| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 1000, | |
| "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": 323.0462646484375, | |
| "learning_rate": 2.9943835616438356e-05, | |
| "loss": 5.0367, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0136986301369863, | |
| "grad_norm": 77.02570343017578, | |
| "learning_rate": 2.987808219178082e-05, | |
| "loss": 3.1287, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02054794520547945, | |
| "grad_norm": 33.8292236328125, | |
| "learning_rate": 2.980958904109589e-05, | |
| "loss": 2.247, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 45.22611618041992, | |
| "learning_rate": 2.974109589041096e-05, | |
| "loss": 1.8313, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03424657534246575, | |
| "grad_norm": 40.95551681518555, | |
| "learning_rate": 2.9672602739726026e-05, | |
| "loss": 1.7307, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0410958904109589, | |
| "grad_norm": 132.5420684814453, | |
| "learning_rate": 2.9604109589041095e-05, | |
| "loss": 1.71, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04794520547945205, | |
| "grad_norm": 46.58742141723633, | |
| "learning_rate": 2.9535616438356165e-05, | |
| "loss": 1.5229, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 60.270259857177734, | |
| "learning_rate": 2.9467123287671234e-05, | |
| "loss": 1.4451, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06164383561643835, | |
| "grad_norm": 36.05481719970703, | |
| "learning_rate": 2.93986301369863e-05, | |
| "loss": 1.4609, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0684931506849315, | |
| "grad_norm": 20.183435440063477, | |
| "learning_rate": 2.933013698630137e-05, | |
| "loss": 1.5096, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.07534246575342465, | |
| "grad_norm": 31.55844497680664, | |
| "learning_rate": 2.926164383561644e-05, | |
| "loss": 1.4921, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 26.991056442260742, | |
| "learning_rate": 2.919315068493151e-05, | |
| "loss": 1.3397, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08904109589041095, | |
| "grad_norm": 22.860877990722656, | |
| "learning_rate": 2.9124657534246575e-05, | |
| "loss": 1.2389, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.0958904109589041, | |
| "grad_norm": 34.633026123046875, | |
| "learning_rate": 2.9056164383561644e-05, | |
| "loss": 1.182, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.10273972602739725, | |
| "grad_norm": 29.118816375732422, | |
| "learning_rate": 2.8987671232876714e-05, | |
| "loss": 1.3518, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 72.57406616210938, | |
| "learning_rate": 2.8919178082191783e-05, | |
| "loss": 1.3301, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.11643835616438356, | |
| "grad_norm": 19.362930297851562, | |
| "learning_rate": 2.885068493150685e-05, | |
| "loss": 1.2366, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1232876712328767, | |
| "grad_norm": 34.462242126464844, | |
| "learning_rate": 2.878219178082192e-05, | |
| "loss": 1.3056, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.13013698630136986, | |
| "grad_norm": 20.275766372680664, | |
| "learning_rate": 2.871369863013699e-05, | |
| "loss": 1.2415, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 34.678558349609375, | |
| "learning_rate": 2.8645205479452058e-05, | |
| "loss": 1.2008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "eval_exact_match": 74.98580889309366, | |
| "eval_f1": 84.12743563305797, | |
| "eval_runtime": 416.5856, | |
| "eval_samples_per_second": 25.373, | |
| "eval_steps_per_second": 1.587, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14383561643835616, | |
| "grad_norm": 27.335071563720703, | |
| "learning_rate": 2.8576712328767124e-05, | |
| "loss": 1.1301, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1506849315068493, | |
| "grad_norm": 36.93644714355469, | |
| "learning_rate": 2.8508219178082194e-05, | |
| "loss": 1.3069, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.15753424657534246, | |
| "grad_norm": 53.08256149291992, | |
| "learning_rate": 2.8439726027397263e-05, | |
| "loss": 1.203, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 70.76780700683594, | |
| "learning_rate": 2.8371232876712332e-05, | |
| "loss": 1.2392, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.17123287671232876, | |
| "grad_norm": 26.65420150756836, | |
| "learning_rate": 2.8302739726027395e-05, | |
| "loss": 1.1829, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1780821917808219, | |
| "grad_norm": 23.404756546020508, | |
| "learning_rate": 2.8234246575342465e-05, | |
| "loss": 1.2175, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.18493150684931506, | |
| "grad_norm": 23.29783058166504, | |
| "learning_rate": 2.8165753424657534e-05, | |
| "loss": 1.1673, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 45.53365707397461, | |
| "learning_rate": 2.8097260273972604e-05, | |
| "loss": 1.1966, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.19863013698630136, | |
| "grad_norm": 13.774870872497559, | |
| "learning_rate": 2.802876712328767e-05, | |
| "loss": 1.2635, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.2054794520547945, | |
| "grad_norm": 14.898510932922363, | |
| "learning_rate": 2.796027397260274e-05, | |
| "loss": 1.1165, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21232876712328766, | |
| "grad_norm": 35.47676467895508, | |
| "learning_rate": 2.789178082191781e-05, | |
| "loss": 1.1877, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 28.7779598236084, | |
| "learning_rate": 2.7823287671232878e-05, | |
| "loss": 1.1162, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.22602739726027396, | |
| "grad_norm": 21.763216018676758, | |
| "learning_rate": 2.7754794520547944e-05, | |
| "loss": 1.2048, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.2328767123287671, | |
| "grad_norm": 32.003700256347656, | |
| "learning_rate": 2.7686301369863014e-05, | |
| "loss": 1.1233, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23972602739726026, | |
| "grad_norm": 31.449691772460938, | |
| "learning_rate": 2.7617808219178083e-05, | |
| "loss": 1.0509, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 18.793638229370117, | |
| "learning_rate": 2.7549315068493153e-05, | |
| "loss": 1.0995, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.2534246575342466, | |
| "grad_norm": 12.931886672973633, | |
| "learning_rate": 2.748082191780822e-05, | |
| "loss": 1.037, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.2602739726027397, | |
| "grad_norm": 14.888827323913574, | |
| "learning_rate": 2.7412328767123288e-05, | |
| "loss": 1.0856, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2671232876712329, | |
| "grad_norm": 42.177330017089844, | |
| "learning_rate": 2.7343835616438358e-05, | |
| "loss": 1.0143, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 52.39680480957031, | |
| "learning_rate": 2.7275342465753427e-05, | |
| "loss": 0.9292, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "eval_exact_match": 78.05108798486282, | |
| "eval_f1": 86.8708882051522, | |
| "eval_runtime": 417.914, | |
| "eval_samples_per_second": 25.292, | |
| "eval_steps_per_second": 1.582, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2808219178082192, | |
| "grad_norm": 79.39705657958984, | |
| "learning_rate": 2.7206849315068493e-05, | |
| "loss": 1.1627, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2876712328767123, | |
| "grad_norm": 13.503669738769531, | |
| "learning_rate": 2.7138356164383563e-05, | |
| "loss": 0.9592, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2945205479452055, | |
| "grad_norm": 34.6471061706543, | |
| "learning_rate": 2.7069863013698632e-05, | |
| "loss": 0.9724, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 28.719350814819336, | |
| "learning_rate": 2.70013698630137e-05, | |
| "loss": 0.9852, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3082191780821918, | |
| "grad_norm": 55.4783935546875, | |
| "learning_rate": 2.6932876712328768e-05, | |
| "loss": 0.9469, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3150684931506849, | |
| "grad_norm": 24.332387924194336, | |
| "learning_rate": 2.6864383561643837e-05, | |
| "loss": 1.0859, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3219178082191781, | |
| "grad_norm": 16.438060760498047, | |
| "learning_rate": 2.6795890410958907e-05, | |
| "loss": 0.9121, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 19.229644775390625, | |
| "learning_rate": 2.6727397260273976e-05, | |
| "loss": 1.0109, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3356164383561644, | |
| "grad_norm": 22.6262149810791, | |
| "learning_rate": 2.6658904109589042e-05, | |
| "loss": 1.077, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 10.421500205993652, | |
| "learning_rate": 2.659041095890411e-05, | |
| "loss": 1.1242, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3493150684931507, | |
| "grad_norm": 44.75662612915039, | |
| "learning_rate": 2.6521917808219178e-05, | |
| "loss": 1.0445, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 10.374776840209961, | |
| "learning_rate": 2.6453424657534247e-05, | |
| "loss": 0.9927, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.363013698630137, | |
| "grad_norm": 21.781341552734375, | |
| "learning_rate": 2.6384931506849313e-05, | |
| "loss": 0.9601, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.3698630136986301, | |
| "grad_norm": 7.096819877624512, | |
| "learning_rate": 2.6316438356164383e-05, | |
| "loss": 1.0105, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3767123287671233, | |
| "grad_norm": 128.46046447753906, | |
| "learning_rate": 2.6247945205479452e-05, | |
| "loss": 0.9627, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 43.575111389160156, | |
| "learning_rate": 2.6179452054794522e-05, | |
| "loss": 1.0495, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.3904109589041096, | |
| "grad_norm": 8.99010944366455, | |
| "learning_rate": 2.6110958904109588e-05, | |
| "loss": 0.9465, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3972602739726027, | |
| "grad_norm": 21.84307289123535, | |
| "learning_rate": 2.6042465753424657e-05, | |
| "loss": 0.9971, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4041095890410959, | |
| "grad_norm": 42.8900032043457, | |
| "learning_rate": 2.5973972602739727e-05, | |
| "loss": 0.907, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 50.012393951416016, | |
| "learning_rate": 2.5905479452054796e-05, | |
| "loss": 0.9572, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "eval_exact_match": 80.07568590350047, | |
| "eval_f1": 88.2866062038965, | |
| "eval_runtime": 417.1017, | |
| "eval_samples_per_second": 25.342, | |
| "eval_steps_per_second": 1.585, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4178082191780822, | |
| "grad_norm": 33.76340103149414, | |
| "learning_rate": 2.5836986301369862e-05, | |
| "loss": 0.9699, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.4246575342465753, | |
| "grad_norm": 14.410927772521973, | |
| "learning_rate": 2.5768493150684932e-05, | |
| "loss": 0.8769, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4315068493150685, | |
| "grad_norm": 13.709430694580078, | |
| "learning_rate": 2.57e-05, | |
| "loss": 0.9086, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 30.455875396728516, | |
| "learning_rate": 2.563150684931507e-05, | |
| "loss": 0.9569, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4452054794520548, | |
| "grad_norm": 10.021869659423828, | |
| "learning_rate": 2.5563013698630137e-05, | |
| "loss": 0.9403, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.4520547945205479, | |
| "grad_norm": 16.0334415435791, | |
| "learning_rate": 2.5494520547945206e-05, | |
| "loss": 0.9413, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4589041095890411, | |
| "grad_norm": 17.812877655029297, | |
| "learning_rate": 2.5426027397260276e-05, | |
| "loss": 0.9739, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 13.129268646240234, | |
| "learning_rate": 2.5357534246575345e-05, | |
| "loss": 0.9456, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4726027397260274, | |
| "grad_norm": 37.476993560791016, | |
| "learning_rate": 2.528904109589041e-05, | |
| "loss": 0.9359, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.4794520547945205, | |
| "grad_norm": 16.441556930541992, | |
| "learning_rate": 2.522054794520548e-05, | |
| "loss": 0.9096, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4863013698630137, | |
| "grad_norm": 11.985368728637695, | |
| "learning_rate": 2.515342465753425e-05, | |
| "loss": 0.9038, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 28.11038589477539, | |
| "learning_rate": 2.5084931506849316e-05, | |
| "loss": 0.8699, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 26.2198543548584, | |
| "learning_rate": 2.5016438356164385e-05, | |
| "loss": 0.9071, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 21.6419620513916, | |
| "learning_rate": 2.4947945205479455e-05, | |
| "loss": 1.1274, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5136986301369864, | |
| "grad_norm": 27.075368881225586, | |
| "learning_rate": 2.4879452054794524e-05, | |
| "loss": 0.9475, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 25.394577026367188, | |
| "learning_rate": 2.481095890410959e-05, | |
| "loss": 0.9166, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5273972602739726, | |
| "grad_norm": 32.769256591796875, | |
| "learning_rate": 2.474246575342466e-05, | |
| "loss": 0.9656, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5342465753424658, | |
| "grad_norm": 9.152816772460938, | |
| "learning_rate": 2.4673972602739726e-05, | |
| "loss": 0.9538, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.541095890410959, | |
| "grad_norm": 24.151294708251953, | |
| "learning_rate": 2.4605479452054795e-05, | |
| "loss": 0.9396, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 10.290411949157715, | |
| "learning_rate": 2.453698630136986e-05, | |
| "loss": 0.971, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_exact_match": 82.82876064333018, | |
| "eval_f1": 90.23705695343135, | |
| "eval_runtime": 418.5094, | |
| "eval_samples_per_second": 25.256, | |
| "eval_steps_per_second": 1.579, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5547945205479452, | |
| "grad_norm": 30.38792610168457, | |
| "learning_rate": 2.446849315068493e-05, | |
| "loss": 0.9951, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.5616438356164384, | |
| "grad_norm": 12.29451847076416, | |
| "learning_rate": 2.44e-05, | |
| "loss": 0.9758, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5684931506849316, | |
| "grad_norm": 56.75043869018555, | |
| "learning_rate": 2.433150684931507e-05, | |
| "loss": 0.8756, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 26.024311065673828, | |
| "learning_rate": 2.4263013698630136e-05, | |
| "loss": 0.9083, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5821917808219178, | |
| "grad_norm": 14.479564666748047, | |
| "learning_rate": 2.4194520547945205e-05, | |
| "loss": 0.9357, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.589041095890411, | |
| "grad_norm": 9.043543815612793, | |
| "learning_rate": 2.4126027397260275e-05, | |
| "loss": 0.9954, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5958904109589042, | |
| "grad_norm": 17.975173950195312, | |
| "learning_rate": 2.4057534246575344e-05, | |
| "loss": 0.8758, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 25.138750076293945, | |
| "learning_rate": 2.398904109589041e-05, | |
| "loss": 0.8641, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6095890410958904, | |
| "grad_norm": 10.003458976745605, | |
| "learning_rate": 2.392054794520548e-05, | |
| "loss": 1.0095, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6164383561643836, | |
| "grad_norm": 48.206233978271484, | |
| "learning_rate": 2.385205479452055e-05, | |
| "loss": 0.924, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6232876712328768, | |
| "grad_norm": 10.344615936279297, | |
| "learning_rate": 2.378356164383562e-05, | |
| "loss": 0.9114, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 15.59772777557373, | |
| "learning_rate": 2.3715068493150685e-05, | |
| "loss": 0.9313, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.636986301369863, | |
| "grad_norm": 4.01880407333374, | |
| "learning_rate": 2.3646575342465754e-05, | |
| "loss": 0.9007, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.6438356164383562, | |
| "grad_norm": 6.648110866546631, | |
| "learning_rate": 2.3578082191780824e-05, | |
| "loss": 0.9106, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6506849315068494, | |
| "grad_norm": 36.89479446411133, | |
| "learning_rate": 2.3509589041095893e-05, | |
| "loss": 0.8683, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 46.464141845703125, | |
| "learning_rate": 2.344109589041096e-05, | |
| "loss": 0.7399, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.6643835616438356, | |
| "grad_norm": 24.97540283203125, | |
| "learning_rate": 2.337260273972603e-05, | |
| "loss": 0.9609, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.6712328767123288, | |
| "grad_norm": 50.79806137084961, | |
| "learning_rate": 2.3304109589041098e-05, | |
| "loss": 0.8724, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.678082191780822, | |
| "grad_norm": 24.82395362854004, | |
| "learning_rate": 2.3235616438356168e-05, | |
| "loss": 0.8565, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 28.389963150024414, | |
| "learning_rate": 2.3167123287671234e-05, | |
| "loss": 0.8048, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "eval_exact_match": 84.12488174077578, | |
| "eval_f1": 91.07595753183573, | |
| "eval_runtime": 417.955, | |
| "eval_samples_per_second": 25.29, | |
| "eval_steps_per_second": 1.582, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6917808219178082, | |
| "grad_norm": 14.127419471740723, | |
| "learning_rate": 2.3098630136986303e-05, | |
| "loss": 0.7959, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.6986301369863014, | |
| "grad_norm": 12.644275665283203, | |
| "learning_rate": 2.3030136986301373e-05, | |
| "loss": 0.9732, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7054794520547946, | |
| "grad_norm": 26.438093185424805, | |
| "learning_rate": 2.296164383561644e-05, | |
| "loss": 0.8267, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 14.37074089050293, | |
| "learning_rate": 2.2893150684931505e-05, | |
| "loss": 0.8899, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7191780821917808, | |
| "grad_norm": 23.79348373413086, | |
| "learning_rate": 2.2824657534246574e-05, | |
| "loss": 0.906, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.726027397260274, | |
| "grad_norm": 12.331747055053711, | |
| "learning_rate": 2.2756164383561644e-05, | |
| "loss": 0.823, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.7328767123287672, | |
| "grad_norm": 17.69827651977539, | |
| "learning_rate": 2.2687671232876713e-05, | |
| "loss": 0.8854, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 30.975631713867188, | |
| "learning_rate": 2.261917808219178e-05, | |
| "loss": 0.8866, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7465753424657534, | |
| "grad_norm": 8.740948677062988, | |
| "learning_rate": 2.255068493150685e-05, | |
| "loss": 0.843, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "grad_norm": 25.42966651916504, | |
| "learning_rate": 2.248219178082192e-05, | |
| "loss": 0.937, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7602739726027398, | |
| "grad_norm": 15.364262580871582, | |
| "learning_rate": 2.2413698630136988e-05, | |
| "loss": 0.8485, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 13.131271362304688, | |
| "learning_rate": 2.2345205479452054e-05, | |
| "loss": 0.8018, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.773972602739726, | |
| "grad_norm": 9.219653129577637, | |
| "learning_rate": 2.2276712328767123e-05, | |
| "loss": 0.8492, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.7808219178082192, | |
| "grad_norm": 20.117149353027344, | |
| "learning_rate": 2.2208219178082193e-05, | |
| "loss": 0.9339, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7876712328767124, | |
| "grad_norm": 10.591473579406738, | |
| "learning_rate": 2.2139726027397262e-05, | |
| "loss": 0.8925, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 25.7907772064209, | |
| "learning_rate": 2.207123287671233e-05, | |
| "loss": 0.9092, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.8013698630136986, | |
| "grad_norm": 12.926765441894531, | |
| "learning_rate": 2.2002739726027398e-05, | |
| "loss": 0.8824, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.8082191780821918, | |
| "grad_norm": 15.865346908569336, | |
| "learning_rate": 2.1934246575342467e-05, | |
| "loss": 0.846, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.815068493150685, | |
| "grad_norm": 20.416015625, | |
| "learning_rate": 2.1865753424657537e-05, | |
| "loss": 0.8126, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 11.727408409118652, | |
| "learning_rate": 2.1797260273972603e-05, | |
| "loss": 0.8308, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "eval_exact_match": 84.57899716177862, | |
| "eval_f1": 91.43611698741225, | |
| "eval_runtime": 417.3274, | |
| "eval_samples_per_second": 25.328, | |
| "eval_steps_per_second": 1.584, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8287671232876712, | |
| "grad_norm": 25.75244903564453, | |
| "learning_rate": 2.1728767123287672e-05, | |
| "loss": 0.8027, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.8356164383561644, | |
| "grad_norm": 10.4827880859375, | |
| "learning_rate": 2.1660273972602742e-05, | |
| "loss": 0.7853, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8424657534246576, | |
| "grad_norm": 20.461090087890625, | |
| "learning_rate": 2.159178082191781e-05, | |
| "loss": 0.9597, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 5.970315456390381, | |
| "learning_rate": 2.1523287671232877e-05, | |
| "loss": 0.8921, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8561643835616438, | |
| "grad_norm": 31.718955993652344, | |
| "learning_rate": 2.1454794520547947e-05, | |
| "loss": 0.7981, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.863013698630137, | |
| "grad_norm": 7.239502906799316, | |
| "learning_rate": 2.1386301369863016e-05, | |
| "loss": 0.8386, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8698630136986302, | |
| "grad_norm": 8.74398422241211, | |
| "learning_rate": 2.1317808219178086e-05, | |
| "loss": 0.8466, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 26.66971206665039, | |
| "learning_rate": 2.1249315068493152e-05, | |
| "loss": 0.8054, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8835616438356164, | |
| "grad_norm": 30.80714988708496, | |
| "learning_rate": 2.1180821917808218e-05, | |
| "loss": 0.8719, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.8904109589041096, | |
| "grad_norm": 10.245841026306152, | |
| "learning_rate": 2.1112328767123287e-05, | |
| "loss": 0.8324, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8972602739726028, | |
| "grad_norm": 11.401703834533691, | |
| "learning_rate": 2.1043835616438357e-05, | |
| "loss": 0.7636, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 10.791388511657715, | |
| "learning_rate": 2.0975342465753423e-05, | |
| "loss": 0.8163, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.910958904109589, | |
| "grad_norm": 73.97101593017578, | |
| "learning_rate": 2.0906849315068493e-05, | |
| "loss": 0.8267, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.9178082191780822, | |
| "grad_norm": 5.0418925285339355, | |
| "learning_rate": 2.0838356164383562e-05, | |
| "loss": 0.7787, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.9246575342465754, | |
| "grad_norm": 16.25297737121582, | |
| "learning_rate": 2.076986301369863e-05, | |
| "loss": 0.8305, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 15.411580085754395, | |
| "learning_rate": 2.0701369863013698e-05, | |
| "loss": 0.7606, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9383561643835616, | |
| "grad_norm": 14.359014511108398, | |
| "learning_rate": 2.0632876712328767e-05, | |
| "loss": 0.8578, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.9452054794520548, | |
| "grad_norm": 21.724008560180664, | |
| "learning_rate": 2.0564383561643836e-05, | |
| "loss": 0.7836, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.952054794520548, | |
| "grad_norm": 16.269027709960938, | |
| "learning_rate": 2.0495890410958906e-05, | |
| "loss": 0.8681, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 33.833797454833984, | |
| "learning_rate": 2.0427397260273972e-05, | |
| "loss": 0.8034, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "eval_exact_match": 84.64522232734153, | |
| "eval_f1": 91.36486838251115, | |
| "eval_runtime": 419.048, | |
| "eval_samples_per_second": 25.224, | |
| "eval_steps_per_second": 1.577, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9657534246575342, | |
| "grad_norm": 11.25650691986084, | |
| "learning_rate": 2.035890410958904e-05, | |
| "loss": 0.8451, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.9726027397260274, | |
| "grad_norm": 42.16932678222656, | |
| "learning_rate": 2.029041095890411e-05, | |
| "loss": 0.8701, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.9794520547945206, | |
| "grad_norm": 23.863306045532227, | |
| "learning_rate": 2.022191780821918e-05, | |
| "loss": 0.8651, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 40.55854034423828, | |
| "learning_rate": 2.0153424657534247e-05, | |
| "loss": 0.8481, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.9931506849315068, | |
| "grad_norm": 14.687817573547363, | |
| "learning_rate": 2.0084931506849316e-05, | |
| "loss": 0.8462, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 27.034555435180664, | |
| "learning_rate": 2.0016438356164386e-05, | |
| "loss": 0.7084, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.0068493150684932, | |
| "grad_norm": 20.166122436523438, | |
| "learning_rate": 1.9947945205479455e-05, | |
| "loss": 0.6103, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 7.819751262664795, | |
| "learning_rate": 1.987945205479452e-05, | |
| "loss": 0.5915, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.0205479452054795, | |
| "grad_norm": 7.091070175170898, | |
| "learning_rate": 1.981095890410959e-05, | |
| "loss": 0.6133, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.0273972602739727, | |
| "grad_norm": 8.464035034179688, | |
| "learning_rate": 1.974246575342466e-05, | |
| "loss": 0.626, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0342465753424657, | |
| "grad_norm": 9.27121353149414, | |
| "learning_rate": 1.9673972602739726e-05, | |
| "loss": 0.6166, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 10.510052680969238, | |
| "learning_rate": 1.9605479452054796e-05, | |
| "loss": 0.5873, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.047945205479452, | |
| "grad_norm": 12.773880958557129, | |
| "learning_rate": 1.9536986301369865e-05, | |
| "loss": 0.6074, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.0547945205479452, | |
| "grad_norm": 5.2743425369262695, | |
| "learning_rate": 1.9468493150684935e-05, | |
| "loss": 0.5606, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.0616438356164384, | |
| "grad_norm": 45.79352569580078, | |
| "learning_rate": 1.9399999999999997e-05, | |
| "loss": 0.6101, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 9.683128356933594, | |
| "learning_rate": 1.9331506849315067e-05, | |
| "loss": 0.521, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.0753424657534247, | |
| "grad_norm": 15.017319679260254, | |
| "learning_rate": 1.9263013698630136e-05, | |
| "loss": 0.6256, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.0821917808219177, | |
| "grad_norm": 17.09823226928711, | |
| "learning_rate": 1.9194520547945206e-05, | |
| "loss": 0.6204, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.0890410958904109, | |
| "grad_norm": 16.04283905029297, | |
| "learning_rate": 1.9126027397260272e-05, | |
| "loss": 0.6075, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 10.665413856506348, | |
| "learning_rate": 1.905753424657534e-05, | |
| "loss": 0.5418, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 85.19394512771996, | |
| "eval_f1": 92.0232735390316, | |
| "eval_runtime": 417.4898, | |
| "eval_samples_per_second": 25.318, | |
| "eval_steps_per_second": 1.583, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1027397260273972, | |
| "grad_norm": 28.214811325073242, | |
| "learning_rate": 1.898904109589041e-05, | |
| "loss": 0.6014, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.1095890410958904, | |
| "grad_norm": 34.26095199584961, | |
| "learning_rate": 1.892054794520548e-05, | |
| "loss": 0.6195, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.1164383561643836, | |
| "grad_norm": 7.891120910644531, | |
| "learning_rate": 1.8852054794520546e-05, | |
| "loss": 0.723, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 21.476566314697266, | |
| "learning_rate": 1.8783561643835616e-05, | |
| "loss": 0.6577, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.13013698630137, | |
| "grad_norm": 15.302453994750977, | |
| "learning_rate": 1.8715068493150685e-05, | |
| "loss": 0.5872, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.1369863013698631, | |
| "grad_norm": 22.77974510192871, | |
| "learning_rate": 1.8646575342465755e-05, | |
| "loss": 0.6372, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.143835616438356, | |
| "grad_norm": 7.8337297439575195, | |
| "learning_rate": 1.857808219178082e-05, | |
| "loss": 0.5487, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 26.61360740661621, | |
| "learning_rate": 1.850958904109589e-05, | |
| "loss": 0.5555, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.1575342465753424, | |
| "grad_norm": 9.870406150817871, | |
| "learning_rate": 1.844109589041096e-05, | |
| "loss": 0.5728, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.1643835616438356, | |
| "grad_norm": 8.367820739746094, | |
| "learning_rate": 1.837260273972603e-05, | |
| "loss": 0.6081, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.1712328767123288, | |
| "grad_norm": 22.267404556274414, | |
| "learning_rate": 1.8304109589041095e-05, | |
| "loss": 0.6432, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 20.5167236328125, | |
| "learning_rate": 1.8235616438356165e-05, | |
| "loss": 0.5952, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.1849315068493151, | |
| "grad_norm": 13.30615520477295, | |
| "learning_rate": 1.8167123287671234e-05, | |
| "loss": 0.6364, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.191780821917808, | |
| "grad_norm": 16.95148277282715, | |
| "learning_rate": 1.8098630136986304e-05, | |
| "loss": 0.676, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.1986301369863013, | |
| "grad_norm": 19.168123245239258, | |
| "learning_rate": 1.803013698630137e-05, | |
| "loss": 0.5939, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 13.934683799743652, | |
| "learning_rate": 1.796164383561644e-05, | |
| "loss": 0.6304, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.2123287671232876, | |
| "grad_norm": 16.010356903076172, | |
| "learning_rate": 1.789315068493151e-05, | |
| "loss": 0.7142, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.2191780821917808, | |
| "grad_norm": 20.430280685424805, | |
| "learning_rate": 1.7824657534246578e-05, | |
| "loss": 0.6254, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.226027397260274, | |
| "grad_norm": 44.09880828857422, | |
| "learning_rate": 1.7756164383561644e-05, | |
| "loss": 0.6566, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 41.73299026489258, | |
| "learning_rate": 1.7687671232876714e-05, | |
| "loss": 0.5479, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "eval_exact_match": 85.38315988647115, | |
| "eval_f1": 92.15019137693149, | |
| "eval_runtime": 418.4718, | |
| "eval_samples_per_second": 25.259, | |
| "eval_steps_per_second": 1.58, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2397260273972603, | |
| "grad_norm": 20.86771011352539, | |
| "learning_rate": 1.761917808219178e-05, | |
| "loss": 0.5645, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.2465753424657535, | |
| "grad_norm": 6.557628631591797, | |
| "learning_rate": 1.755068493150685e-05, | |
| "loss": 0.6265, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.2534246575342465, | |
| "grad_norm": 6.776758193969727, | |
| "learning_rate": 1.7482191780821915e-05, | |
| "loss": 0.5482, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 20.538101196289062, | |
| "learning_rate": 1.7413698630136985e-05, | |
| "loss": 0.5793, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.2671232876712328, | |
| "grad_norm": 19.532896041870117, | |
| "learning_rate": 1.7345205479452054e-05, | |
| "loss": 0.6236, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.273972602739726, | |
| "grad_norm": 16.863801956176758, | |
| "learning_rate": 1.7276712328767124e-05, | |
| "loss": 0.5566, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.2808219178082192, | |
| "grad_norm": 46.87477493286133, | |
| "learning_rate": 1.720821917808219e-05, | |
| "loss": 0.5785, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 10.397550582885742, | |
| "learning_rate": 1.713972602739726e-05, | |
| "loss": 0.7018, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.2945205479452055, | |
| "grad_norm": 6.6365275382995605, | |
| "learning_rate": 1.7072602739726028e-05, | |
| "loss": 0.5387, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.3013698630136985, | |
| "grad_norm": 24.18856430053711, | |
| "learning_rate": 1.7004109589041094e-05, | |
| "loss": 0.5884, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.308219178082192, | |
| "grad_norm": 33.152278900146484, | |
| "learning_rate": 1.6935616438356164e-05, | |
| "loss": 0.6385, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 4.547526836395264, | |
| "learning_rate": 1.6867123287671233e-05, | |
| "loss": 0.5276, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.321917808219178, | |
| "grad_norm": 34.75554656982422, | |
| "learning_rate": 1.6798630136986303e-05, | |
| "loss": 0.6225, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.3287671232876712, | |
| "grad_norm": 15.248549461364746, | |
| "learning_rate": 1.673013698630137e-05, | |
| "loss": 0.5654, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.3356164383561644, | |
| "grad_norm": 19.826614379882812, | |
| "learning_rate": 1.6661643835616438e-05, | |
| "loss": 0.6356, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 29.037817001342773, | |
| "learning_rate": 1.6593150684931508e-05, | |
| "loss": 0.5997, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.3493150684931507, | |
| "grad_norm": 5.311439037322998, | |
| "learning_rate": 1.6524657534246577e-05, | |
| "loss": 0.5873, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.356164383561644, | |
| "grad_norm": 44.800636291503906, | |
| "learning_rate": 1.6456164383561643e-05, | |
| "loss": 0.5232, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.3630136986301369, | |
| "grad_norm": 19.861589431762695, | |
| "learning_rate": 1.6387671232876713e-05, | |
| "loss": 0.5877, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 8.236838340759277, | |
| "learning_rate": 1.6319178082191782e-05, | |
| "loss": 0.5612, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "eval_exact_match": 84.95742667928099, | |
| "eval_f1": 91.94903795865467, | |
| "eval_runtime": 417.7906, | |
| "eval_samples_per_second": 25.3, | |
| "eval_steps_per_second": 1.582, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.3767123287671232, | |
| "grad_norm": 5.004638671875, | |
| "learning_rate": 1.625068493150685e-05, | |
| "loss": 0.6488, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.3835616438356164, | |
| "grad_norm": 6.322940349578857, | |
| "learning_rate": 1.6182191780821918e-05, | |
| "loss": 0.53, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.3904109589041096, | |
| "grad_norm": 18.412700653076172, | |
| "learning_rate": 1.6113698630136987e-05, | |
| "loss": 0.6148, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 10.37777042388916, | |
| "learning_rate": 1.6045205479452057e-05, | |
| "loss": 0.5429, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.404109589041096, | |
| "grad_norm": 14.783160209655762, | |
| "learning_rate": 1.5976712328767126e-05, | |
| "loss": 0.5469, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.410958904109589, | |
| "grad_norm": 9.634529113769531, | |
| "learning_rate": 1.5908219178082192e-05, | |
| "loss": 0.5273, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.4178082191780823, | |
| "grad_norm": 8.868270874023438, | |
| "learning_rate": 1.5839726027397258e-05, | |
| "loss": 0.6968, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 2.925807476043701, | |
| "learning_rate": 1.5771232876712328e-05, | |
| "loss": 0.5857, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.4315068493150684, | |
| "grad_norm": 14.901018142700195, | |
| "learning_rate": 1.5702739726027397e-05, | |
| "loss": 0.5813, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.4383561643835616, | |
| "grad_norm": 18.872957229614258, | |
| "learning_rate": 1.5634246575342463e-05, | |
| "loss": 0.6373, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.4452054794520548, | |
| "grad_norm": 35.207847595214844, | |
| "learning_rate": 1.5565753424657533e-05, | |
| "loss": 0.5539, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 11.02206802368164, | |
| "learning_rate": 1.5497260273972602e-05, | |
| "loss": 0.5957, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.4589041095890412, | |
| "grad_norm": 25.235326766967773, | |
| "learning_rate": 1.5428767123287672e-05, | |
| "loss": 0.6574, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.4657534246575343, | |
| "grad_norm": 32.09264373779297, | |
| "learning_rate": 1.5360273972602738e-05, | |
| "loss": 0.5925, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.4726027397260273, | |
| "grad_norm": 19.04875373840332, | |
| "learning_rate": 1.5291780821917807e-05, | |
| "loss": 0.5259, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 16.185894012451172, | |
| "learning_rate": 1.5223287671232877e-05, | |
| "loss": 0.6109, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.4863013698630136, | |
| "grad_norm": 17.464332580566406, | |
| "learning_rate": 1.5154794520547946e-05, | |
| "loss": 0.5961, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.4931506849315068, | |
| "grad_norm": 8.539608001708984, | |
| "learning_rate": 1.5086301369863012e-05, | |
| "loss": 0.5867, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 28.3868465423584, | |
| "learning_rate": 1.5017808219178082e-05, | |
| "loss": 0.6098, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 6.247623443603516, | |
| "learning_rate": 1.4949315068493151e-05, | |
| "loss": 0.6059, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "eval_exact_match": 85.22232734153263, | |
| "eval_f1": 92.06076201446909, | |
| "eval_runtime": 416.847, | |
| "eval_samples_per_second": 25.357, | |
| "eval_steps_per_second": 1.586, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.5136986301369864, | |
| "grad_norm": 22.93709373474121, | |
| "learning_rate": 1.4880821917808219e-05, | |
| "loss": 0.5952, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 1.5205479452054793, | |
| "grad_norm": 40.23944091796875, | |
| "learning_rate": 1.4812328767123289e-05, | |
| "loss": 0.5395, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.5273972602739727, | |
| "grad_norm": 18.1081485748291, | |
| "learning_rate": 1.4743835616438356e-05, | |
| "loss": 0.5989, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 34.16764831542969, | |
| "learning_rate": 1.4675342465753426e-05, | |
| "loss": 0.5738, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 1.541095890410959, | |
| "grad_norm": 8.558218002319336, | |
| "learning_rate": 1.4606849315068494e-05, | |
| "loss": 0.5767, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.547945205479452, | |
| "grad_norm": 50.398895263671875, | |
| "learning_rate": 1.4538356164383563e-05, | |
| "loss": 0.581, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 1.5547945205479452, | |
| "grad_norm": 8.642769813537598, | |
| "learning_rate": 1.4469863013698629e-05, | |
| "loss": 0.6654, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 22.660131454467773, | |
| "learning_rate": 1.4401369863013699e-05, | |
| "loss": 0.6377, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.5684931506849316, | |
| "grad_norm": 17.38098907470703, | |
| "learning_rate": 1.4332876712328766e-05, | |
| "loss": 0.5625, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 1.5753424657534247, | |
| "grad_norm": 9.987977981567383, | |
| "learning_rate": 1.4264383561643836e-05, | |
| "loss": 0.5191, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.5821917808219177, | |
| "grad_norm": 42.332763671875, | |
| "learning_rate": 1.4195890410958904e-05, | |
| "loss": 0.6211, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 8.304152488708496, | |
| "learning_rate": 1.4127397260273973e-05, | |
| "loss": 0.6175, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 1.595890410958904, | |
| "grad_norm": 142.1208038330078, | |
| "learning_rate": 1.4058904109589041e-05, | |
| "loss": 0.5902, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 1.6027397260273972, | |
| "grad_norm": 13.38881778717041, | |
| "learning_rate": 1.399041095890411e-05, | |
| "loss": 0.5528, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.6095890410958904, | |
| "grad_norm": 8.86255168914795, | |
| "learning_rate": 1.3921917808219178e-05, | |
| "loss": 0.6205, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 33.42393493652344, | |
| "learning_rate": 1.3853424657534248e-05, | |
| "loss": 0.6477, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 1.6232876712328768, | |
| "grad_norm": 15.220393180847168, | |
| "learning_rate": 1.3784931506849315e-05, | |
| "loss": 0.5493, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 1.6301369863013697, | |
| "grad_norm": 9.7684907913208, | |
| "learning_rate": 1.3716438356164385e-05, | |
| "loss": 0.5256, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 1.6369863013698631, | |
| "grad_norm": 19.64045524597168, | |
| "learning_rate": 1.3647945205479453e-05, | |
| "loss": 0.5653, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 28.879430770874023, | |
| "learning_rate": 1.357945205479452e-05, | |
| "loss": 0.5989, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_exact_match": 85.12771996215704, | |
| "eval_f1": 91.80306824412318, | |
| "eval_runtime": 417.2891, | |
| "eval_samples_per_second": 25.33, | |
| "eval_steps_per_second": 1.584, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.6506849315068495, | |
| "grad_norm": 7.393039226531982, | |
| "learning_rate": 1.3510958904109588e-05, | |
| "loss": 0.5271, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 1.6575342465753424, | |
| "grad_norm": 10.611188888549805, | |
| "learning_rate": 1.3442465753424658e-05, | |
| "loss": 0.5939, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 1.6643835616438356, | |
| "grad_norm": 9.510908126831055, | |
| "learning_rate": 1.3373972602739725e-05, | |
| "loss": 0.5574, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 15.351234436035156, | |
| "learning_rate": 1.3305479452054795e-05, | |
| "loss": 0.5861, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 1.678082191780822, | |
| "grad_norm": 31.311676025390625, | |
| "learning_rate": 1.3236986301369863e-05, | |
| "loss": 0.5568, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.6849315068493151, | |
| "grad_norm": 4.728596210479736, | |
| "learning_rate": 1.3168493150684932e-05, | |
| "loss": 0.5431, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.691780821917808, | |
| "grad_norm": 21.880786895751953, | |
| "learning_rate": 1.31e-05, | |
| "loss": 0.5993, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 5.625259876251221, | |
| "learning_rate": 1.303150684931507e-05, | |
| "loss": 0.6345, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 1.7054794520547945, | |
| "grad_norm": 12.03124713897705, | |
| "learning_rate": 1.2963013698630137e-05, | |
| "loss": 0.6195, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 1.7123287671232876, | |
| "grad_norm": 11.920297622680664, | |
| "learning_rate": 1.2894520547945207e-05, | |
| "loss": 0.5523, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.7191780821917808, | |
| "grad_norm": 12.449995994567871, | |
| "learning_rate": 1.2826027397260274e-05, | |
| "loss": 0.5442, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 15.602882385253906, | |
| "learning_rate": 1.2757534246575342e-05, | |
| "loss": 0.4794, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.7328767123287672, | |
| "grad_norm": 27.904523849487305, | |
| "learning_rate": 1.268904109589041e-05, | |
| "loss": 0.5184, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 1.7397260273972601, | |
| "grad_norm": 6.819875717163086, | |
| "learning_rate": 1.262054794520548e-05, | |
| "loss": 0.4637, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 1.7465753424657535, | |
| "grad_norm": 17.69037437438965, | |
| "learning_rate": 1.2552054794520547e-05, | |
| "loss": 0.5248, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 25.76197052001953, | |
| "learning_rate": 1.2483561643835617e-05, | |
| "loss": 0.6165, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 1.7602739726027399, | |
| "grad_norm": 5.317371845245361, | |
| "learning_rate": 1.2415068493150685e-05, | |
| "loss": 0.5206, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 1.7671232876712328, | |
| "grad_norm": 8.703845977783203, | |
| "learning_rate": 1.2346575342465754e-05, | |
| "loss": 0.5994, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.773972602739726, | |
| "grad_norm": 16.243268966674805, | |
| "learning_rate": 1.2278082191780822e-05, | |
| "loss": 0.5416, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 15.478755950927734, | |
| "learning_rate": 1.2209589041095891e-05, | |
| "loss": 0.6375, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "eval_exact_match": 85.65752128666036, | |
| "eval_f1": 92.15631910787795, | |
| "eval_runtime": 416.7008, | |
| "eval_samples_per_second": 25.366, | |
| "eval_steps_per_second": 1.586, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7876712328767124, | |
| "grad_norm": 14.403656005859375, | |
| "learning_rate": 1.2141095890410959e-05, | |
| "loss": 0.6117, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 1.7945205479452055, | |
| "grad_norm": 22.657033920288086, | |
| "learning_rate": 1.2072602739726028e-05, | |
| "loss": 0.5757, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 1.8013698630136985, | |
| "grad_norm": 13.059876441955566, | |
| "learning_rate": 1.2004109589041096e-05, | |
| "loss": 0.5765, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 2.8631415367126465, | |
| "learning_rate": 1.1935616438356166e-05, | |
| "loss": 0.5688, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.8150684931506849, | |
| "grad_norm": 8.850284576416016, | |
| "learning_rate": 1.1867123287671232e-05, | |
| "loss": 0.5713, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 1.821917808219178, | |
| "grad_norm": 9.604997634887695, | |
| "learning_rate": 1.1798630136986301e-05, | |
| "loss": 0.5602, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 1.8287671232876712, | |
| "grad_norm": 27.603858947753906, | |
| "learning_rate": 1.1730136986301369e-05, | |
| "loss": 0.6105, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 10.399881362915039, | |
| "learning_rate": 1.1661643835616439e-05, | |
| "loss": 0.5744, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 1.8424657534246576, | |
| "grad_norm": 16.775104522705078, | |
| "learning_rate": 1.1593150684931506e-05, | |
| "loss": 0.5776, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 1.8493150684931505, | |
| "grad_norm": 8.317610740661621, | |
| "learning_rate": 1.1524657534246576e-05, | |
| "loss": 0.6075, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.856164383561644, | |
| "grad_norm": 18.899354934692383, | |
| "learning_rate": 1.1456164383561644e-05, | |
| "loss": 0.5967, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 10.251896858215332, | |
| "learning_rate": 1.1387671232876713e-05, | |
| "loss": 0.6121, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 1.8698630136986303, | |
| "grad_norm": 24.907438278198242, | |
| "learning_rate": 1.131917808219178e-05, | |
| "loss": 0.6348, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 1.8767123287671232, | |
| "grad_norm": 17.239213943481445, | |
| "learning_rate": 1.125068493150685e-05, | |
| "loss": 0.5835, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 1.8835616438356164, | |
| "grad_norm": 14.36588191986084, | |
| "learning_rate": 1.1182191780821918e-05, | |
| "loss": 0.5681, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 10.424467086791992, | |
| "learning_rate": 1.1113698630136988e-05, | |
| "loss": 0.5334, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.8972602739726028, | |
| "grad_norm": 11.122437477111816, | |
| "learning_rate": 1.1045205479452055e-05, | |
| "loss": 0.5594, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 1.904109589041096, | |
| "grad_norm": 10.735795021057129, | |
| "learning_rate": 1.0976712328767123e-05, | |
| "loss": 0.5565, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 1.910958904109589, | |
| "grad_norm": 10.91677474975586, | |
| "learning_rate": 1.0908219178082191e-05, | |
| "loss": 0.5945, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 7.375208377838135, | |
| "learning_rate": 1.083972602739726e-05, | |
| "loss": 0.536, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "eval_exact_match": 86.10217596972564, | |
| "eval_f1": 92.4099426929563, | |
| "eval_runtime": 417.908, | |
| "eval_samples_per_second": 25.293, | |
| "eval_steps_per_second": 1.582, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9246575342465753, | |
| "grad_norm": 32.034141540527344, | |
| "learning_rate": 1.0771232876712328e-05, | |
| "loss": 0.5696, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 1.9315068493150684, | |
| "grad_norm": 21.64228057861328, | |
| "learning_rate": 1.0702739726027398e-05, | |
| "loss": 0.6152, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.9383561643835616, | |
| "grad_norm": 21.606149673461914, | |
| "learning_rate": 1.0634246575342465e-05, | |
| "loss": 0.5286, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 29.915918350219727, | |
| "learning_rate": 1.0565753424657535e-05, | |
| "loss": 0.5781, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 1.952054794520548, | |
| "grad_norm": 17.88494873046875, | |
| "learning_rate": 1.0497260273972603e-05, | |
| "loss": 0.6066, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 1.958904109589041, | |
| "grad_norm": 24.821956634521484, | |
| "learning_rate": 1.0428767123287672e-05, | |
| "loss": 0.5819, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 1.9657534246575343, | |
| "grad_norm": 34.31120681762695, | |
| "learning_rate": 1.036027397260274e-05, | |
| "loss": 0.6531, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 13.344314575195312, | |
| "learning_rate": 1.029178082191781e-05, | |
| "loss": 0.5585, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.9794520547945207, | |
| "grad_norm": 40.556358337402344, | |
| "learning_rate": 1.0223287671232877e-05, | |
| "loss": 0.5257, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 1.9863013698630136, | |
| "grad_norm": 11.713874816894531, | |
| "learning_rate": 1.0154794520547947e-05, | |
| "loss": 0.6307, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.9931506849315068, | |
| "grad_norm": 9.927220344543457, | |
| "learning_rate": 1.0086301369863013e-05, | |
| "loss": 0.5814, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 14.175002098083496, | |
| "learning_rate": 1.0017808219178082e-05, | |
| "loss": 0.6428, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.006849315068493, | |
| "grad_norm": 28.156503677368164, | |
| "learning_rate": 9.94931506849315e-06, | |
| "loss": 0.3428, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.0136986301369864, | |
| "grad_norm": 12.531486511230469, | |
| "learning_rate": 9.88082191780822e-06, | |
| "loss": 0.3625, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.0205479452054793, | |
| "grad_norm": 6.839471340179443, | |
| "learning_rate": 9.812328767123287e-06, | |
| "loss": 0.3675, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 9.022225379943848, | |
| "learning_rate": 9.743835616438357e-06, | |
| "loss": 0.3613, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.0342465753424657, | |
| "grad_norm": 9.764174461364746, | |
| "learning_rate": 9.675342465753424e-06, | |
| "loss": 0.3353, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.041095890410959, | |
| "grad_norm": 8.139081001281738, | |
| "learning_rate": 9.606849315068494e-06, | |
| "loss": 0.3749, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.047945205479452, | |
| "grad_norm": 19.99053955078125, | |
| "learning_rate": 9.538356164383562e-06, | |
| "loss": 0.3434, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 4.129643440246582, | |
| "learning_rate": 9.469863013698631e-06, | |
| "loss": 0.3753, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "eval_exact_match": 85.34531693472091, | |
| "eval_f1": 92.12097175946987, | |
| "eval_runtime": 417.8979, | |
| "eval_samples_per_second": 25.293, | |
| "eval_steps_per_second": 1.582, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.0616438356164384, | |
| "grad_norm": 33.54549026489258, | |
| "learning_rate": 9.401369863013699e-06, | |
| "loss": 0.3592, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.0684931506849313, | |
| "grad_norm": 24.34673500061035, | |
| "learning_rate": 9.332876712328768e-06, | |
| "loss": 0.3847, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.0753424657534247, | |
| "grad_norm": 5.306540489196777, | |
| "learning_rate": 9.264383561643836e-06, | |
| "loss": 0.3157, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 5.992457866668701, | |
| "learning_rate": 9.195890410958904e-06, | |
| "loss": 0.3308, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.089041095890411, | |
| "grad_norm": 15.749208450317383, | |
| "learning_rate": 9.127397260273972e-06, | |
| "loss": 0.3631, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.095890410958904, | |
| "grad_norm": 10.97288703918457, | |
| "learning_rate": 9.058904109589041e-06, | |
| "loss": 0.3107, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.1027397260273974, | |
| "grad_norm": 5.519676208496094, | |
| "learning_rate": 8.990410958904109e-06, | |
| "loss": 0.3225, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 24.270828247070312, | |
| "learning_rate": 8.921917808219179e-06, | |
| "loss": 0.3833, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.1164383561643834, | |
| "grad_norm": 17.408357620239258, | |
| "learning_rate": 8.854794520547946e-06, | |
| "loss": 0.3484, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.1232876712328768, | |
| "grad_norm": 19.13545036315918, | |
| "learning_rate": 8.786301369863013e-06, | |
| "loss": 0.3417, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.1301369863013697, | |
| "grad_norm": 12.084640502929688, | |
| "learning_rate": 8.717808219178083e-06, | |
| "loss": 0.3475, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 32.05975341796875, | |
| "learning_rate": 8.64931506849315e-06, | |
| "loss": 0.3406, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.143835616438356, | |
| "grad_norm": 19.190528869628906, | |
| "learning_rate": 8.58082191780822e-06, | |
| "loss": 0.4047, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.1506849315068495, | |
| "grad_norm": 2.928743600845337, | |
| "learning_rate": 8.512328767123288e-06, | |
| "loss": 0.3339, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.1575342465753424, | |
| "grad_norm": 11.721294403076172, | |
| "learning_rate": 8.443835616438357e-06, | |
| "loss": 0.3314, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 8.114335060119629, | |
| "learning_rate": 8.375342465753425e-06, | |
| "loss": 0.3736, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.171232876712329, | |
| "grad_norm": 14.555135726928711, | |
| "learning_rate": 8.306849315068495e-06, | |
| "loss": 0.3332, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.1780821917808217, | |
| "grad_norm": 81.77395629882812, | |
| "learning_rate": 8.23835616438356e-06, | |
| "loss": 0.3049, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.184931506849315, | |
| "grad_norm": 8.095141410827637, | |
| "learning_rate": 8.16986301369863e-06, | |
| "loss": 0.2637, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 17.374752044677734, | |
| "learning_rate": 8.101369863013698e-06, | |
| "loss": 0.4439, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 85.61021759697256, | |
| "eval_f1": 92.16749429035703, | |
| "eval_runtime": 416.4255, | |
| "eval_samples_per_second": 25.383, | |
| "eval_steps_per_second": 1.587, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.1986301369863015, | |
| "grad_norm": 8.427570343017578, | |
| "learning_rate": 8.032876712328767e-06, | |
| "loss": 0.2902, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.2054794520547945, | |
| "grad_norm": 6.139337539672852, | |
| "learning_rate": 7.964383561643835e-06, | |
| "loss": 0.3242, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.212328767123288, | |
| "grad_norm": 26.255056381225586, | |
| "learning_rate": 7.895890410958905e-06, | |
| "loss": 0.3496, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 16.986501693725586, | |
| "learning_rate": 7.827397260273972e-06, | |
| "loss": 0.3297, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.2260273972602738, | |
| "grad_norm": 17.626201629638672, | |
| "learning_rate": 7.758904109589042e-06, | |
| "loss": 0.3733, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.232876712328767, | |
| "grad_norm": 20.750221252441406, | |
| "learning_rate": 7.69041095890411e-06, | |
| "loss": 0.3316, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.23972602739726, | |
| "grad_norm": 42.70435333251953, | |
| "learning_rate": 7.621917808219179e-06, | |
| "loss": 0.3721, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 11.971570014953613, | |
| "learning_rate": 7.553424657534246e-06, | |
| "loss": 0.3056, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 2.2534246575342465, | |
| "grad_norm": 7.298882484436035, | |
| "learning_rate": 7.484931506849315e-06, | |
| "loss": 0.4077, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 2.26027397260274, | |
| "grad_norm": 8.013861656188965, | |
| "learning_rate": 7.416438356164383e-06, | |
| "loss": 0.3488, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.267123287671233, | |
| "grad_norm": 10.631738662719727, | |
| "learning_rate": 7.347945205479452e-06, | |
| "loss": 0.3567, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 47.24718475341797, | |
| "learning_rate": 7.2794520547945206e-06, | |
| "loss": 0.3319, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 2.280821917808219, | |
| "grad_norm": 31.66643714904785, | |
| "learning_rate": 7.210958904109589e-06, | |
| "loss": 0.3768, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 2.287671232876712, | |
| "grad_norm": 34.2202033996582, | |
| "learning_rate": 7.142465753424657e-06, | |
| "loss": 0.4115, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 2.2945205479452055, | |
| "grad_norm": 35.246585845947266, | |
| "learning_rate": 7.073972602739726e-06, | |
| "loss": 0.3538, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 17.598068237304688, | |
| "learning_rate": 7.005479452054794e-06, | |
| "loss": 0.3189, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 2.308219178082192, | |
| "grad_norm": 31.153356552124023, | |
| "learning_rate": 6.936986301369863e-06, | |
| "loss": 0.4082, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 2.315068493150685, | |
| "grad_norm": 3.9330878257751465, | |
| "learning_rate": 6.8684931506849315e-06, | |
| "loss": 0.366, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 2.3219178082191783, | |
| "grad_norm": 13.598389625549316, | |
| "learning_rate": 6.8e-06, | |
| "loss": 0.3723, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 12.808009147644043, | |
| "learning_rate": 6.731506849315069e-06, | |
| "loss": 0.323, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "eval_exact_match": 85.30747398297068, | |
| "eval_f1": 92.19853158333738, | |
| "eval_runtime": 418.1668, | |
| "eval_samples_per_second": 25.277, | |
| "eval_steps_per_second": 1.581, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.3356164383561646, | |
| "grad_norm": 7.125260353088379, | |
| "learning_rate": 6.6630136986301365e-06, | |
| "loss": 0.3353, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 2.3424657534246576, | |
| "grad_norm": 11.679519653320312, | |
| "learning_rate": 6.594520547945205e-06, | |
| "loss": 0.3704, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 2.3493150684931505, | |
| "grad_norm": 12.352055549621582, | |
| "learning_rate": 6.526027397260274e-06, | |
| "loss": 0.3419, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 5.7046709060668945, | |
| "learning_rate": 6.457534246575342e-06, | |
| "loss": 0.3601, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 2.363013698630137, | |
| "grad_norm": 6.679656505584717, | |
| "learning_rate": 6.389041095890411e-06, | |
| "loss": 0.3537, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.3698630136986303, | |
| "grad_norm": 1.9317164421081543, | |
| "learning_rate": 6.32054794520548e-06, | |
| "loss": 0.3496, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 2.3767123287671232, | |
| "grad_norm": 7.0557026863098145, | |
| "learning_rate": 6.2520547945205474e-06, | |
| "loss": 0.3469, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 10.497695922851562, | |
| "learning_rate": 6.183561643835616e-06, | |
| "loss": 0.3477, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.3904109589041096, | |
| "grad_norm": 7.504631996154785, | |
| "learning_rate": 6.115068493150685e-06, | |
| "loss": 0.3188, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 2.3972602739726026, | |
| "grad_norm": 2.7814149856567383, | |
| "learning_rate": 6.046575342465753e-06, | |
| "loss": 0.3582, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.404109589041096, | |
| "grad_norm": 4.986628532409668, | |
| "learning_rate": 5.978082191780822e-06, | |
| "loss": 0.3234, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 10.343502044677734, | |
| "learning_rate": 5.9095890410958906e-06, | |
| "loss": 0.3475, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 2.4178082191780823, | |
| "grad_norm": 19.3160400390625, | |
| "learning_rate": 5.841095890410958e-06, | |
| "loss": 0.3722, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 2.4246575342465753, | |
| "grad_norm": 11.461263656616211, | |
| "learning_rate": 5.772602739726027e-06, | |
| "loss": 0.3842, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.4315068493150687, | |
| "grad_norm": 9.144445419311523, | |
| "learning_rate": 5.704109589041096e-06, | |
| "loss": 0.3361, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 6.6573686599731445, | |
| "learning_rate": 5.635616438356164e-06, | |
| "loss": 0.3334, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 2.4452054794520546, | |
| "grad_norm": 8.072036743164062, | |
| "learning_rate": 5.567123287671233e-06, | |
| "loss": 0.2792, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 2.452054794520548, | |
| "grad_norm": 28.824522018432617, | |
| "learning_rate": 5.4986301369863015e-06, | |
| "loss": 0.3438, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 2.458904109589041, | |
| "grad_norm": 13.252080917358398, | |
| "learning_rate": 5.43013698630137e-06, | |
| "loss": 0.3496, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 46.41205596923828, | |
| "learning_rate": 5.361643835616438e-06, | |
| "loss": 0.381, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "eval_exact_match": 85.30747398297068, | |
| "eval_f1": 92.19996078048693, | |
| "eval_runtime": 416.6834, | |
| "eval_samples_per_second": 25.367, | |
| "eval_steps_per_second": 1.586, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4726027397260273, | |
| "grad_norm": 10.28970718383789, | |
| "learning_rate": 5.2931506849315065e-06, | |
| "loss": 0.396, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 2.4794520547945207, | |
| "grad_norm": 15.0628080368042, | |
| "learning_rate": 5.224657534246575e-06, | |
| "loss": 0.3326, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 2.4863013698630136, | |
| "grad_norm": 11.597463607788086, | |
| "learning_rate": 5.156164383561644e-06, | |
| "loss": 0.3538, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 7.035673141479492, | |
| "learning_rate": 5.087671232876712e-06, | |
| "loss": 0.3049, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 24.056955337524414, | |
| "learning_rate": 5.019178082191781e-06, | |
| "loss": 0.4231, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.506849315068493, | |
| "grad_norm": 19.858354568481445, | |
| "learning_rate": 4.950684931506849e-06, | |
| "loss": 0.3143, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.5136986301369864, | |
| "grad_norm": 10.258318901062012, | |
| "learning_rate": 4.8821917808219174e-06, | |
| "loss": 0.3242, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 14.595016479492188, | |
| "learning_rate": 4.813698630136986e-06, | |
| "loss": 0.3642, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 2.5273972602739727, | |
| "grad_norm": 6.389834403991699, | |
| "learning_rate": 4.745205479452055e-06, | |
| "loss": 0.3086, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 2.5342465753424657, | |
| "grad_norm": 5.80775260925293, | |
| "learning_rate": 4.676712328767123e-06, | |
| "loss": 0.3142, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.541095890410959, | |
| "grad_norm": 4.96169376373291, | |
| "learning_rate": 4.608219178082192e-06, | |
| "loss": 0.3292, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 22.449113845825195, | |
| "learning_rate": 4.5397260273972606e-06, | |
| "loss": 0.4852, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.5547945205479454, | |
| "grad_norm": 23.349910736083984, | |
| "learning_rate": 4.471232876712328e-06, | |
| "loss": 0.3374, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 2.5616438356164384, | |
| "grad_norm": 9.687122344970703, | |
| "learning_rate": 4.402739726027397e-06, | |
| "loss": 0.3257, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 2.5684931506849313, | |
| "grad_norm": 8.347723007202148, | |
| "learning_rate": 4.334246575342466e-06, | |
| "loss": 0.2819, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 11.261799812316895, | |
| "learning_rate": 4.265753424657534e-06, | |
| "loss": 0.2929, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 2.5821917808219177, | |
| "grad_norm": 7.757129192352295, | |
| "learning_rate": 4.197260273972603e-06, | |
| "loss": 0.2746, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 2.589041095890411, | |
| "grad_norm": 13.809581756591797, | |
| "learning_rate": 4.1287671232876715e-06, | |
| "loss": 0.3915, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.595890410958904, | |
| "grad_norm": 15.997642517089844, | |
| "learning_rate": 4.060273972602739e-06, | |
| "loss": 0.3181, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 27.046567916870117, | |
| "learning_rate": 3.991780821917808e-06, | |
| "loss": 0.309, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "eval_exact_match": 85.57237464522233, | |
| "eval_f1": 92.37303362559862, | |
| "eval_runtime": 417.6236, | |
| "eval_samples_per_second": 25.31, | |
| "eval_steps_per_second": 1.583, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.6095890410958904, | |
| "grad_norm": 12.608302116394043, | |
| "learning_rate": 3.9232876712328765e-06, | |
| "loss": 0.3295, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 2.616438356164384, | |
| "grad_norm": 8.650862693786621, | |
| "learning_rate": 3.854794520547945e-06, | |
| "loss": 0.3119, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 2.6232876712328768, | |
| "grad_norm": 8.674386978149414, | |
| "learning_rate": 3.7863013698630138e-06, | |
| "loss": 0.2546, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 5.919064998626709, | |
| "learning_rate": 3.717808219178082e-06, | |
| "loss": 0.3697, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.636986301369863, | |
| "grad_norm": 7.016098499298096, | |
| "learning_rate": 3.6493150684931506e-06, | |
| "loss": 0.3302, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.643835616438356, | |
| "grad_norm": 8.310173034667969, | |
| "learning_rate": 3.5808219178082192e-06, | |
| "loss": 0.349, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 2.6506849315068495, | |
| "grad_norm": 8.628369331359863, | |
| "learning_rate": 3.5123287671232874e-06, | |
| "loss": 0.3424, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 13.100319862365723, | |
| "learning_rate": 3.443835616438356e-06, | |
| "loss": 0.3611, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 2.6643835616438354, | |
| "grad_norm": 5.1128692626953125, | |
| "learning_rate": 3.3753424657534247e-06, | |
| "loss": 0.3575, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 2.671232876712329, | |
| "grad_norm": 15.877631187438965, | |
| "learning_rate": 3.306849315068493e-06, | |
| "loss": 0.3791, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.678082191780822, | |
| "grad_norm": 14.39353084564209, | |
| "learning_rate": 3.2383561643835615e-06, | |
| "loss": 0.2931, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 6.899120807647705, | |
| "learning_rate": 3.16986301369863e-06, | |
| "loss": 0.3782, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 2.691780821917808, | |
| "grad_norm": 10.824689865112305, | |
| "learning_rate": 3.1013698630136988e-06, | |
| "loss": 0.3049, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 2.6986301369863015, | |
| "grad_norm": 18.822484970092773, | |
| "learning_rate": 3.032876712328767e-06, | |
| "loss": 0.4038, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 2.7054794520547945, | |
| "grad_norm": 15.976180076599121, | |
| "learning_rate": 2.9643835616438356e-06, | |
| "loss": 0.2828, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 27.864824295043945, | |
| "learning_rate": 2.8958904109589042e-06, | |
| "loss": 0.2974, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.719178082191781, | |
| "grad_norm": 16.251916885375977, | |
| "learning_rate": 2.8273972602739724e-06, | |
| "loss": 0.3124, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 2.7260273972602738, | |
| "grad_norm": 55.30453872680664, | |
| "learning_rate": 2.758904109589041e-06, | |
| "loss": 0.3211, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 2.732876712328767, | |
| "grad_norm": 15.253798484802246, | |
| "learning_rate": 2.6904109589041097e-06, | |
| "loss": 0.3369, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 67.18714904785156, | |
| "learning_rate": 2.623287671232877e-06, | |
| "loss": 0.376, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "eval_exact_match": 85.26963103122044, | |
| "eval_f1": 92.09863360357713, | |
| "eval_runtime": 417.6598, | |
| "eval_samples_per_second": 25.308, | |
| "eval_steps_per_second": 1.583, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.7465753424657535, | |
| "grad_norm": 11.747102737426758, | |
| "learning_rate": 2.5547945205479453e-06, | |
| "loss": 0.3359, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 2.7534246575342465, | |
| "grad_norm": 16.32077980041504, | |
| "learning_rate": 2.486301369863014e-06, | |
| "loss": 0.3529, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.76027397260274, | |
| "grad_norm": 40.523311614990234, | |
| "learning_rate": 2.4178082191780826e-06, | |
| "loss": 0.3346, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 16.648462295532227, | |
| "learning_rate": 2.3493150684931508e-06, | |
| "loss": 0.3167, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 2.7739726027397262, | |
| "grad_norm": 13.732590675354004, | |
| "learning_rate": 2.2808219178082194e-06, | |
| "loss": 0.3431, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 2.780821917808219, | |
| "grad_norm": 10.560003280639648, | |
| "learning_rate": 2.2123287671232876e-06, | |
| "loss": 0.3688, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 2.787671232876712, | |
| "grad_norm": 6.171405792236328, | |
| "learning_rate": 2.1438356164383562e-06, | |
| "loss": 0.3314, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 11.62563419342041, | |
| "learning_rate": 2.0753424657534244e-06, | |
| "loss": 0.3371, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.8013698630136985, | |
| "grad_norm": 12.698870658874512, | |
| "learning_rate": 2.006849315068493e-06, | |
| "loss": 0.3429, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 2.808219178082192, | |
| "grad_norm": 12.162155151367188, | |
| "learning_rate": 1.9383561643835617e-06, | |
| "loss": 0.3383, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.815068493150685, | |
| "grad_norm": 11.249317169189453, | |
| "learning_rate": 1.8698630136986303e-06, | |
| "loss": 0.3228, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 15.184226989746094, | |
| "learning_rate": 1.8013698630136987e-06, | |
| "loss": 0.3299, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 2.828767123287671, | |
| "grad_norm": 21.128114700317383, | |
| "learning_rate": 1.7328767123287671e-06, | |
| "loss": 0.2959, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 2.8356164383561646, | |
| "grad_norm": 18.958660125732422, | |
| "learning_rate": 1.6643835616438358e-06, | |
| "loss": 0.2943, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.8424657534246576, | |
| "grad_norm": 11.99002456665039, | |
| "learning_rate": 1.5958904109589042e-06, | |
| "loss": 0.3151, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 41.37422561645508, | |
| "learning_rate": 1.5273972602739726e-06, | |
| "loss": 0.3659, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 2.856164383561644, | |
| "grad_norm": 63.95098876953125, | |
| "learning_rate": 1.4589041095890412e-06, | |
| "loss": 0.3366, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 2.863013698630137, | |
| "grad_norm": 7.210549354553223, | |
| "learning_rate": 1.3904109589041096e-06, | |
| "loss": 0.3207, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 2.8698630136986303, | |
| "grad_norm": 21.405946731567383, | |
| "learning_rate": 1.3219178082191783e-06, | |
| "loss": 0.314, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 40.97032165527344, | |
| "learning_rate": 1.2534246575342467e-06, | |
| "loss": 0.3598, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "eval_exact_match": 85.58183538315988, | |
| "eval_f1": 92.20238705674495, | |
| "eval_runtime": 416.9361, | |
| "eval_samples_per_second": 25.352, | |
| "eval_steps_per_second": 1.585, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.883561643835616, | |
| "grad_norm": 28.014732360839844, | |
| "learning_rate": 1.184931506849315e-06, | |
| "loss": 0.2415, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 2.8904109589041096, | |
| "grad_norm": 28.053062438964844, | |
| "learning_rate": 1.1164383561643837e-06, | |
| "loss": 0.3468, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 2.897260273972603, | |
| "grad_norm": 7.82937479019165, | |
| "learning_rate": 1.0479452054794521e-06, | |
| "loss": 0.3102, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 9.58170223236084, | |
| "learning_rate": 9.794520547945205e-07, | |
| "loss": 0.3115, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 2.910958904109589, | |
| "grad_norm": 15.694490432739258, | |
| "learning_rate": 9.109589041095891e-07, | |
| "loss": 0.2848, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 2.9178082191780823, | |
| "grad_norm": 11.358589172363281, | |
| "learning_rate": 8.424657534246575e-07, | |
| "loss": 0.3353, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.9246575342465753, | |
| "grad_norm": 42.85386276245117, | |
| "learning_rate": 7.73972602739726e-07, | |
| "loss": 0.271, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 4.216646194458008, | |
| "learning_rate": 7.054794520547945e-07, | |
| "loss": 0.352, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 2.9383561643835616, | |
| "grad_norm": 18.55265235900879, | |
| "learning_rate": 6.36986301369863e-07, | |
| "loss": 0.3376, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 2.9452054794520546, | |
| "grad_norm": 10.340827941894531, | |
| "learning_rate": 5.684931506849316e-07, | |
| "loss": 0.2902, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.952054794520548, | |
| "grad_norm": 13.433025360107422, | |
| "learning_rate": 5e-07, | |
| "loss": 0.3152, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 26.042661666870117, | |
| "learning_rate": 4.315068493150685e-07, | |
| "loss": 0.3431, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.9657534246575343, | |
| "grad_norm": 19.630699157714844, | |
| "learning_rate": 3.63013698630137e-07, | |
| "loss": 0.296, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 2.9726027397260273, | |
| "grad_norm": 25.595293045043945, | |
| "learning_rate": 2.945205479452055e-07, | |
| "loss": 0.352, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 2.9794520547945207, | |
| "grad_norm": 11.616922378540039, | |
| "learning_rate": 2.2602739726027398e-07, | |
| "loss": 0.3358, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 5.852312088012695, | |
| "learning_rate": 1.5753424657534248e-07, | |
| "loss": 0.3017, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 2.993150684931507, | |
| "grad_norm": 30.67576789855957, | |
| "learning_rate": 8.904109589041097e-08, | |
| "loss": 0.3171, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 5.107680320739746, | |
| "learning_rate": 2.0547945205479452e-08, | |
| "loss": 0.3315, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 21900, | |
| "total_flos": 2.006738209660207e+18, | |
| "train_loss": 0.6596583454358523, | |
| "train_runtime": 64667.9553, | |
| "train_samples_per_second": 4.064, | |
| "train_steps_per_second": 0.339 | |
| } | |
| ], | |
| "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 | |
| } | |