| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 2000, | |
| "global_step": 21900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00684931506849315, | |
| "grad_norm": 55.08233642578125, | |
| "learning_rate": 2.9943835616438356e-05, | |
| "loss": 4.9626, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0136986301369863, | |
| "grad_norm": 61.599609375, | |
| "learning_rate": 2.987808219178082e-05, | |
| "loss": 3.2637, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02054794520547945, | |
| "grad_norm": 54.39605712890625, | |
| "learning_rate": 2.980958904109589e-05, | |
| "loss": 2.5155, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 35.411338806152344, | |
| "learning_rate": 2.974109589041096e-05, | |
| "loss": 2.3836, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03424657534246575, | |
| "grad_norm": 38.87581253051758, | |
| "learning_rate": 2.9672602739726026e-05, | |
| "loss": 2.1592, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0410958904109589, | |
| "grad_norm": 43.26409149169922, | |
| "learning_rate": 2.9604109589041095e-05, | |
| "loss": 2.0851, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04794520547945205, | |
| "grad_norm": 32.2227897644043, | |
| "learning_rate": 2.9535616438356165e-05, | |
| "loss": 1.833, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 34.12604904174805, | |
| "learning_rate": 2.9467123287671234e-05, | |
| "loss": 1.7324, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06164383561643835, | |
| "grad_norm": 802.5269165039062, | |
| "learning_rate": 2.93986301369863e-05, | |
| "loss": 1.715, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0684931506849315, | |
| "grad_norm": 21.16217803955078, | |
| "learning_rate": 2.933013698630137e-05, | |
| "loss": 1.7966, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.07534246575342465, | |
| "grad_norm": 53.79206466674805, | |
| "learning_rate": 2.926164383561644e-05, | |
| "loss": 1.7756, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 69.78266906738281, | |
| "learning_rate": 2.919315068493151e-05, | |
| "loss": 1.6611, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08904109589041095, | |
| "grad_norm": 19.355026245117188, | |
| "learning_rate": 2.9124657534246575e-05, | |
| "loss": 1.5454, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.0958904109589041, | |
| "grad_norm": 32.797298431396484, | |
| "learning_rate": 2.9056164383561644e-05, | |
| "loss": 1.533, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.10273972602739725, | |
| "grad_norm": 22.2159423828125, | |
| "learning_rate": 2.8987671232876714e-05, | |
| "loss": 1.555, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 41.200904846191406, | |
| "learning_rate": 2.8919178082191783e-05, | |
| "loss": 1.5322, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.11643835616438356, | |
| "grad_norm": 22.64232635498047, | |
| "learning_rate": 2.885068493150685e-05, | |
| "loss": 1.5337, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1232876712328767, | |
| "grad_norm": 38.510257720947266, | |
| "learning_rate": 2.878219178082192e-05, | |
| "loss": 1.5001, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.13013698630136986, | |
| "grad_norm": 23.04859733581543, | |
| "learning_rate": 2.871369863013699e-05, | |
| "loss": 1.4488, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 23.516109466552734, | |
| "learning_rate": 2.8645205479452058e-05, | |
| "loss": 1.3247, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14383561643835616, | |
| "grad_norm": 24.6121883392334, | |
| "learning_rate": 2.8576712328767124e-05, | |
| "loss": 1.3537, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1506849315068493, | |
| "grad_norm": 29.713895797729492, | |
| "learning_rate": 2.8508219178082194e-05, | |
| "loss": 1.4116, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.15753424657534246, | |
| "grad_norm": 102.77496337890625, | |
| "learning_rate": 2.8439726027397263e-05, | |
| "loss": 1.3196, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 46.55476379394531, | |
| "learning_rate": 2.8371232876712332e-05, | |
| "loss": 1.3987, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.17123287671232876, | |
| "grad_norm": 22.365116119384766, | |
| "learning_rate": 2.8302739726027395e-05, | |
| "loss": 1.3027, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1780821917808219, | |
| "grad_norm": 48.17926788330078, | |
| "learning_rate": 2.8234246575342465e-05, | |
| "loss": 1.3698, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.18493150684931506, | |
| "grad_norm": 29.022974014282227, | |
| "learning_rate": 2.8165753424657534e-05, | |
| "loss": 1.2422, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 23.812314987182617, | |
| "learning_rate": 2.8097260273972604e-05, | |
| "loss": 1.2691, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.19863013698630136, | |
| "grad_norm": 18.73370361328125, | |
| "learning_rate": 2.802876712328767e-05, | |
| "loss": 1.3611, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.2054794520547945, | |
| "grad_norm": 23.951034545898438, | |
| "learning_rate": 2.796027397260274e-05, | |
| "loss": 1.3148, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21232876712328766, | |
| "grad_norm": 33.09256362915039, | |
| "learning_rate": 2.789178082191781e-05, | |
| "loss": 1.3978, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 23.944395065307617, | |
| "learning_rate": 2.7823287671232878e-05, | |
| "loss": 1.2173, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.22602739726027396, | |
| "grad_norm": 16.93844223022461, | |
| "learning_rate": 2.7754794520547944e-05, | |
| "loss": 1.3082, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.2328767123287671, | |
| "grad_norm": 21.014097213745117, | |
| "learning_rate": 2.7686301369863014e-05, | |
| "loss": 1.3363, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23972602739726026, | |
| "grad_norm": 34.768096923828125, | |
| "learning_rate": 2.7617808219178083e-05, | |
| "loss": 1.1646, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 19.525144577026367, | |
| "learning_rate": 2.7549315068493153e-05, | |
| "loss": 1.2566, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.2534246575342466, | |
| "grad_norm": 17.319896697998047, | |
| "learning_rate": 2.748082191780822e-05, | |
| "loss": 1.2123, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.2602739726027397, | |
| "grad_norm": 21.581501007080078, | |
| "learning_rate": 2.7412328767123288e-05, | |
| "loss": 1.4083, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2671232876712329, | |
| "grad_norm": 38.267120361328125, | |
| "learning_rate": 2.7343835616438358e-05, | |
| "loss": 1.1987, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 64.0147933959961, | |
| "learning_rate": 2.7275342465753427e-05, | |
| "loss": 1.1474, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "eval_exact_match": 71.11636707663197, | |
| "eval_f1": 81.37469585294956, | |
| "eval_runtime": 406.8449, | |
| "eval_samples_per_second": 25.98, | |
| "eval_steps_per_second": 1.625, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2808219178082192, | |
| "grad_norm": 22.60133171081543, | |
| "learning_rate": 2.7206849315068493e-05, | |
| "loss": 1.3887, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2876712328767123, | |
| "grad_norm": 51.95814514160156, | |
| "learning_rate": 2.7138356164383563e-05, | |
| "loss": 1.2049, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2945205479452055, | |
| "grad_norm": 11.237629890441895, | |
| "learning_rate": 2.7069863013698632e-05, | |
| "loss": 1.1684, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 23.52985954284668, | |
| "learning_rate": 2.70013698630137e-05, | |
| "loss": 1.2162, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3082191780821918, | |
| "grad_norm": 19.681074142456055, | |
| "learning_rate": 2.6932876712328768e-05, | |
| "loss": 1.0853, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3150684931506849, | |
| "grad_norm": 13.592674255371094, | |
| "learning_rate": 2.6864383561643837e-05, | |
| "loss": 1.2977, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3219178082191781, | |
| "grad_norm": 30.62900733947754, | |
| "learning_rate": 2.6795890410958907e-05, | |
| "loss": 1.1385, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 15.438138961791992, | |
| "learning_rate": 2.6727397260273976e-05, | |
| "loss": 1.1983, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3356164383561644, | |
| "grad_norm": 21.04355812072754, | |
| "learning_rate": 2.6658904109589042e-05, | |
| "loss": 1.2473, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 101.09291076660156, | |
| "learning_rate": 2.659041095890411e-05, | |
| "loss": 1.2827, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3493150684931507, | |
| "grad_norm": 24.547096252441406, | |
| "learning_rate": 2.6521917808219178e-05, | |
| "loss": 1.2088, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 17.362468719482422, | |
| "learning_rate": 2.6453424657534247e-05, | |
| "loss": 1.1586, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.363013698630137, | |
| "grad_norm": 22.94930076599121, | |
| "learning_rate": 2.6384931506849313e-05, | |
| "loss": 1.1435, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.3698630136986301, | |
| "grad_norm": 9.875961303710938, | |
| "learning_rate": 2.6316438356164383e-05, | |
| "loss": 1.2004, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3767123287671233, | |
| "grad_norm": 39.17120361328125, | |
| "learning_rate": 2.6247945205479452e-05, | |
| "loss": 1.1009, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 29.799320220947266, | |
| "learning_rate": 2.6179452054794522e-05, | |
| "loss": 1.1917, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.3904109589041096, | |
| "grad_norm": 15.327347755432129, | |
| "learning_rate": 2.6110958904109588e-05, | |
| "loss": 1.1141, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3972602739726027, | |
| "grad_norm": 63.62096405029297, | |
| "learning_rate": 2.6042465753424657e-05, | |
| "loss": 1.1083, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4041095890410959, | |
| "grad_norm": 77.25872802734375, | |
| "learning_rate": 2.5973972602739727e-05, | |
| "loss": 1.0962, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 26.37015724182129, | |
| "learning_rate": 2.5905479452054796e-05, | |
| "loss": 1.1158, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4178082191780822, | |
| "grad_norm": 30.89618682861328, | |
| "learning_rate": 2.5836986301369862e-05, | |
| "loss": 1.1356, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.4246575342465753, | |
| "grad_norm": 26.392322540283203, | |
| "learning_rate": 2.5768493150684932e-05, | |
| "loss": 0.9842, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4315068493150685, | |
| "grad_norm": 19.746112823486328, | |
| "learning_rate": 2.57e-05, | |
| "loss": 1.0477, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 21.304296493530273, | |
| "learning_rate": 2.563150684931507e-05, | |
| "loss": 1.1522, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4452054794520548, | |
| "grad_norm": 21.67245864868164, | |
| "learning_rate": 2.5563013698630137e-05, | |
| "loss": 1.0677, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.4520547945205479, | |
| "grad_norm": 17.96895980834961, | |
| "learning_rate": 2.5494520547945206e-05, | |
| "loss": 1.0657, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4589041095890411, | |
| "grad_norm": 19.414440155029297, | |
| "learning_rate": 2.5426027397260276e-05, | |
| "loss": 1.1109, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 20.33131217956543, | |
| "learning_rate": 2.5357534246575345e-05, | |
| "loss": 1.0958, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4726027397260274, | |
| "grad_norm": 35.86481475830078, | |
| "learning_rate": 2.528904109589041e-05, | |
| "loss": 1.1134, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.4794520547945205, | |
| "grad_norm": 20.007892608642578, | |
| "learning_rate": 2.522054794520548e-05, | |
| "loss": 1.0236, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4863013698630137, | |
| "grad_norm": 8.771307945251465, | |
| "learning_rate": 2.515205479452055e-05, | |
| "loss": 1.0518, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 31.425060272216797, | |
| "learning_rate": 2.508356164383562e-05, | |
| "loss": 1.0397, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 33.32624816894531, | |
| "learning_rate": 2.5015068493150686e-05, | |
| "loss": 1.0374, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 15.538783073425293, | |
| "learning_rate": 2.4946575342465755e-05, | |
| "loss": 1.2677, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5136986301369864, | |
| "grad_norm": 32.81166076660156, | |
| "learning_rate": 2.4878082191780825e-05, | |
| "loss": 1.0274, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 40.134246826171875, | |
| "learning_rate": 2.480958904109589e-05, | |
| "loss": 1.0493, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5273972602739726, | |
| "grad_norm": 19.939796447753906, | |
| "learning_rate": 2.4741095890410957e-05, | |
| "loss": 1.1492, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5342465753424658, | |
| "grad_norm": 23.643762588500977, | |
| "learning_rate": 2.4672602739726026e-05, | |
| "loss": 1.0892, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.541095890410959, | |
| "grad_norm": 28.48002052307129, | |
| "learning_rate": 2.4604109589041096e-05, | |
| "loss": 1.144, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 15.827131271362305, | |
| "learning_rate": 2.4535616438356165e-05, | |
| "loss": 1.117, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_exact_match": 78.61873226111636, | |
| "eval_f1": 86.50538268692839, | |
| "eval_runtime": 407.223, | |
| "eval_samples_per_second": 25.956, | |
| "eval_steps_per_second": 1.623, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5547945205479452, | |
| "grad_norm": 17.803165435791016, | |
| "learning_rate": 2.446712328767123e-05, | |
| "loss": 1.1428, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.5616438356164384, | |
| "grad_norm": 15.47031307220459, | |
| "learning_rate": 2.43986301369863e-05, | |
| "loss": 1.1178, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5684931506849316, | |
| "grad_norm": 20.780733108520508, | |
| "learning_rate": 2.433013698630137e-05, | |
| "loss": 1.0105, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 33.10541915893555, | |
| "learning_rate": 2.426164383561644e-05, | |
| "loss": 1.0502, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5821917808219178, | |
| "grad_norm": 19.134233474731445, | |
| "learning_rate": 2.4193150684931506e-05, | |
| "loss": 1.1227, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.589041095890411, | |
| "grad_norm": 10.8368558883667, | |
| "learning_rate": 2.4124657534246575e-05, | |
| "loss": 1.0697, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5958904109589042, | |
| "grad_norm": 18.496990203857422, | |
| "learning_rate": 2.4057534246575344e-05, | |
| "loss": 1.0705, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 16.98480987548828, | |
| "learning_rate": 2.398904109589041e-05, | |
| "loss": 1.135, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6095890410958904, | |
| "grad_norm": 35.130393981933594, | |
| "learning_rate": 2.392054794520548e-05, | |
| "loss": 1.1933, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6164383561643836, | |
| "grad_norm": 71.43773651123047, | |
| "learning_rate": 2.385205479452055e-05, | |
| "loss": 1.0783, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6232876712328768, | |
| "grad_norm": 21.982946395874023, | |
| "learning_rate": 2.378356164383562e-05, | |
| "loss": 1.0517, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 21.33681297302246, | |
| "learning_rate": 2.3715068493150685e-05, | |
| "loss": 1.0829, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.636986301369863, | |
| "grad_norm": 14.706864356994629, | |
| "learning_rate": 2.3646575342465754e-05, | |
| "loss": 1.0452, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.6438356164383562, | |
| "grad_norm": 19.40340805053711, | |
| "learning_rate": 2.3578082191780824e-05, | |
| "loss": 1.0843, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6506849315068494, | |
| "grad_norm": 66.0240478515625, | |
| "learning_rate": 2.3509589041095893e-05, | |
| "loss": 0.9722, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 30.19057273864746, | |
| "learning_rate": 2.344109589041096e-05, | |
| "loss": 0.8738, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.6643835616438356, | |
| "grad_norm": 24.212846755981445, | |
| "learning_rate": 2.337260273972603e-05, | |
| "loss": 0.9865, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.6712328767123288, | |
| "grad_norm": 24.953773498535156, | |
| "learning_rate": 2.3304109589041098e-05, | |
| "loss": 0.9409, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.678082191780822, | |
| "grad_norm": 32.05400466918945, | |
| "learning_rate": 2.3235616438356168e-05, | |
| "loss": 0.9674, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 20.69804573059082, | |
| "learning_rate": 2.3167123287671234e-05, | |
| "loss": 0.9547, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6917808219178082, | |
| "grad_norm": 9.982364654541016, | |
| "learning_rate": 2.3098630136986303e-05, | |
| "loss": 0.9724, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.6986301369863014, | |
| "grad_norm": 24.489803314208984, | |
| "learning_rate": 2.3030136986301373e-05, | |
| "loss": 1.0141, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7054794520547946, | |
| "grad_norm": 22.05426597595215, | |
| "learning_rate": 2.296164383561644e-05, | |
| "loss": 1.0011, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 11.475118637084961, | |
| "learning_rate": 2.2893150684931505e-05, | |
| "loss": 1.0293, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7191780821917808, | |
| "grad_norm": 24.616731643676758, | |
| "learning_rate": 2.2824657534246574e-05, | |
| "loss": 1.0347, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.726027397260274, | |
| "grad_norm": 13.638339042663574, | |
| "learning_rate": 2.2756164383561644e-05, | |
| "loss": 0.9843, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.7328767123287672, | |
| "grad_norm": 26.214601516723633, | |
| "learning_rate": 2.2687671232876713e-05, | |
| "loss": 0.9909, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 18.321882247924805, | |
| "learning_rate": 2.261917808219178e-05, | |
| "loss": 0.9893, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7465753424657534, | |
| "grad_norm": 14.493846893310547, | |
| "learning_rate": 2.255068493150685e-05, | |
| "loss": 1.0089, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "grad_norm": 29.463472366333008, | |
| "learning_rate": 2.248219178082192e-05, | |
| "loss": 1.052, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7602739726027398, | |
| "grad_norm": 16.674697875976562, | |
| "learning_rate": 2.2413698630136988e-05, | |
| "loss": 0.977, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 20.96110725402832, | |
| "learning_rate": 2.2345205479452054e-05, | |
| "loss": 0.9393, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.773972602739726, | |
| "grad_norm": 10.15765380859375, | |
| "learning_rate": 2.2276712328767123e-05, | |
| "loss": 0.9647, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.7808219178082192, | |
| "grad_norm": 18.204395294189453, | |
| "learning_rate": 2.2208219178082193e-05, | |
| "loss": 1.0376, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7876712328767124, | |
| "grad_norm": 12.090936660766602, | |
| "learning_rate": 2.2139726027397262e-05, | |
| "loss": 1.0562, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 23.062904357910156, | |
| "learning_rate": 2.207123287671233e-05, | |
| "loss": 1.0397, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.8013698630136986, | |
| "grad_norm": 13.404748916625977, | |
| "learning_rate": 2.2002739726027398e-05, | |
| "loss": 1.0177, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.8082191780821918, | |
| "grad_norm": 16.615171432495117, | |
| "learning_rate": 2.1934246575342467e-05, | |
| "loss": 0.9627, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.815068493150685, | |
| "grad_norm": 24.662452697753906, | |
| "learning_rate": 2.1865753424657537e-05, | |
| "loss": 0.9447, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 9.534379959106445, | |
| "learning_rate": 2.1797260273972603e-05, | |
| "loss": 1.0024, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "eval_exact_match": 80.7379375591296, | |
| "eval_f1": 88.15133262306199, | |
| "eval_runtime": 405.8112, | |
| "eval_samples_per_second": 26.047, | |
| "eval_steps_per_second": 1.629, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8287671232876712, | |
| "grad_norm": 31.16766929626465, | |
| "learning_rate": 2.1728767123287672e-05, | |
| "loss": 0.9153, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.8356164383561644, | |
| "grad_norm": 21.178213119506836, | |
| "learning_rate": 2.1660273972602742e-05, | |
| "loss": 0.919, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8424657534246576, | |
| "grad_norm": 16.18219566345215, | |
| "learning_rate": 2.159178082191781e-05, | |
| "loss": 1.0232, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 10.178057670593262, | |
| "learning_rate": 2.1523287671232877e-05, | |
| "loss": 1.0113, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8561643835616438, | |
| "grad_norm": 32.234249114990234, | |
| "learning_rate": 2.1454794520547947e-05, | |
| "loss": 0.9115, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.863013698630137, | |
| "grad_norm": 10.931483268737793, | |
| "learning_rate": 2.1386301369863016e-05, | |
| "loss": 0.9493, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8698630136986302, | |
| "grad_norm": 18.961946487426758, | |
| "learning_rate": 2.1317808219178086e-05, | |
| "loss": 0.9662, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 32.1358642578125, | |
| "learning_rate": 2.125068493150685e-05, | |
| "loss": 0.9122, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8835616438356164, | |
| "grad_norm": 27.434032440185547, | |
| "learning_rate": 2.118219178082192e-05, | |
| "loss": 0.9665, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.8904109589041096, | |
| "grad_norm": 13.604903221130371, | |
| "learning_rate": 2.1113698630136987e-05, | |
| "loss": 0.9752, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8972602739726028, | |
| "grad_norm": 13.965779304504395, | |
| "learning_rate": 2.1045205479452053e-05, | |
| "loss": 0.8704, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 18.716514587402344, | |
| "learning_rate": 2.0976712328767122e-05, | |
| "loss": 0.9159, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.910958904109589, | |
| "grad_norm": 19.964460372924805, | |
| "learning_rate": 2.0908219178082192e-05, | |
| "loss": 0.9359, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.9178082191780822, | |
| "grad_norm": 13.847062110900879, | |
| "learning_rate": 2.083972602739726e-05, | |
| "loss": 0.8623, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.9246575342465754, | |
| "grad_norm": 25.339975357055664, | |
| "learning_rate": 2.0771232876712327e-05, | |
| "loss": 0.9804, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 16.102794647216797, | |
| "learning_rate": 2.0702739726027397e-05, | |
| "loss": 0.8721, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9383561643835616, | |
| "grad_norm": 15.555779457092285, | |
| "learning_rate": 2.0634246575342466e-05, | |
| "loss": 0.9942, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.9452054794520548, | |
| "grad_norm": 14.534811019897461, | |
| "learning_rate": 2.0565753424657536e-05, | |
| "loss": 0.9333, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.952054794520548, | |
| "grad_norm": 24.963340759277344, | |
| "learning_rate": 2.0497260273972602e-05, | |
| "loss": 0.966, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 42.927734375, | |
| "learning_rate": 2.042876712328767e-05, | |
| "loss": 0.913, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9657534246575342, | |
| "grad_norm": 14.816558837890625, | |
| "learning_rate": 2.036027397260274e-05, | |
| "loss": 0.9717, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.9726027397260274, | |
| "grad_norm": 27.83946418762207, | |
| "learning_rate": 2.029178082191781e-05, | |
| "loss": 0.9855, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.9794520547945206, | |
| "grad_norm": 23.87200164794922, | |
| "learning_rate": 2.0223287671232876e-05, | |
| "loss": 0.9934, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 30.313343048095703, | |
| "learning_rate": 2.0154794520547946e-05, | |
| "loss": 1.0148, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.9931506849315068, | |
| "grad_norm": 23.50067138671875, | |
| "learning_rate": 2.0086301369863015e-05, | |
| "loss": 0.9578, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 17.968570709228516, | |
| "learning_rate": 2.0017808219178085e-05, | |
| "loss": 0.821, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.0068493150684932, | |
| "grad_norm": 44.593414306640625, | |
| "learning_rate": 1.994931506849315e-05, | |
| "loss": 0.6823, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 12.959754943847656, | |
| "learning_rate": 1.988082191780822e-05, | |
| "loss": 0.6937, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.0205479452054795, | |
| "grad_norm": 20.291378021240234, | |
| "learning_rate": 1.981232876712329e-05, | |
| "loss": 0.6457, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.0273972602739727, | |
| "grad_norm": 5.460036277770996, | |
| "learning_rate": 1.974383561643836e-05, | |
| "loss": 0.729, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0342465753424657, | |
| "grad_norm": 17.41657257080078, | |
| "learning_rate": 1.9675342465753425e-05, | |
| "loss": 0.7032, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 10.595620155334473, | |
| "learning_rate": 1.9606849315068495e-05, | |
| "loss": 0.6788, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.047945205479452, | |
| "grad_norm": 32.301509857177734, | |
| "learning_rate": 1.9538356164383564e-05, | |
| "loss": 0.669, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.0547945205479452, | |
| "grad_norm": 5.369728088378906, | |
| "learning_rate": 1.9469863013698634e-05, | |
| "loss": 0.6169, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.0616438356164384, | |
| "grad_norm": 38.604183197021484, | |
| "learning_rate": 1.94013698630137e-05, | |
| "loss": 0.6585, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 12.535406112670898, | |
| "learning_rate": 1.9332876712328766e-05, | |
| "loss": 0.6448, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.0753424657534247, | |
| "grad_norm": 11.292516708374023, | |
| "learning_rate": 1.9264383561643835e-05, | |
| "loss": 0.6429, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.0821917808219177, | |
| "grad_norm": 15.400158882141113, | |
| "learning_rate": 1.9195890410958905e-05, | |
| "loss": 0.6553, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.0890410958904109, | |
| "grad_norm": 28.782424926757812, | |
| "learning_rate": 1.912739726027397e-05, | |
| "loss": 0.7737, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 18.407026290893555, | |
| "learning_rate": 1.905890410958904e-05, | |
| "loss": 0.5717, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 81.46641438032167, | |
| "eval_f1": 88.73804254841154, | |
| "eval_runtime": 407.6234, | |
| "eval_samples_per_second": 25.931, | |
| "eval_steps_per_second": 1.622, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1027397260273972, | |
| "grad_norm": 54.460880279541016, | |
| "learning_rate": 1.899041095890411e-05, | |
| "loss": 0.6957, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.1095890410958904, | |
| "grad_norm": 10.459880828857422, | |
| "learning_rate": 1.892191780821918e-05, | |
| "loss": 0.6871, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.1164383561643836, | |
| "grad_norm": 21.884292602539062, | |
| "learning_rate": 1.8853424657534245e-05, | |
| "loss": 0.7586, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 18.18182373046875, | |
| "learning_rate": 1.8784931506849315e-05, | |
| "loss": 0.7294, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.13013698630137, | |
| "grad_norm": 14.117548942565918, | |
| "learning_rate": 1.8716438356164384e-05, | |
| "loss": 0.6837, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.1369863013698631, | |
| "grad_norm": 20.044261932373047, | |
| "learning_rate": 1.8647945205479454e-05, | |
| "loss": 0.7353, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.143835616438356, | |
| "grad_norm": 16.077611923217773, | |
| "learning_rate": 1.857945205479452e-05, | |
| "loss": 0.6038, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 28.85369300842285, | |
| "learning_rate": 1.851095890410959e-05, | |
| "loss": 0.6414, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.1575342465753424, | |
| "grad_norm": 14.509927749633789, | |
| "learning_rate": 1.844246575342466e-05, | |
| "loss": 0.7068, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.1643835616438356, | |
| "grad_norm": 14.299630165100098, | |
| "learning_rate": 1.837397260273973e-05, | |
| "loss": 0.7094, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.1712328767123288, | |
| "grad_norm": 11.768505096435547, | |
| "learning_rate": 1.8305479452054794e-05, | |
| "loss": 0.6914, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 42.37126922607422, | |
| "learning_rate": 1.8236986301369864e-05, | |
| "loss": 0.6539, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.1849315068493151, | |
| "grad_norm": 14.21442985534668, | |
| "learning_rate": 1.8168493150684933e-05, | |
| "loss": 0.7302, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.191780821917808, | |
| "grad_norm": 19.04937171936035, | |
| "learning_rate": 1.8100000000000003e-05, | |
| "loss": 0.7685, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.1986301369863013, | |
| "grad_norm": 17.215967178344727, | |
| "learning_rate": 1.803150684931507e-05, | |
| "loss": 0.6888, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 16.23516082763672, | |
| "learning_rate": 1.796301369863014e-05, | |
| "loss": 0.6825, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.2123287671232876, | |
| "grad_norm": 39.78145217895508, | |
| "learning_rate": 1.7894520547945208e-05, | |
| "loss": 0.7799, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.2191780821917808, | |
| "grad_norm": 26.684986114501953, | |
| "learning_rate": 1.7826027397260277e-05, | |
| "loss": 0.7425, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.226027397260274, | |
| "grad_norm": 35.67079544067383, | |
| "learning_rate": 1.7757534246575343e-05, | |
| "loss": 0.7405, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 18.228994369506836, | |
| "learning_rate": 1.7689041095890413e-05, | |
| "loss": 0.5968, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2397260273972603, | |
| "grad_norm": 15.768519401550293, | |
| "learning_rate": 1.7620547945205482e-05, | |
| "loss": 0.6791, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.2465753424657535, | |
| "grad_norm": 26.350936889648438, | |
| "learning_rate": 1.7552054794520545e-05, | |
| "loss": 0.7312, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.2534246575342465, | |
| "grad_norm": 22.138206481933594, | |
| "learning_rate": 1.7483561643835615e-05, | |
| "loss": 0.6496, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 20.530515670776367, | |
| "learning_rate": 1.7415068493150684e-05, | |
| "loss": 0.6592, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.2671232876712328, | |
| "grad_norm": 18.60872459411621, | |
| "learning_rate": 1.7346575342465754e-05, | |
| "loss": 0.7109, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.273972602739726, | |
| "grad_norm": 10.934627532958984, | |
| "learning_rate": 1.727808219178082e-05, | |
| "loss": 0.6364, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.2808219178082192, | |
| "grad_norm": 24.400938034057617, | |
| "learning_rate": 1.720958904109589e-05, | |
| "loss": 0.6624, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 33.15473556518555, | |
| "learning_rate": 1.714109589041096e-05, | |
| "loss": 0.7526, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.2945205479452055, | |
| "grad_norm": 4.9253339767456055, | |
| "learning_rate": 1.7072602739726028e-05, | |
| "loss": 0.6094, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.3013698630136985, | |
| "grad_norm": 26.45025634765625, | |
| "learning_rate": 1.7004109589041094e-05, | |
| "loss": 0.6779, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.308219178082192, | |
| "grad_norm": 12.181562423706055, | |
| "learning_rate": 1.6935616438356164e-05, | |
| "loss": 0.7728, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 6.9998040199279785, | |
| "learning_rate": 1.6867123287671233e-05, | |
| "loss": 0.6221, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.321917808219178, | |
| "grad_norm": 40.702369689941406, | |
| "learning_rate": 1.6798630136986303e-05, | |
| "loss": 0.7124, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.3287671232876712, | |
| "grad_norm": 18.84299659729004, | |
| "learning_rate": 1.673013698630137e-05, | |
| "loss": 0.6579, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.3356164383561644, | |
| "grad_norm": 24.911535263061523, | |
| "learning_rate": 1.6661643835616438e-05, | |
| "loss": 0.7197, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 24.64927101135254, | |
| "learning_rate": 1.6593150684931508e-05, | |
| "loss": 0.6767, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.3493150684931507, | |
| "grad_norm": 11.854528427124023, | |
| "learning_rate": 1.6524657534246577e-05, | |
| "loss": 0.6726, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.356164383561644, | |
| "grad_norm": 38.699310302734375, | |
| "learning_rate": 1.6456164383561643e-05, | |
| "loss": 0.6484, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.3630136986301369, | |
| "grad_norm": 21.76763916015625, | |
| "learning_rate": 1.6387671232876713e-05, | |
| "loss": 0.6959, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 15.432331085205078, | |
| "learning_rate": 1.6319178082191782e-05, | |
| "loss": 0.6585, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "eval_exact_match": 80.78524124881741, | |
| "eval_f1": 88.39324523394289, | |
| "eval_runtime": 407.758, | |
| "eval_samples_per_second": 25.922, | |
| "eval_steps_per_second": 1.621, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.3767123287671232, | |
| "grad_norm": 21.714828491210938, | |
| "learning_rate": 1.625068493150685e-05, | |
| "loss": 0.6976, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.3835616438356164, | |
| "grad_norm": 6.602792739868164, | |
| "learning_rate": 1.6182191780821918e-05, | |
| "loss": 0.6094, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.3904109589041096, | |
| "grad_norm": 42.471412658691406, | |
| "learning_rate": 1.6113698630136987e-05, | |
| "loss": 0.7151, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 11.658584594726562, | |
| "learning_rate": 1.6045205479452057e-05, | |
| "loss": 0.639, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.404109589041096, | |
| "grad_norm": 37.821659088134766, | |
| "learning_rate": 1.5976712328767126e-05, | |
| "loss": 0.6079, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.410958904109589, | |
| "grad_norm": 5.837065696716309, | |
| "learning_rate": 1.5908219178082192e-05, | |
| "loss": 0.5878, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.4178082191780823, | |
| "grad_norm": 67.67562866210938, | |
| "learning_rate": 1.5839726027397258e-05, | |
| "loss": 0.7724, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 2.624040365219116, | |
| "learning_rate": 1.5771232876712328e-05, | |
| "loss": 0.6753, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.4315068493150684, | |
| "grad_norm": 32.9188232421875, | |
| "learning_rate": 1.5702739726027397e-05, | |
| "loss": 0.6534, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.4383561643835616, | |
| "grad_norm": 32.16576385498047, | |
| "learning_rate": 1.5634246575342463e-05, | |
| "loss": 0.716, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.4452054794520548, | |
| "grad_norm": 19.11595344543457, | |
| "learning_rate": 1.5565753424657533e-05, | |
| "loss": 0.6081, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 13.033523559570312, | |
| "learning_rate": 1.5497260273972602e-05, | |
| "loss": 0.7286, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.4589041095890412, | |
| "grad_norm": 22.972614288330078, | |
| "learning_rate": 1.5428767123287672e-05, | |
| "loss": 0.6916, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.4657534246575343, | |
| "grad_norm": 14.032968521118164, | |
| "learning_rate": 1.5360273972602738e-05, | |
| "loss": 0.6643, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.4726027397260273, | |
| "grad_norm": 60.66258239746094, | |
| "learning_rate": 1.5291780821917807e-05, | |
| "loss": 0.5659, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 18.83857536315918, | |
| "learning_rate": 1.5223287671232877e-05, | |
| "loss": 0.6599, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.4863013698630136, | |
| "grad_norm": 16.09821128845215, | |
| "learning_rate": 1.5154794520547946e-05, | |
| "loss": 0.6523, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.4931506849315068, | |
| "grad_norm": 8.068375587463379, | |
| "learning_rate": 1.5086301369863012e-05, | |
| "loss": 0.6724, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 16.360702514648438, | |
| "learning_rate": 1.5017808219178082e-05, | |
| "loss": 0.6371, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 17.347824096679688, | |
| "learning_rate": 1.4949315068493151e-05, | |
| "loss": 0.6732, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.5136986301369864, | |
| "grad_norm": 45.4647216796875, | |
| "learning_rate": 1.4880821917808219e-05, | |
| "loss": 0.721, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 1.5205479452054793, | |
| "grad_norm": 53.550174713134766, | |
| "learning_rate": 1.4812328767123289e-05, | |
| "loss": 0.6261, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.5273972602739727, | |
| "grad_norm": 27.888072967529297, | |
| "learning_rate": 1.4743835616438356e-05, | |
| "loss": 0.7, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 23.892024993896484, | |
| "learning_rate": 1.4675342465753426e-05, | |
| "loss": 0.6329, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 1.541095890410959, | |
| "grad_norm": 12.964653968811035, | |
| "learning_rate": 1.4606849315068494e-05, | |
| "loss": 0.6893, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.547945205479452, | |
| "grad_norm": 16.30516242980957, | |
| "learning_rate": 1.4538356164383563e-05, | |
| "loss": 0.6674, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 1.5547945205479452, | |
| "grad_norm": 20.829771041870117, | |
| "learning_rate": 1.4469863013698629e-05, | |
| "loss": 0.7394, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 21.59797477722168, | |
| "learning_rate": 1.4401369863013699e-05, | |
| "loss": 0.7592, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.5684931506849316, | |
| "grad_norm": 14.089178085327148, | |
| "learning_rate": 1.4332876712328766e-05, | |
| "loss": 0.6692, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 1.5753424657534247, | |
| "grad_norm": 9.009936332702637, | |
| "learning_rate": 1.4264383561643836e-05, | |
| "loss": 0.5945, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.5821917808219177, | |
| "grad_norm": 14.968718528747559, | |
| "learning_rate": 1.4195890410958904e-05, | |
| "loss": 0.7055, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 8.840102195739746, | |
| "learning_rate": 1.4127397260273973e-05, | |
| "loss": 0.7345, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 1.595890410958904, | |
| "grad_norm": 17.11764907836914, | |
| "learning_rate": 1.4058904109589041e-05, | |
| "loss": 0.7073, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 1.6027397260273972, | |
| "grad_norm": 11.691437721252441, | |
| "learning_rate": 1.399041095890411e-05, | |
| "loss": 0.6319, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.6095890410958904, | |
| "grad_norm": 12.635778427124023, | |
| "learning_rate": 1.3921917808219178e-05, | |
| "loss": 0.6966, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 8.899778366088867, | |
| "learning_rate": 1.3853424657534248e-05, | |
| "loss": 0.7087, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 1.6232876712328768, | |
| "grad_norm": 20.235586166381836, | |
| "learning_rate": 1.3784931506849315e-05, | |
| "loss": 0.6193, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 1.6301369863013697, | |
| "grad_norm": 36.906707763671875, | |
| "learning_rate": 1.3716438356164385e-05, | |
| "loss": 0.6291, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 1.6369863013698631, | |
| "grad_norm": 11.22154712677002, | |
| "learning_rate": 1.3647945205479453e-05, | |
| "loss": 0.7093, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 27.797801971435547, | |
| "learning_rate": 1.357945205479452e-05, | |
| "loss": 0.6589, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_exact_match": 82.42194891201514, | |
| "eval_f1": 89.45487444483695, | |
| "eval_runtime": 406.0814, | |
| "eval_samples_per_second": 26.029, | |
| "eval_steps_per_second": 1.628, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.6506849315068495, | |
| "grad_norm": 39.78184509277344, | |
| "learning_rate": 1.3510958904109588e-05, | |
| "loss": 0.6151, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 1.6575342465753424, | |
| "grad_norm": 8.190918922424316, | |
| "learning_rate": 1.3442465753424658e-05, | |
| "loss": 0.6776, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 1.6643835616438356, | |
| "grad_norm": 33.2342529296875, | |
| "learning_rate": 1.3373972602739725e-05, | |
| "loss": 0.6841, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 20.676816940307617, | |
| "learning_rate": 1.3305479452054795e-05, | |
| "loss": 0.6696, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 1.678082191780822, | |
| "grad_norm": 28.78380584716797, | |
| "learning_rate": 1.3236986301369863e-05, | |
| "loss": 0.6833, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.6849315068493151, | |
| "grad_norm": 7.916901588439941, | |
| "learning_rate": 1.3168493150684932e-05, | |
| "loss": 0.6699, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.691780821917808, | |
| "grad_norm": 19.101404190063477, | |
| "learning_rate": 1.31e-05, | |
| "loss": 0.698, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 7.5608978271484375, | |
| "learning_rate": 1.303150684931507e-05, | |
| "loss": 0.6808, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 1.7054794520547945, | |
| "grad_norm": 24.476348876953125, | |
| "learning_rate": 1.2963013698630137e-05, | |
| "loss": 0.7041, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 1.7123287671232876, | |
| "grad_norm": 22.60247039794922, | |
| "learning_rate": 1.2894520547945207e-05, | |
| "loss": 0.6406, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.7191780821917808, | |
| "grad_norm": 18.08481216430664, | |
| "learning_rate": 1.2826027397260274e-05, | |
| "loss": 0.6683, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 254.4781951904297, | |
| "learning_rate": 1.2757534246575342e-05, | |
| "loss": 0.5617, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.7328767123287672, | |
| "grad_norm": 24.120647430419922, | |
| "learning_rate": 1.268904109589041e-05, | |
| "loss": 0.6121, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 1.7397260273972601, | |
| "grad_norm": 25.768285751342773, | |
| "learning_rate": 1.262054794520548e-05, | |
| "loss": 0.5745, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 1.7465753424657535, | |
| "grad_norm": 13.516427993774414, | |
| "learning_rate": 1.2552054794520547e-05, | |
| "loss": 0.6538, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 26.595272064208984, | |
| "learning_rate": 1.2483561643835617e-05, | |
| "loss": 0.7562, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 1.7602739726027399, | |
| "grad_norm": 9.390408515930176, | |
| "learning_rate": 1.2415068493150685e-05, | |
| "loss": 0.6179, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 1.7671232876712328, | |
| "grad_norm": 25.727460861206055, | |
| "learning_rate": 1.2346575342465754e-05, | |
| "loss": 0.6364, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.773972602739726, | |
| "grad_norm": 17.040943145751953, | |
| "learning_rate": 1.2278082191780822e-05, | |
| "loss": 0.6312, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 47.64375305175781, | |
| "learning_rate": 1.2209589041095891e-05, | |
| "loss": 0.6916, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7876712328767124, | |
| "grad_norm": 7.669281005859375, | |
| "learning_rate": 1.2141095890410959e-05, | |
| "loss": 0.6765, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 1.7945205479452055, | |
| "grad_norm": 27.307491302490234, | |
| "learning_rate": 1.2072602739726028e-05, | |
| "loss": 0.6479, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 1.8013698630136985, | |
| "grad_norm": 34.39345932006836, | |
| "learning_rate": 1.2004109589041096e-05, | |
| "loss": 0.6739, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 3.3462016582489014, | |
| "learning_rate": 1.1935616438356166e-05, | |
| "loss": 0.6523, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.8150684931506849, | |
| "grad_norm": 6.420201301574707, | |
| "learning_rate": 1.1867123287671232e-05, | |
| "loss": 0.6458, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 1.821917808219178, | |
| "grad_norm": 60.785194396972656, | |
| "learning_rate": 1.1798630136986301e-05, | |
| "loss": 0.621, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 1.8287671232876712, | |
| "grad_norm": 25.992483139038086, | |
| "learning_rate": 1.1730136986301369e-05, | |
| "loss": 0.6654, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 26.358213424682617, | |
| "learning_rate": 1.1661643835616439e-05, | |
| "loss": 0.6936, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 1.8424657534246576, | |
| "grad_norm": 19.42777442932129, | |
| "learning_rate": 1.1593150684931506e-05, | |
| "loss": 0.6549, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 1.8493150684931505, | |
| "grad_norm": 35.19224548339844, | |
| "learning_rate": 1.1524657534246576e-05, | |
| "loss": 0.6872, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.856164383561644, | |
| "grad_norm": 24.116058349609375, | |
| "learning_rate": 1.1456164383561644e-05, | |
| "loss": 0.6744, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 29.181964874267578, | |
| "learning_rate": 1.1387671232876713e-05, | |
| "loss": 0.7206, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 1.8698630136986303, | |
| "grad_norm": 33.76987838745117, | |
| "learning_rate": 1.131917808219178e-05, | |
| "loss": 0.6819, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 1.8767123287671232, | |
| "grad_norm": 17.98587417602539, | |
| "learning_rate": 1.125068493150685e-05, | |
| "loss": 0.6991, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 1.8835616438356164, | |
| "grad_norm": 30.777263641357422, | |
| "learning_rate": 1.1182191780821918e-05, | |
| "loss": 0.6692, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 9.233376502990723, | |
| "learning_rate": 1.1113698630136988e-05, | |
| "loss": 0.6068, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.8972602739726028, | |
| "grad_norm": 6.541473388671875, | |
| "learning_rate": 1.1045205479452055e-05, | |
| "loss": 0.6263, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 1.904109589041096, | |
| "grad_norm": 10.6819486618042, | |
| "learning_rate": 1.0976712328767123e-05, | |
| "loss": 0.5947, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 1.910958904109589, | |
| "grad_norm": 11.901646614074707, | |
| "learning_rate": 1.0908219178082191e-05, | |
| "loss": 0.722, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 11.687748908996582, | |
| "learning_rate": 1.083972602739726e-05, | |
| "loss": 0.6237, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "eval_exact_match": 82.82876064333018, | |
| "eval_f1": 89.81835180261338, | |
| "eval_runtime": 406.8019, | |
| "eval_samples_per_second": 25.983, | |
| "eval_steps_per_second": 1.625, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9246575342465753, | |
| "grad_norm": 22.835041046142578, | |
| "learning_rate": 1.0771232876712328e-05, | |
| "loss": 0.657, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 1.9315068493150684, | |
| "grad_norm": 15.053028106689453, | |
| "learning_rate": 1.0702739726027398e-05, | |
| "loss": 0.7027, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.9383561643835616, | |
| "grad_norm": 18.73754119873047, | |
| "learning_rate": 1.0634246575342465e-05, | |
| "loss": 0.6092, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 22.99529266357422, | |
| "learning_rate": 1.0565753424657535e-05, | |
| "loss": 0.6337, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 1.952054794520548, | |
| "grad_norm": 11.403436660766602, | |
| "learning_rate": 1.0497260273972603e-05, | |
| "loss": 0.7014, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 1.958904109589041, | |
| "grad_norm": 10.497685432434082, | |
| "learning_rate": 1.0428767123287672e-05, | |
| "loss": 0.624, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 1.9657534246575343, | |
| "grad_norm": 39.24784851074219, | |
| "learning_rate": 1.036027397260274e-05, | |
| "loss": 0.7035, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 10.58161449432373, | |
| "learning_rate": 1.029178082191781e-05, | |
| "loss": 0.6705, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.9794520547945207, | |
| "grad_norm": 30.54125213623047, | |
| "learning_rate": 1.0223287671232877e-05, | |
| "loss": 0.6178, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 1.9863013698630136, | |
| "grad_norm": 11.545398712158203, | |
| "learning_rate": 1.0154794520547947e-05, | |
| "loss": 0.6929, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.9931506849315068, | |
| "grad_norm": 12.356890678405762, | |
| "learning_rate": 1.0086301369863013e-05, | |
| "loss": 0.6686, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 12.808874130249023, | |
| "learning_rate": 1.0017808219178082e-05, | |
| "loss": 0.709, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.006849315068493, | |
| "grad_norm": 20.87654685974121, | |
| "learning_rate": 9.94931506849315e-06, | |
| "loss": 0.3807, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.0136986301369864, | |
| "grad_norm": 11.455994606018066, | |
| "learning_rate": 9.88082191780822e-06, | |
| "loss": 0.4346, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.0205479452054793, | |
| "grad_norm": 9.450268745422363, | |
| "learning_rate": 9.812328767123287e-06, | |
| "loss": 0.4046, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 11.238616943359375, | |
| "learning_rate": 9.743835616438357e-06, | |
| "loss": 0.4077, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.0342465753424657, | |
| "grad_norm": 7.271957874298096, | |
| "learning_rate": 9.675342465753424e-06, | |
| "loss": 0.4039, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.041095890410959, | |
| "grad_norm": 13.240756034851074, | |
| "learning_rate": 9.606849315068494e-06, | |
| "loss": 0.4307, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.047945205479452, | |
| "grad_norm": 17.84387969970703, | |
| "learning_rate": 9.538356164383562e-06, | |
| "loss": 0.4196, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 15.702322959899902, | |
| "learning_rate": 9.469863013698631e-06, | |
| "loss": 0.39, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.0616438356164384, | |
| "grad_norm": 17.96023178100586, | |
| "learning_rate": 9.401369863013699e-06, | |
| "loss": 0.3965, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.0684931506849313, | |
| "grad_norm": 29.622323989868164, | |
| "learning_rate": 9.332876712328768e-06, | |
| "loss": 0.4599, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.0753424657534247, | |
| "grad_norm": 10.847167015075684, | |
| "learning_rate": 9.264383561643836e-06, | |
| "loss": 0.3994, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 9.122156143188477, | |
| "learning_rate": 9.195890410958904e-06, | |
| "loss": 0.3829, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.089041095890411, | |
| "grad_norm": 61.343101501464844, | |
| "learning_rate": 9.127397260273972e-06, | |
| "loss": 0.4434, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.095890410958904, | |
| "grad_norm": 14.082651138305664, | |
| "learning_rate": 9.058904109589041e-06, | |
| "loss": 0.377, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.1027397260273974, | |
| "grad_norm": 10.202653884887695, | |
| "learning_rate": 8.990410958904109e-06, | |
| "loss": 0.3685, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 18.169658660888672, | |
| "learning_rate": 8.921917808219179e-06, | |
| "loss": 0.4122, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.1164383561643834, | |
| "grad_norm": 54.33354568481445, | |
| "learning_rate": 8.853424657534246e-06, | |
| "loss": 0.3906, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.1232876712328768, | |
| "grad_norm": 6.232911109924316, | |
| "learning_rate": 8.784931506849316e-06, | |
| "loss": 0.3766, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.1301369863013697, | |
| "grad_norm": 40.90781784057617, | |
| "learning_rate": 8.716438356164384e-06, | |
| "loss": 0.4178, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 43.94190979003906, | |
| "learning_rate": 8.647945205479453e-06, | |
| "loss": 0.4115, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.143835616438356, | |
| "grad_norm": 14.004490852355957, | |
| "learning_rate": 8.57945205479452e-06, | |
| "loss": 0.4495, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.1506849315068495, | |
| "grad_norm": 10.205181121826172, | |
| "learning_rate": 8.51095890410959e-06, | |
| "loss": 0.3702, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.1575342465753424, | |
| "grad_norm": 12.333789825439453, | |
| "learning_rate": 8.442465753424658e-06, | |
| "loss": 0.4153, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 10.010555267333984, | |
| "learning_rate": 8.373972602739728e-06, | |
| "loss": 0.3854, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.171232876712329, | |
| "grad_norm": 25.08806037902832, | |
| "learning_rate": 8.305479452054794e-06, | |
| "loss": 0.3935, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.1780821917808217, | |
| "grad_norm": 5.474767684936523, | |
| "learning_rate": 8.236986301369863e-06, | |
| "loss": 0.3408, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.184931506849315, | |
| "grad_norm": 19.28006362915039, | |
| "learning_rate": 8.168493150684931e-06, | |
| "loss": 0.3059, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 1.0139840841293335, | |
| "learning_rate": 8.1e-06, | |
| "loss": 0.4515, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 82.21381267738883, | |
| "eval_f1": 89.63473536642901, | |
| "eval_runtime": 406.5079, | |
| "eval_samples_per_second": 26.002, | |
| "eval_steps_per_second": 1.626, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.1986301369863015, | |
| "grad_norm": 26.130165100097656, | |
| "learning_rate": 8.031506849315068e-06, | |
| "loss": 0.3657, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.2054794520547945, | |
| "grad_norm": 12.132906913757324, | |
| "learning_rate": 7.963013698630138e-06, | |
| "loss": 0.3715, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.212328767123288, | |
| "grad_norm": 19.634357452392578, | |
| "learning_rate": 7.894520547945205e-06, | |
| "loss": 0.3868, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 12.86025333404541, | |
| "learning_rate": 7.826027397260275e-06, | |
| "loss": 0.4119, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.2260273972602738, | |
| "grad_norm": 13.171170234680176, | |
| "learning_rate": 7.757534246575343e-06, | |
| "loss": 0.4362, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.232876712328767, | |
| "grad_norm": 32.56090545654297, | |
| "learning_rate": 7.689041095890412e-06, | |
| "loss": 0.3995, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.23972602739726, | |
| "grad_norm": 31.1318416595459, | |
| "learning_rate": 7.620547945205479e-06, | |
| "loss": 0.4267, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 8.52885913848877, | |
| "learning_rate": 7.5520547945205485e-06, | |
| "loss": 0.4065, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 2.2534246575342465, | |
| "grad_norm": 5.443692684173584, | |
| "learning_rate": 7.483561643835616e-06, | |
| "loss": 0.4486, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 2.26027397260274, | |
| "grad_norm": 25.305814743041992, | |
| "learning_rate": 7.415068493150685e-06, | |
| "loss": 0.4277, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.267123287671233, | |
| "grad_norm": 32.161224365234375, | |
| "learning_rate": 7.3465753424657536e-06, | |
| "loss": 0.4175, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 12.804214477539062, | |
| "learning_rate": 7.278082191780822e-06, | |
| "loss": 0.3335, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 2.280821917808219, | |
| "grad_norm": 22.67701530456543, | |
| "learning_rate": 7.20958904109589e-06, | |
| "loss": 0.4323, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 2.287671232876712, | |
| "grad_norm": 31.13144302368164, | |
| "learning_rate": 7.142465753424657e-06, | |
| "loss": 0.4542, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 2.2945205479452055, | |
| "grad_norm": 20.63867950439453, | |
| "learning_rate": 7.073972602739726e-06, | |
| "loss": 0.4023, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 21.299816131591797, | |
| "learning_rate": 7.005479452054794e-06, | |
| "loss": 0.3852, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 2.308219178082192, | |
| "grad_norm": 27.046512603759766, | |
| "learning_rate": 6.936986301369863e-06, | |
| "loss": 0.4234, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 2.315068493150685, | |
| "grad_norm": 23.793231964111328, | |
| "learning_rate": 6.8684931506849315e-06, | |
| "loss": 0.4347, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 2.3219178082191783, | |
| "grad_norm": 19.59113883972168, | |
| "learning_rate": 6.8e-06, | |
| "loss": 0.4505, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 14.70582389831543, | |
| "learning_rate": 6.731506849315069e-06, | |
| "loss": 0.36, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.3356164383561646, | |
| "grad_norm": 49.76693344116211, | |
| "learning_rate": 6.6630136986301365e-06, | |
| "loss": 0.3639, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 2.3424657534246576, | |
| "grad_norm": 44.807167053222656, | |
| "learning_rate": 6.594520547945205e-06, | |
| "loss": 0.3981, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 2.3493150684931505, | |
| "grad_norm": 9.841875076293945, | |
| "learning_rate": 6.526027397260274e-06, | |
| "loss": 0.3733, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 50.942108154296875, | |
| "learning_rate": 6.457534246575342e-06, | |
| "loss": 0.4233, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 2.363013698630137, | |
| "grad_norm": 9.675677299499512, | |
| "learning_rate": 6.389041095890411e-06, | |
| "loss": 0.4035, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.3698630136986303, | |
| "grad_norm": 5.541302680969238, | |
| "learning_rate": 6.32054794520548e-06, | |
| "loss": 0.4161, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 2.3767123287671232, | |
| "grad_norm": 8.581879615783691, | |
| "learning_rate": 6.2520547945205474e-06, | |
| "loss": 0.3986, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 9.531363487243652, | |
| "learning_rate": 6.183561643835616e-06, | |
| "loss": 0.3827, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.3904109589041096, | |
| "grad_norm": 7.351466178894043, | |
| "learning_rate": 6.115068493150685e-06, | |
| "loss": 0.3796, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 2.3972602739726026, | |
| "grad_norm": 8.607413291931152, | |
| "learning_rate": 6.046575342465753e-06, | |
| "loss": 0.3983, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.404109589041096, | |
| "grad_norm": 8.33619499206543, | |
| "learning_rate": 5.978082191780822e-06, | |
| "loss": 0.3695, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 6.319853782653809, | |
| "learning_rate": 5.9095890410958906e-06, | |
| "loss": 0.3843, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 2.4178082191780823, | |
| "grad_norm": 19.26327133178711, | |
| "learning_rate": 5.841095890410958e-06, | |
| "loss": 0.4723, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 2.4246575342465753, | |
| "grad_norm": 16.118913650512695, | |
| "learning_rate": 5.772602739726027e-06, | |
| "loss": 0.4518, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.4315068493150687, | |
| "grad_norm": 7.736336708068848, | |
| "learning_rate": 5.704109589041096e-06, | |
| "loss": 0.3713, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 11.381460189819336, | |
| "learning_rate": 5.635616438356164e-06, | |
| "loss": 0.4285, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 2.4452054794520546, | |
| "grad_norm": 401.49493408203125, | |
| "learning_rate": 5.567123287671233e-06, | |
| "loss": 0.3637, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 2.452054794520548, | |
| "grad_norm": 20.71985626220703, | |
| "learning_rate": 5.4986301369863015e-06, | |
| "loss": 0.4036, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 2.458904109589041, | |
| "grad_norm": 14.313848495483398, | |
| "learning_rate": 5.43013698630137e-06, | |
| "loss": 0.4131, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 14.69888687133789, | |
| "learning_rate": 5.361643835616438e-06, | |
| "loss": 0.4162, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "eval_exact_match": 82.2705771050142, | |
| "eval_f1": 89.66159681312358, | |
| "eval_runtime": 407.322, | |
| "eval_samples_per_second": 25.95, | |
| "eval_steps_per_second": 1.623, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4726027397260273, | |
| "grad_norm": 15.124028205871582, | |
| "learning_rate": 5.2931506849315065e-06, | |
| "loss": 0.4069, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 2.4794520547945207, | |
| "grad_norm": 32.08819580078125, | |
| "learning_rate": 5.224657534246575e-06, | |
| "loss": 0.3963, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 2.4863013698630136, | |
| "grad_norm": 10.072858810424805, | |
| "learning_rate": 5.156164383561644e-06, | |
| "loss": 0.3669, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 7.13778018951416, | |
| "learning_rate": 5.087671232876712e-06, | |
| "loss": 0.3498, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 22.02096176147461, | |
| "learning_rate": 5.019178082191781e-06, | |
| "loss": 0.4829, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.506849315068493, | |
| "grad_norm": 11.320915222167969, | |
| "learning_rate": 4.950684931506849e-06, | |
| "loss": 0.3595, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.5136986301369864, | |
| "grad_norm": 12.33502197265625, | |
| "learning_rate": 4.8821917808219174e-06, | |
| "loss": 0.3876, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 13.46191692352295, | |
| "learning_rate": 4.813698630136986e-06, | |
| "loss": 0.454, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 2.5273972602739727, | |
| "grad_norm": 13.49188232421875, | |
| "learning_rate": 4.746575342465753e-06, | |
| "loss": 0.3855, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 2.5342465753424657, | |
| "grad_norm": 17.15059471130371, | |
| "learning_rate": 4.678082191780822e-06, | |
| "loss": 0.366, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.541095890410959, | |
| "grad_norm": 8.814437866210938, | |
| "learning_rate": 4.60958904109589e-06, | |
| "loss": 0.4034, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 22.521684646606445, | |
| "learning_rate": 4.541095890410959e-06, | |
| "loss": 0.4877, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.5547945205479454, | |
| "grad_norm": 12.521992683410645, | |
| "learning_rate": 4.4726027397260276e-06, | |
| "loss": 0.39, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 2.5616438356164384, | |
| "grad_norm": 8.952247619628906, | |
| "learning_rate": 4.404109589041095e-06, | |
| "loss": 0.3797, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 2.5684931506849313, | |
| "grad_norm": 22.96004867553711, | |
| "learning_rate": 4.335616438356164e-06, | |
| "loss": 0.3642, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 15.552525520324707, | |
| "learning_rate": 4.267123287671233e-06, | |
| "loss": 0.3611, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 2.5821917808219177, | |
| "grad_norm": 6.638945579528809, | |
| "learning_rate": 4.198630136986301e-06, | |
| "loss": 0.3153, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 2.589041095890411, | |
| "grad_norm": 22.522302627563477, | |
| "learning_rate": 4.13013698630137e-06, | |
| "loss": 0.4371, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.595890410958904, | |
| "grad_norm": 9.711662292480469, | |
| "learning_rate": 4.0616438356164385e-06, | |
| "loss": 0.3819, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 13.974068641662598, | |
| "learning_rate": 3.993150684931506e-06, | |
| "loss": 0.3498, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.6095890410958904, | |
| "grad_norm": 16.810501098632812, | |
| "learning_rate": 3.924657534246575e-06, | |
| "loss": 0.3499, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 2.616438356164384, | |
| "grad_norm": 12.136087417602539, | |
| "learning_rate": 3.8561643835616435e-06, | |
| "loss": 0.3913, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 2.6232876712328768, | |
| "grad_norm": 20.14099884033203, | |
| "learning_rate": 3.787671232876712e-06, | |
| "loss": 0.3576, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 3.969587802886963, | |
| "learning_rate": 3.719178082191781e-06, | |
| "loss": 0.4321, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.636986301369863, | |
| "grad_norm": 2.799744129180908, | |
| "learning_rate": 3.6506849315068494e-06, | |
| "loss": 0.3585, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.643835616438356, | |
| "grad_norm": 48.521202087402344, | |
| "learning_rate": 3.582191780821918e-06, | |
| "loss": 0.406, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 2.6506849315068495, | |
| "grad_norm": 15.20323371887207, | |
| "learning_rate": 3.5136986301369866e-06, | |
| "loss": 0.3498, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 14.380341529846191, | |
| "learning_rate": 3.445205479452055e-06, | |
| "loss": 0.3952, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 2.6643835616438354, | |
| "grad_norm": 21.444190979003906, | |
| "learning_rate": 3.3767123287671235e-06, | |
| "loss": 0.4067, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 2.671232876712329, | |
| "grad_norm": 12.899758338928223, | |
| "learning_rate": 3.308219178082192e-06, | |
| "loss": 0.4587, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.678082191780822, | |
| "grad_norm": 9.967342376708984, | |
| "learning_rate": 3.2397260273972603e-06, | |
| "loss": 0.3656, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 5.67423677444458, | |
| "learning_rate": 3.171232876712329e-06, | |
| "loss": 0.4351, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 2.691780821917808, | |
| "grad_norm": 14.479686737060547, | |
| "learning_rate": 3.1027397260273976e-06, | |
| "loss": 0.3801, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 2.6986301369863015, | |
| "grad_norm": 17.046621322631836, | |
| "learning_rate": 3.034246575342466e-06, | |
| "loss": 0.4302, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 2.7054794520547945, | |
| "grad_norm": 10.318126678466797, | |
| "learning_rate": 2.9657534246575344e-06, | |
| "loss": 0.3594, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 24.85907554626465, | |
| "learning_rate": 2.897260273972603e-06, | |
| "loss": 0.3195, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.719178082191781, | |
| "grad_norm": 16.129796981811523, | |
| "learning_rate": 2.8287671232876716e-06, | |
| "loss": 0.3899, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 2.7260273972602738, | |
| "grad_norm": 13.561001777648926, | |
| "learning_rate": 2.76027397260274e-06, | |
| "loss": 0.3348, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 2.732876712328767, | |
| "grad_norm": 32.10982131958008, | |
| "learning_rate": 2.6917808219178085e-06, | |
| "loss": 0.3648, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 49.121280670166016, | |
| "learning_rate": 2.623287671232877e-06, | |
| "loss": 0.4146, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "eval_exact_match": 82.77199621570483, | |
| "eval_f1": 89.87472419974556, | |
| "eval_runtime": 407.372, | |
| "eval_samples_per_second": 25.947, | |
| "eval_steps_per_second": 1.623, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.7465753424657535, | |
| "grad_norm": 10.713053703308105, | |
| "learning_rate": 2.5547945205479453e-06, | |
| "loss": 0.3653, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 2.7534246575342465, | |
| "grad_norm": 20.47135353088379, | |
| "learning_rate": 2.486301369863014e-06, | |
| "loss": 0.4231, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.76027397260274, | |
| "grad_norm": 40.13731002807617, | |
| "learning_rate": 2.4178082191780826e-06, | |
| "loss": 0.3993, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 12.861891746520996, | |
| "learning_rate": 2.3493150684931508e-06, | |
| "loss": 0.3912, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 2.7739726027397262, | |
| "grad_norm": 7.921535968780518, | |
| "learning_rate": 2.2808219178082194e-06, | |
| "loss": 0.408, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 2.780821917808219, | |
| "grad_norm": 26.30048179626465, | |
| "learning_rate": 2.2123287671232876e-06, | |
| "loss": 0.4308, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 2.787671232876712, | |
| "grad_norm": 22.600740432739258, | |
| "learning_rate": 2.1438356164383562e-06, | |
| "loss": 0.363, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 10.7540283203125, | |
| "learning_rate": 2.0753424657534244e-06, | |
| "loss": 0.3956, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.8013698630136985, | |
| "grad_norm": 13.891115188598633, | |
| "learning_rate": 2.006849315068493e-06, | |
| "loss": 0.3441, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 2.808219178082192, | |
| "grad_norm": 14.40695571899414, | |
| "learning_rate": 1.9383561643835617e-06, | |
| "loss": 0.3875, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.815068493150685, | |
| "grad_norm": 19.043682098388672, | |
| "learning_rate": 1.8698630136986303e-06, | |
| "loss": 0.4015, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 26.111764907836914, | |
| "learning_rate": 1.8013698630136987e-06, | |
| "loss": 0.3877, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 2.828767123287671, | |
| "grad_norm": 13.53073787689209, | |
| "learning_rate": 1.7328767123287671e-06, | |
| "loss": 0.3292, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 2.8356164383561646, | |
| "grad_norm": 12.833224296569824, | |
| "learning_rate": 1.6643835616438358e-06, | |
| "loss": 0.3202, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.8424657534246576, | |
| "grad_norm": 12.937023162841797, | |
| "learning_rate": 1.5958904109589042e-06, | |
| "loss": 0.3709, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 30.91938018798828, | |
| "learning_rate": 1.5273972602739726e-06, | |
| "loss": 0.4258, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 2.856164383561644, | |
| "grad_norm": 50.359283447265625, | |
| "learning_rate": 1.4589041095890412e-06, | |
| "loss": 0.3572, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 2.863013698630137, | |
| "grad_norm": 7.755626201629639, | |
| "learning_rate": 1.3904109589041096e-06, | |
| "loss": 0.3985, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 2.8698630136986303, | |
| "grad_norm": 33.82756805419922, | |
| "learning_rate": 1.3219178082191783e-06, | |
| "loss": 0.3494, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 44.668338775634766, | |
| "learning_rate": 1.2534246575342467e-06, | |
| "loss": 0.4596, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.883561643835616, | |
| "grad_norm": 24.665861129760742, | |
| "learning_rate": 1.184931506849315e-06, | |
| "loss": 0.3069, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 2.8904109589041096, | |
| "grad_norm": 13.378949165344238, | |
| "learning_rate": 1.1164383561643837e-06, | |
| "loss": 0.3873, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 2.897260273972603, | |
| "grad_norm": 8.084388732910156, | |
| "learning_rate": 1.0479452054794521e-06, | |
| "loss": 0.3911, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 8.717424392700195, | |
| "learning_rate": 9.794520547945205e-07, | |
| "loss": 0.4187, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 2.910958904109589, | |
| "grad_norm": 10.450674057006836, | |
| "learning_rate": 9.123287671232876e-07, | |
| "loss": 0.3498, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 2.9178082191780823, | |
| "grad_norm": 8.336039543151855, | |
| "learning_rate": 8.438356164383562e-07, | |
| "loss": 0.3815, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.9246575342465753, | |
| "grad_norm": 25.949018478393555, | |
| "learning_rate": 7.753424657534247e-07, | |
| "loss": 0.3061, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 3.685792922973633, | |
| "learning_rate": 7.068493150684931e-07, | |
| "loss": 0.4067, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 2.9383561643835616, | |
| "grad_norm": 20.176740646362305, | |
| "learning_rate": 6.383561643835616e-07, | |
| "loss": 0.4145, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 2.9452054794520546, | |
| "grad_norm": 6.9156270027160645, | |
| "learning_rate": 5.698630136986301e-07, | |
| "loss": 0.3644, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.952054794520548, | |
| "grad_norm": 16.190839767456055, | |
| "learning_rate": 5.013698630136987e-07, | |
| "loss": 0.3783, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 14.747089385986328, | |
| "learning_rate": 4.3287671232876714e-07, | |
| "loss": 0.4057, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.9657534246575343, | |
| "grad_norm": 28.770313262939453, | |
| "learning_rate": 3.643835616438356e-07, | |
| "loss": 0.3261, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 2.9726027397260273, | |
| "grad_norm": 6.209177017211914, | |
| "learning_rate": 2.958904109589041e-07, | |
| "loss": 0.4036, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 2.9794520547945207, | |
| "grad_norm": 28.698171615600586, | |
| "learning_rate": 2.273972602739726e-07, | |
| "loss": 0.3678, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 7.778885364532471, | |
| "learning_rate": 1.589041095890411e-07, | |
| "loss": 0.3814, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 2.993150684931507, | |
| "grad_norm": 28.04308319091797, | |
| "learning_rate": 9.04109589041096e-08, | |
| "loss": 0.3538, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.1987411975860596, | |
| "learning_rate": 2.1917808219178083e-08, | |
| "loss": 0.3481, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 21900, | |
| "total_flos": 2.006738209660207e+18, | |
| "train_loss": 0.7586759792397556, | |
| "train_runtime": 82472.0176, | |
| "train_samples_per_second": 3.186, | |
| "train_steps_per_second": 0.266 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 21900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.006738209660207e+18, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |