | |
| "grad_norm": 26.615070343017578, | |
| "learning_rate": 1.1298891786179922e-05, | |
| "loss": 0.4606, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 3.127037157757497, | |
| "grad_norm": 35.78953552246094, | |
| "learning_rate": 1.123777705345502e-05, | |
| "loss": 0.4754, | |
| "step": 153500 | |
| }, | |
| { | |
| "epoch": 3.1372229465449806, | |
| "grad_norm": 17.106658935546875, | |
| "learning_rate": 1.1176662320730117e-05, | |
| "loss": 0.46, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 3.1474087353324642, | |
| "grad_norm": 5.81476354598999, | |
| "learning_rate": 1.1115547588005214e-05, | |
| "loss": 0.455, | |
| "step": 154500 | |
| }, | |
| { | |
| "epoch": 3.157594524119948, | |
| "grad_norm": 13.205123901367188, | |
| "learning_rate": 1.1054432855280312e-05, | |
| "loss": 0.4443, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 3.1677803129074316, | |
| "grad_norm": 44.934471130371094, | |
| "learning_rate": 1.0993318122555411e-05, | |
| "loss": 0.447, | |
| "step": 155500 | |
| }, | |
| { | |
| "epoch": 3.1779661016949152, | |
| "grad_norm": 8.879769325256348, | |
| "learning_rate": 1.0932203389830509e-05, | |
| "loss": 0.4536, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 3.188151890482399, | |
| "grad_norm": 22.281354904174805, | |
| "learning_rate": 1.0871088657105608e-05, | |
| "loss": 0.4541, | |
| "step": 156500 | |
| }, | |
| { | |
| "epoch": 3.1983376792698825, | |
| "grad_norm": 46.504112243652344, | |
| "learning_rate": 1.0809973924380705e-05, | |
| "loss": 0.4467, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 3.208523468057366, | |
| "grad_norm": 1.4721815586090088, | |
| "learning_rate": 1.0748859191655803e-05, | |
| "loss": 0.4474, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 3.21870925684485, | |
| "grad_norm": 18.72897720336914, | |
| "learning_rate": 1.06877444589309e-05, | |
| "loss": 0.4465, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 3.228895045632334, | |
| "grad_norm": 13.687529563903809, | |
| "learning_rate": 1.0626629726205998e-05, | |
| "loss": 0.4674, | |
| "step": 158500 | |
| }, | |
| { | |
| "epoch": 3.2390808344198176, | |
| "grad_norm": 2.0822532176971436, | |
| "learning_rate": 1.0565514993481096e-05, | |
| "loss": 0.4516, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 3.2492666232073013, | |
| "grad_norm": 15.965363502502441, | |
| "learning_rate": 1.0504400260756193e-05, | |
| "loss": 0.4582, | |
| "step": 159500 | |
| }, | |
| { | |
| "epoch": 3.259452411994785, | |
| "grad_norm": 19.683805465698242, | |
| "learning_rate": 1.044328552803129e-05, | |
| "loss": 0.452, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 3.2696382007822686, | |
| "grad_norm": 30.53873062133789, | |
| "learning_rate": 1.0382170795306388e-05, | |
| "loss": 0.4532, | |
| "step": 160500 | |
| }, | |
| { | |
| "epoch": 3.2798239895697523, | |
| "grad_norm": 14.192473411560059, | |
| "learning_rate": 1.0321056062581486e-05, | |
| "loss": 0.4576, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 3.290009778357236, | |
| "grad_norm": 19.519214630126953, | |
| "learning_rate": 1.0259941329856583e-05, | |
| "loss": 0.4531, | |
| "step": 161500 | |
| }, | |
| { | |
| "epoch": 3.3001955671447196, | |
| "grad_norm": 11.308916091918945, | |
| "learning_rate": 1.0198826597131683e-05, | |
| "loss": 0.4374, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 3.3103813559322033, | |
| "grad_norm": 22.54949378967285, | |
| "learning_rate": 1.013771186440678e-05, | |
| "loss": 0.4641, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 3.320567144719687, | |
| "grad_norm": 27.339025497436523, | |
| "learning_rate": 1.0076597131681878e-05, | |
| "loss": 0.4603, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 3.3307529335071706, | |
| "grad_norm": 10.614120483398438, | |
| "learning_rate": 1.0015482398956975e-05, | |
| "loss": 0.4482, | |
| "step": 163500 | |
| }, | |
| { | |
| "epoch": 3.3409387222946547, | |
| "grad_norm": 19.92940902709961, | |
| "learning_rate": 9.954367666232073e-06, | |
| "loss": 0.4503, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 3.3511245110821384, | |
| "grad_norm": 21.356348037719727, | |
| "learning_rate": 9.89325293350717e-06, | |
| "loss": 0.4542, | |
| "step": 164500 | |
| }, | |
| { | |
| "epoch": 3.361310299869622, | |
| "grad_norm": 19.11351203918457, | |
| "learning_rate": 9.83213820078227e-06, | |
| "loss": 0.4653, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 3.3714960886571057, | |
| "grad_norm": 15.949396133422852, | |
| "learning_rate": 9.771023468057367e-06, | |
| "loss": 0.4583, | |
| "step": 165500 | |
| }, | |
| { | |
| "epoch": 3.3816818774445894, | |
| "grad_norm": 10.171916961669922, | |
| "learning_rate": 9.709908735332465e-06, | |
| "loss": 0.4747, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 3.391867666232073, | |
| "grad_norm": 17.548404693603516, | |
| "learning_rate": 9.648794002607562e-06, | |
| "loss": 0.4252, | |
| "step": 166500 | |
| }, | |
| { | |
| "epoch": 3.4020534550195567, | |
| "grad_norm": 26.116491317749023, | |
| "learning_rate": 9.58767926988266e-06, | |
| "loss": 0.4429, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 3.4122392438070404, | |
| "grad_norm": 8.717327117919922, | |
| "learning_rate": 9.526564537157757e-06, | |
| "loss": 0.4752, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 3.422425032594524, | |
| "grad_norm": 5.2715535163879395, | |
| "learning_rate": 9.465449804432857e-06, | |
| "loss": 0.4689, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 3.4326108213820077, | |
| "grad_norm": 15.369921684265137, | |
| "learning_rate": 9.404335071707954e-06, | |
| "loss": 0.474, | |
| "step": 168500 | |
| }, | |
| { | |
| "epoch": 3.4427966101694913, | |
| "grad_norm": 23.541234970092773, | |
| "learning_rate": 9.343220338983052e-06, | |
| "loss": 0.4717, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 3.4529823989569755, | |
| "grad_norm": 25.710681915283203, | |
| "learning_rate": 9.28210560625815e-06, | |
| "loss": 0.459, | |
| "step": 169500 | |
| }, | |
| { | |
| "epoch": 3.463168187744459, | |
| "grad_norm": 36.22392272949219, | |
| "learning_rate": 9.220990873533247e-06, | |
| "loss": 0.4374, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 3.4733539765319428, | |
| "grad_norm": 17.486045837402344, | |
| "learning_rate": 9.159876140808344e-06, | |
| "loss": 0.4682, | |
| "step": 170500 | |
| }, | |
| { | |
| "epoch": 3.4835397653194264, | |
| "grad_norm": 21.079090118408203, | |
| "learning_rate": 9.098761408083442e-06, | |
| "loss": 0.461, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 3.49372555410691, | |
| "grad_norm": 47.44207763671875, | |
| "learning_rate": 9.03764667535854e-06, | |
| "loss": 0.428, | |
| "step": 171500 | |
| }, | |
| { | |
| "epoch": 3.5039113428943938, | |
| "grad_norm": 10.158509254455566, | |
| "learning_rate": 8.976531942633637e-06, | |
| "loss": 0.4705, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 3.5140971316818774, | |
| "grad_norm": 2.3025388717651367, | |
| "learning_rate": 8.915417209908735e-06, | |
| "loss": 0.4794, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 3.524282920469361, | |
| "grad_norm": 7.106923580169678, | |
| "learning_rate": 8.854302477183832e-06, | |
| "loss": 0.4596, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 3.5344687092568448, | |
| "grad_norm": 11.846363067626953, | |
| "learning_rate": 8.79318774445893e-06, | |
| "loss": 0.4548, | |
| "step": 173500 | |
| }, | |
| { | |
| "epoch": 3.5446544980443284, | |
| "grad_norm": 23.35795783996582, | |
| "learning_rate": 8.73207301173403e-06, | |
| "loss": 0.4627, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 3.554840286831812, | |
| "grad_norm": 51.494873046875, | |
| "learning_rate": 8.670958279009128e-06, | |
| "loss": 0.443, | |
| "step": 174500 | |
| }, | |
| { | |
| "epoch": 3.565026075619296, | |
| "grad_norm": 17.30136489868164, | |
| "learning_rate": 8.609843546284226e-06, | |
| "loss": 0.4431, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 3.5752118644067794, | |
| "grad_norm": 6.228327751159668, | |
| "learning_rate": 8.548728813559323e-06, | |
| "loss": 0.4494, | |
| "step": 175500 | |
| }, | |
| { | |
| "epoch": 3.5853976531942635, | |
| "grad_norm": 7.742433071136475, | |
| "learning_rate": 8.48761408083442e-06, | |
| "loss": 0.4591, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 3.595583441981747, | |
| "grad_norm": 17.392818450927734, | |
| "learning_rate": 8.426499348109518e-06, | |
| "loss": 0.4572, | |
| "step": 176500 | |
| }, | |
| { | |
| "epoch": 3.605769230769231, | |
| "grad_norm": 9.273465156555176, | |
| "learning_rate": 8.365384615384616e-06, | |
| "loss": 0.4592, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 3.6159550195567145, | |
| "grad_norm": 10.859349250793457, | |
| "learning_rate": 8.304269882659713e-06, | |
| "loss": 0.4507, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 3.626140808344198, | |
| "grad_norm": 16.180192947387695, | |
| "learning_rate": 8.243155149934811e-06, | |
| "loss": 0.4615, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 3.636326597131682, | |
| "grad_norm": 7.720341682434082, | |
| "learning_rate": 8.182040417209908e-06, | |
| "loss": 0.4614, | |
| "step": 178500 | |
| }, | |
| { | |
| "epoch": 3.6465123859191655, | |
| "grad_norm": 10.102255821228027, | |
| "learning_rate": 8.120925684485006e-06, | |
| "loss": 0.4504, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 3.656698174706649, | |
| "grad_norm": 9.698884010314941, | |
| "learning_rate": 8.059810951760104e-06, | |
| "loss": 0.4791, | |
| "step": 179500 | |
| }, | |
| { | |
| "epoch": 3.666883963494133, | |
| "grad_norm": 13.645587921142578, | |
| "learning_rate": 7.998696219035203e-06, | |
| "loss": 0.4655, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 3.677069752281617, | |
| "grad_norm": 6.0963358879089355, | |
| "learning_rate": 7.9375814863103e-06, | |
| "loss": 0.4522, | |
| "step": 180500 | |
| }, | |
| { | |
| "epoch": 3.6872555410691, | |
| "grad_norm": 19.082395553588867, | |
| "learning_rate": 7.876466753585398e-06, | |
| "loss": 0.4606, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 3.6974413298565842, | |
| "grad_norm": 21.68135643005371, | |
| "learning_rate": 7.815352020860495e-06, | |
| "loss": 0.475, | |
| "step": 181500 | |
| }, | |
| { | |
| "epoch": 3.707627118644068, | |
| "grad_norm": 22.7216796875, | |
| "learning_rate": 7.754237288135593e-06, | |
| "loss": 0.4561, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 3.7178129074315516, | |
| "grad_norm": 5.14304780960083, | |
| "learning_rate": 7.69312255541069e-06, | |
| "loss": 0.4561, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 3.7279986962190352, | |
| "grad_norm": 12.879047393798828, | |
| "learning_rate": 7.63200782268579e-06, | |
| "loss": 0.4619, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 3.738184485006519, | |
| "grad_norm": 20.963592529296875, | |
| "learning_rate": 7.570893089960887e-06, | |
| "loss": 0.4488, | |
| "step": 183500 | |
| }, | |
| { | |
| "epoch": 3.7483702737940026, | |
| "grad_norm": 15.685653686523438, | |
| "learning_rate": 7.509778357235985e-06, | |
| "loss": 0.4515, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 3.7585560625814862, | |
| "grad_norm": 14.296714782714844, | |
| "learning_rate": 7.4486636245110824e-06, | |
| "loss": 0.4461, | |
| "step": 184500 | |
| }, | |
| { | |
| "epoch": 3.76874185136897, | |
| "grad_norm": 8.592365264892578, | |
| "learning_rate": 7.38754889178618e-06, | |
| "loss": 0.4571, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 3.7789276401564535, | |
| "grad_norm": 32.55515670776367, | |
| "learning_rate": 7.326434159061278e-06, | |
| "loss": 0.4601, | |
| "step": 185500 | |
| }, | |
| { | |
| "epoch": 3.7891134289439377, | |
| "grad_norm": 19.179519653320312, | |
| "learning_rate": 7.265319426336376e-06, | |
| "loss": 0.4591, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 3.799299217731421, | |
| "grad_norm": 11.681869506835938, | |
| "learning_rate": 7.2042046936114735e-06, | |
| "loss": 0.4638, | |
| "step": 186500 | |
| }, | |
| { | |
| "epoch": 3.809485006518905, | |
| "grad_norm": 17.548564910888672, | |
| "learning_rate": 7.143089960886571e-06, | |
| "loss": 0.4595, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 3.8196707953063886, | |
| "grad_norm": 34.2673225402832, | |
| "learning_rate": 7.0819752281616686e-06, | |
| "loss": 0.4622, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 3.8298565840938723, | |
| "grad_norm": 25.60137939453125, | |
| "learning_rate": 7.020860495436767e-06, | |
| "loss": 0.4653, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 3.840042372881356, | |
| "grad_norm": 14.652885437011719, | |
| "learning_rate": 6.9597457627118645e-06, | |
| "loss": 0.4484, | |
| "step": 188500 | |
| }, | |
| { | |
| "epoch": 3.8502281616688396, | |
| "grad_norm": 20.71872901916504, | |
| "learning_rate": 6.898631029986962e-06, | |
| "loss": 0.4717, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 3.8604139504563233, | |
| "grad_norm": 6.4770989418029785, | |
| "learning_rate": 6.83751629726206e-06, | |
| "loss": 0.462, | |
| "step": 189500 | |
| }, | |
| { | |
| "epoch": 3.870599739243807, | |
| "grad_norm": 21.029338836669922, | |
| "learning_rate": 6.776401564537158e-06, | |
| "loss": 0.4312, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 3.8807855280312906, | |
| "grad_norm": 27.267345428466797, | |
| "learning_rate": 6.7152868318122556e-06, | |
| "loss": 0.4623, | |
| "step": 190500 | |
| }, | |
| { | |
| "epoch": 3.8909713168187743, | |
| "grad_norm": 16.270328521728516, | |
| "learning_rate": 6.654172099087354e-06, | |
| "loss": 0.4594, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 3.9011571056062584, | |
| "grad_norm": 15.952796936035156, | |
| "learning_rate": 6.5930573663624515e-06, | |
| "loss": 0.4478, | |
| "step": 191500 | |
| }, | |
| { | |
| "epoch": 3.9113428943937416, | |
| "grad_norm": 24.082805633544922, | |
| "learning_rate": 6.531942633637549e-06, | |
| "loss": 0.4572, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 3.9215286831812257, | |
| "grad_norm": 17.784955978393555, | |
| "learning_rate": 6.470827900912647e-06, | |
| "loss": 0.453, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 3.9317144719687094, | |
| "grad_norm": 15.306330680847168, | |
| "learning_rate": 6.409713168187744e-06, | |
| "loss": 0.4516, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 3.941900260756193, | |
| "grad_norm": 17.07221031188965, | |
| "learning_rate": 6.348598435462842e-06, | |
| "loss": 0.4513, | |
| "step": 193500 | |
| }, | |
| { | |
| "epoch": 3.9520860495436767, | |
| "grad_norm": 1.9630801677703857, | |
| "learning_rate": 6.287483702737941e-06, | |
| "loss": 0.4339, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 3.9622718383311604, | |
| "grad_norm": 23.103532791137695, | |
| "learning_rate": 6.2263689700130385e-06, | |
| "loss": 0.475, | |
| "step": 194500 | |
| }, | |
| { | |
| "epoch": 3.972457627118644, | |
| "grad_norm": 9.09752082824707, | |
| "learning_rate": 6.165254237288136e-06, | |
| "loss": 0.4422, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 3.9826434159061277, | |
| "grad_norm": 23.36123275756836, | |
| "learning_rate": 6.104139504563234e-06, | |
| "loss": 0.4578, | |
| "step": 195500 | |
| }, | |
| { | |
| "epoch": 3.9928292046936114, | |
| "grad_norm": 15.44927978515625, | |
| "learning_rate": 6.043024771838331e-06, | |
| "loss": 0.4815, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7300326228141785, | |
| "eval_loss": 0.8482908010482788, | |
| "eval_runtime": 7.2526, | |
| "eval_samples_per_second": 676.722, | |
| "eval_steps_per_second": 84.659, | |
| "step": 196352 | |
| }, | |
| { | |
| "epoch": 4.003014993481095, | |
| "grad_norm": 18.638896942138672, | |
| "learning_rate": 5.981910039113429e-06, | |
| "loss": 0.4162, | |
| "step": 196500 | |
| }, | |
| { | |
| "epoch": 4.013200782268579, | |
| "grad_norm": 23.482778549194336, | |
| "learning_rate": 5.920795306388527e-06, | |
| "loss": 0.3857, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 4.023386571056062, | |
| "grad_norm": 19.941246032714844, | |
| "learning_rate": 5.859680573663625e-06, | |
| "loss": 0.4055, | |
| "step": 197500 | |
| }, | |
| { | |
| "epoch": 4.0335723598435465, | |
| "grad_norm": 55.201602935791016, | |
| "learning_rate": 5.798565840938722e-06, | |
| "loss": 0.4054, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 4.04375814863103, | |
| "grad_norm": 9.683618545532227, | |
| "learning_rate": 5.7374511082138206e-06, | |
| "loss": 0.3772, | |
| "step": 198500 | |
| }, | |
| { | |
| "epoch": 4.053943937418514, | |
| "grad_norm": 14.284856796264648, | |
| "learning_rate": 5.676336375488918e-06, | |
| "loss": 0.3761, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 4.064129726205997, | |
| "grad_norm": 33.600894927978516, | |
| "learning_rate": 5.615221642764016e-06, | |
| "loss": 0.394, | |
| "step": 199500 | |
| }, | |
| { | |
| "epoch": 4.074315514993481, | |
| "grad_norm": 4.879152774810791, | |
| "learning_rate": 5.554106910039114e-06, | |
| "loss": 0.3754, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 4.084501303780965, | |
| "grad_norm": 4.289336681365967, | |
| "learning_rate": 5.492992177314212e-06, | |
| "loss": 0.3822, | |
| "step": 200500 | |
| }, | |
| { | |
| "epoch": 4.094687092568448, | |
| "grad_norm": 13.225927352905273, | |
| "learning_rate": 5.431877444589309e-06, | |
| "loss": 0.385, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 4.1048728813559325, | |
| "grad_norm": 30.11479949951172, | |
| "learning_rate": 5.370762711864407e-06, | |
| "loss": 0.4043, | |
| "step": 201500 | |
| }, | |
| { | |
| "epoch": 4.115058670143416, | |
| "grad_norm": 30.226152420043945, | |
| "learning_rate": 5.309647979139504e-06, | |
| "loss": 0.3785, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 4.1252444589309, | |
| "grad_norm": 57.69206619262695, | |
| "learning_rate": 5.248533246414602e-06, | |
| "loss": 0.3783, | |
| "step": 202500 | |
| }, | |
| { | |
| "epoch": 4.135430247718383, | |
| "grad_norm": 0.5498570203781128, | |
| "learning_rate": 5.187418513689701e-06, | |
| "loss": 0.387, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 4.145616036505867, | |
| "grad_norm": 1.2860363721847534, | |
| "learning_rate": 5.126303780964799e-06, | |
| "loss": 0.3788, | |
| "step": 203500 | |
| }, | |
| { | |
| "epoch": 4.15580182529335, | |
| "grad_norm": 24.312036514282227, | |
| "learning_rate": 5.065189048239896e-06, | |
| "loss": 0.3958, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 4.1659876140808345, | |
| "grad_norm": 31.0595645904541, | |
| "learning_rate": 5.004074315514994e-06, | |
| "loss": 0.368, | |
| "step": 204500 | |
| }, | |
| { | |
| "epoch": 4.176173402868318, | |
| "grad_norm": 30.00829315185547, | |
| "learning_rate": 4.942959582790091e-06, | |
| "loss": 0.387, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 4.186359191655802, | |
| "grad_norm": 5.656859874725342, | |
| "learning_rate": 4.881844850065189e-06, | |
| "loss": 0.3876, | |
| "step": 205500 | |
| }, | |
| { | |
| "epoch": 4.196544980443286, | |
| "grad_norm": 28.364885330200195, | |
| "learning_rate": 4.820730117340287e-06, | |
| "loss": 0.3974, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 4.206730769230769, | |
| "grad_norm": 21.450817108154297, | |
| "learning_rate": 4.759615384615385e-06, | |
| "loss": 0.4304, | |
| "step": 206500 | |
| }, | |
| { | |
| "epoch": 4.216916558018253, | |
| "grad_norm": 26.55447769165039, | |
| "learning_rate": 4.698500651890482e-06, | |
| "loss": 0.3598, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 4.2271023468057365, | |
| "grad_norm": 63.503299713134766, | |
| "learning_rate": 4.637385919165581e-06, | |
| "loss": 0.3972, | |
| "step": 207500 | |
| }, | |
| { | |
| "epoch": 4.237288135593221, | |
| "grad_norm": 6.910649299621582, | |
| "learning_rate": 4.576271186440678e-06, | |
| "loss": 0.3788, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 4.247473924380704, | |
| "grad_norm": 6.5545172691345215, | |
| "learning_rate": 4.515156453715776e-06, | |
| "loss": 0.3978, | |
| "step": 208500 | |
| }, | |
| { | |
| "epoch": 4.257659713168188, | |
| "grad_norm": 17.815195083618164, | |
| "learning_rate": 4.454041720990874e-06, | |
| "loss": 0.4014, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 4.267845501955671, | |
| "grad_norm": 33.210113525390625, | |
| "learning_rate": 4.392926988265972e-06, | |
| "loss": 0.3891, | |
| "step": 209500 | |
| }, | |
| { | |
| "epoch": 4.278031290743155, | |
| "grad_norm": 33.96702194213867, | |
| "learning_rate": 4.331812255541069e-06, | |
| "loss": 0.3891, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 4.2882170795306385, | |
| "grad_norm": 13.674240112304688, | |
| "learning_rate": 4.270697522816167e-06, | |
| "loss": 0.3743, | |
| "step": 210500 | |
| }, | |
| { | |
| "epoch": 4.298402868318123, | |
| "grad_norm": 1.694655179977417, | |
| "learning_rate": 4.209582790091264e-06, | |
| "loss": 0.4072, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 4.308588657105606, | |
| "grad_norm": 10.407938957214355, | |
| "learning_rate": 4.148468057366362e-06, | |
| "loss": 0.3788, | |
| "step": 211500 | |
| }, | |
| { | |
| "epoch": 4.31877444589309, | |
| "grad_norm": 28.451736450195312, | |
| "learning_rate": 4.087353324641461e-06, | |
| "loss": 0.44, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 4.328960234680574, | |
| "grad_norm": 40.655616760253906, | |
| "learning_rate": 4.026238591916559e-06, | |
| "loss": 0.3846, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 4.339146023468057, | |
| "grad_norm": 33.640682220458984, | |
| "learning_rate": 3.965123859191656e-06, | |
| "loss": 0.3967, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 4.349331812255541, | |
| "grad_norm": 77.51473236083984, | |
| "learning_rate": 3.904009126466754e-06, | |
| "loss": 0.384, | |
| "step": 213500 | |
| }, | |
| { | |
| "epoch": 4.3595176010430245, | |
| "grad_norm": 15.64946174621582, | |
| "learning_rate": 3.842894393741851e-06, | |
| "loss": 0.3935, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 4.369703389830509, | |
| "grad_norm": 38.03608703613281, | |
| "learning_rate": 3.7817796610169493e-06, | |
| "loss": 0.3933, | |
| "step": 214500 | |
| }, | |
| { | |
| "epoch": 4.379889178617992, | |
| "grad_norm": 24.385404586791992, | |
| "learning_rate": 3.720664928292047e-06, | |
| "loss": 0.3879, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 4.390074967405476, | |
| "grad_norm": 19.359683990478516, | |
| "learning_rate": 3.659550195567145e-06, | |
| "loss": 0.4027, | |
| "step": 215500 | |
| }, | |
| { | |
| "epoch": 4.400260756192959, | |
| "grad_norm": 20.058656692504883, | |
| "learning_rate": 3.598435462842243e-06, | |
| "loss": 0.3947, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 4.410446544980443, | |
| "grad_norm": 22.407522201538086, | |
| "learning_rate": 3.5373207301173403e-06, | |
| "loss": 0.3692, | |
| "step": 216500 | |
| }, | |
| { | |
| "epoch": 4.4206323337679265, | |
| "grad_norm": 0.43725308775901794, | |
| "learning_rate": 3.476205997392438e-06, | |
| "loss": 0.3876, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 4.430818122555411, | |
| "grad_norm": 6.114492416381836, | |
| "learning_rate": 3.4150912646675363e-06, | |
| "loss": 0.3812, | |
| "step": 217500 | |
| }, | |
| { | |
| "epoch": 4.441003911342895, | |
| "grad_norm": 8.737237930297852, | |
| "learning_rate": 3.353976531942634e-06, | |
| "loss": 0.4218, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 4.451189700130378, | |
| "grad_norm": 21.46657371520996, | |
| "learning_rate": 3.2928617992177314e-06, | |
| "loss": 0.3878, | |
| "step": 218500 | |
| }, | |
| { | |
| "epoch": 4.461375488917862, | |
| "grad_norm": 28.781583786010742, | |
| "learning_rate": 3.2317470664928294e-06, | |
| "loss": 0.4014, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 4.471561277705345, | |
| "grad_norm": 65.7254867553711, | |
| "learning_rate": 3.170632333767927e-06, | |
| "loss": 0.3846, | |
| "step": 219500 | |
| }, | |
| { | |
| "epoch": 4.481747066492829, | |
| "grad_norm": 42.237693786621094, | |
| "learning_rate": 3.109517601043025e-06, | |
| "loss": 0.3945, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 4.491932855280313, | |
| "grad_norm": 22.627384185791016, | |
| "learning_rate": 3.048402868318123e-06, | |
| "loss": 0.3846, | |
| "step": 220500 | |
| }, | |
| { | |
| "epoch": 4.502118644067797, | |
| "grad_norm": 5.28376579284668, | |
| "learning_rate": 2.9872881355932204e-06, | |
| "loss": 0.3951, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 4.51230443285528, | |
| "grad_norm": 29.31270980834961, | |
| "learning_rate": 2.926173402868318e-06, | |
| "loss": 0.387, | |
| "step": 221500 | |
| }, | |
| { | |
| "epoch": 4.522490221642764, | |
| "grad_norm": 28.06836700439453, | |
| "learning_rate": 2.8650586701434163e-06, | |
| "loss": 0.3758, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 4.532676010430247, | |
| "grad_norm": 19.28433609008789, | |
| "learning_rate": 2.803943937418514e-06, | |
| "loss": 0.4011, | |
| "step": 222500 | |
| }, | |
| { | |
| "epoch": 4.542861799217731, | |
| "grad_norm": 28.01453399658203, | |
| "learning_rate": 2.7428292046936114e-06, | |
| "loss": 0.4123, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 4.5530475880052155, | |
| "grad_norm": 22.563831329345703, | |
| "learning_rate": 2.6817144719687094e-06, | |
| "loss": 0.3901, | |
| "step": 223500 | |
| }, | |
| { | |
| "epoch": 4.563233376792699, | |
| "grad_norm": 6.881279945373535, | |
| "learning_rate": 2.620599739243807e-06, | |
| "loss": 0.3641, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 4.573419165580183, | |
| "grad_norm": 10.799131393432617, | |
| "learning_rate": 2.559485006518905e-06, | |
| "loss": 0.388, | |
| "step": 224500 | |
| }, | |
| { | |
| "epoch": 4.583604954367666, | |
| "grad_norm": 8.748978614807129, | |
| "learning_rate": 2.498370273794003e-06, | |
| "loss": 0.3855, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 4.59379074315515, | |
| "grad_norm": 30.88077163696289, | |
| "learning_rate": 2.4372555410691004e-06, | |
| "loss": 0.3846, | |
| "step": 225500 | |
| }, | |
| { | |
| "epoch": 4.603976531942633, | |
| "grad_norm": 9.857905387878418, | |
| "learning_rate": 2.376140808344198e-06, | |
| "loss": 0.3887, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 4.6141623207301175, | |
| "grad_norm": 18.84724235534668, | |
| "learning_rate": 2.3150260756192964e-06, | |
| "loss": 0.3981, | |
| "step": 226500 | |
| }, | |
| { | |
| "epoch": 4.624348109517601, | |
| "grad_norm": 1.5743709802627563, | |
| "learning_rate": 2.253911342894394e-06, | |
| "loss": 0.4198, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 4.634533898305085, | |
| "grad_norm": 1.2036515474319458, | |
| "learning_rate": 2.1927966101694915e-06, | |
| "loss": 0.3883, | |
| "step": 227500 | |
| }, | |
| { | |
| "epoch": 4.644719687092568, | |
| "grad_norm": 25.730968475341797, | |
| "learning_rate": 2.1316818774445895e-06, | |
| "loss": 0.3791, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 4.654905475880052, | |
| "grad_norm": 32.65802764892578, | |
| "learning_rate": 2.070567144719687e-06, | |
| "loss": 0.3772, | |
| "step": 228500 | |
| }, | |
| { | |
| "epoch": 4.665091264667536, | |
| "grad_norm": 3.2796003818511963, | |
| "learning_rate": 2.009452411994785e-06, | |
| "loss": 0.3994, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 4.675277053455019, | |
| "grad_norm": 12.903874397277832, | |
| "learning_rate": 1.948337679269883e-06, | |
| "loss": 0.3725, | |
| "step": 229500 | |
| }, | |
| { | |
| "epoch": 4.6854628422425035, | |
| "grad_norm": 1.0021297931671143, | |
| "learning_rate": 1.8872229465449805e-06, | |
| "loss": 0.393, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 4.695648631029987, | |
| "grad_norm": 5.547977447509766, | |
| "learning_rate": 1.8261082138200783e-06, | |
| "loss": 0.3907, | |
| "step": 230500 | |
| }, | |
| { | |
| "epoch": 4.705834419817471, | |
| "grad_norm": 5.375828742980957, | |
| "learning_rate": 1.764993481095176e-06, | |
| "loss": 0.4023, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 4.716020208604954, | |
| "grad_norm": 20.087383270263672, | |
| "learning_rate": 1.703878748370274e-06, | |
| "loss": 0.3652, | |
| "step": 231500 | |
| }, | |
| { | |
| "epoch": 4.726205997392438, | |
| "grad_norm": 19.617931365966797, | |
| "learning_rate": 1.6427640156453715e-06, | |
| "loss": 0.3842, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 4.736391786179921, | |
| "grad_norm": 22.911376953125, | |
| "learning_rate": 1.5816492829204695e-06, | |
| "loss": 0.3871, | |
| "step": 232500 | |
| }, | |
| { | |
| "epoch": 4.7465775749674055, | |
| "grad_norm": 31.00206184387207, | |
| "learning_rate": 1.5205345501955673e-06, | |
| "loss": 0.3918, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 4.756763363754889, | |
| "grad_norm": 39.066619873046875, | |
| "learning_rate": 1.4594198174706648e-06, | |
| "loss": 0.3743, | |
| "step": 233500 | |
| }, | |
| { | |
| "epoch": 4.766949152542373, | |
| "grad_norm": 9.460115432739258, | |
| "learning_rate": 1.3983050847457628e-06, | |
| "loss": 0.388, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 4.777134941329857, | |
| "grad_norm": 10.770241737365723, | |
| "learning_rate": 1.3371903520208605e-06, | |
| "loss": 0.3986, | |
| "step": 234500 | |
| }, | |
| { | |
| "epoch": 4.78732073011734, | |
| "grad_norm": 29.36595916748047, | |
| "learning_rate": 1.2760756192959583e-06, | |
| "loss": 0.3836, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 4.797506518904824, | |
| "grad_norm": 21.498857498168945, | |
| "learning_rate": 1.214960886571056e-06, | |
| "loss": 0.3839, | |
| "step": 235500 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "grad_norm": 2.3236119747161865, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 0.3983, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 4.817878096479792, | |
| "grad_norm": 10.451170921325684, | |
| "learning_rate": 1.0927314211212516e-06, | |
| "loss": 0.3853, | |
| "step": 236500 | |
| }, | |
| { | |
| "epoch": 4.828063885267275, | |
| "grad_norm": 18.35541343688965, | |
| "learning_rate": 1.0316166883963496e-06, | |
| "loss": 0.3779, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 4.838249674054759, | |
| "grad_norm": 56.08652877807617, | |
| "learning_rate": 9.705019556714473e-07, | |
| "loss": 0.3608, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 4.848435462842242, | |
| "grad_norm": 30.668987274169922, | |
| "learning_rate": 9.09387222946545e-07, | |
| "loss": 0.3901, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 4.858621251629726, | |
| "grad_norm": 44.03152084350586, | |
| "learning_rate": 8.482724902216428e-07, | |
| "loss": 0.3742, | |
| "step": 238500 | |
| }, | |
| { | |
| "epoch": 4.8688070404172095, | |
| "grad_norm": 46.421485900878906, | |
| "learning_rate": 7.871577574967406e-07, | |
| "loss": 0.4039, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 4.878992829204694, | |
| "grad_norm": 41.14004898071289, | |
| "learning_rate": 7.260430247718384e-07, | |
| "loss": 0.4113, | |
| "step": 239500 | |
| }, | |
| { | |
| "epoch": 4.889178617992178, | |
| "grad_norm": 3.9679319858551025, | |
| "learning_rate": 6.649282920469362e-07, | |
| "loss": 0.3856, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 4.899364406779661, | |
| "grad_norm": 30.101577758789062, | |
| "learning_rate": 6.038135593220339e-07, | |
| "loss": 0.3742, | |
| "step": 240500 | |
| }, | |
| { | |
| "epoch": 4.909550195567145, | |
| "grad_norm": 12.514345169067383, | |
| "learning_rate": 5.426988265971316e-07, | |
| "loss": 0.3795, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 4.919735984354628, | |
| "grad_norm": 33.0556526184082, | |
| "learning_rate": 4.815840938722295e-07, | |
| "loss": 0.3865, | |
| "step": 241500 | |
| }, | |
| { | |
| "epoch": 4.929921773142112, | |
| "grad_norm": 36.99669647216797, | |
| "learning_rate": 4.2046936114732726e-07, | |
| "loss": 0.4271, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 4.9401075619295955, | |
| "grad_norm": 2.8075144290924072, | |
| "learning_rate": 3.59354628422425e-07, | |
| "loss": 0.3867, | |
| "step": 242500 | |
| }, | |
| { | |
| "epoch": 4.95029335071708, | |
| "grad_norm": 0.9676657319068909, | |
| "learning_rate": 2.9823989569752284e-07, | |
| "loss": 0.3994, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 4.960479139504563, | |
| "grad_norm": 26.419252395629883, | |
| "learning_rate": 2.3712516297262062e-07, | |
| "loss": 0.3956, | |
| "step": 243500 | |
| }, | |
| { | |
| "epoch": 4.970664928292047, | |
| "grad_norm": 37.89094543457031, | |
| "learning_rate": 1.7601043024771838e-07, | |
| "loss": 0.3856, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 4.98085071707953, | |
| "grad_norm": 20.763473510742188, | |
| "learning_rate": 1.1489569752281617e-07, | |
| "loss": 0.4161, | |
| "step": 244500 | |
| }, | |
| { | |
| "epoch": 4.991036505867014, | |
| "grad_norm": 10.952351570129395, | |
| "learning_rate": 5.378096479791395e-08, | |
| "loss": 0.4074, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7310513257980347, | |
| "eval_loss": 1.0102216005325317, | |
| "eval_runtime": 7.6909, | |
| "eval_samples_per_second": 638.153, | |
| "eval_steps_per_second": 79.834, | |
| "step": 245440 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 245440, | |
| "total_flos": 1.158449687808768e+17, | |
| "train_loss": 0.5704954994082295, | |
| "train_runtime": 14864.7666, | |
| "train_samples_per_second": 132.092, | |
| "train_steps_per_second": 16.512 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 245440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.001 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.158449687808768e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |