| { | |
| "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": 341.1622619628906, | |
| "learning_rate": 2.9945205479452056e-05, | |
| "loss": 4.9651, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0136986301369863, | |
| "grad_norm": 47.20112609863281, | |
| "learning_rate": 2.9876712328767122e-05, | |
| "loss": 2.5643, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02054794520547945, | |
| "grad_norm": 136.01480102539062, | |
| "learning_rate": 2.980821917808219e-05, | |
| "loss": 2.492, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 34.74509048461914, | |
| "learning_rate": 2.973972602739726e-05, | |
| "loss": 1.8855, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03424657534246575, | |
| "grad_norm": 415.1113586425781, | |
| "learning_rate": 2.967123287671233e-05, | |
| "loss": 1.8193, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0410958904109589, | |
| "grad_norm": 46.783470153808594, | |
| "learning_rate": 2.9602739726027396e-05, | |
| "loss": 1.6777, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04794520547945205, | |
| "grad_norm": 49.481300354003906, | |
| "learning_rate": 2.9534246575342466e-05, | |
| "loss": 1.6324, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 54.750755310058594, | |
| "learning_rate": 2.9467123287671234e-05, | |
| "loss": 1.4367, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06164383561643835, | |
| "grad_norm": 49.342838287353516, | |
| "learning_rate": 2.93986301369863e-05, | |
| "loss": 1.4989, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0684931506849315, | |
| "grad_norm": 11.364151000976562, | |
| "learning_rate": 2.933013698630137e-05, | |
| "loss": 1.5344, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.07534246575342465, | |
| "grad_norm": 77.4140396118164, | |
| "learning_rate": 2.926164383561644e-05, | |
| "loss": 1.51, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 20.35340690612793, | |
| "learning_rate": 2.919315068493151e-05, | |
| "loss": 1.4193, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.08904109589041095, | |
| "grad_norm": 45.406463623046875, | |
| "learning_rate": 2.9124657534246575e-05, | |
| "loss": 1.3835, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.0958904109589041, | |
| "grad_norm": 21.045604705810547, | |
| "learning_rate": 2.9056164383561644e-05, | |
| "loss": 1.2977, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.10273972602739725, | |
| "grad_norm": 45.87715530395508, | |
| "learning_rate": 2.8987671232876714e-05, | |
| "loss": 1.3941, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 50.6860466003418, | |
| "learning_rate": 2.8919178082191783e-05, | |
| "loss": 1.4132, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.11643835616438356, | |
| "grad_norm": 35.442527770996094, | |
| "learning_rate": 2.885068493150685e-05, | |
| "loss": 1.3313, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1232876712328767, | |
| "grad_norm": 37.033992767333984, | |
| "learning_rate": 2.878219178082192e-05, | |
| "loss": 1.2819, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.13013698630136986, | |
| "grad_norm": 21.093502044677734, | |
| "learning_rate": 2.871369863013699e-05, | |
| "loss": 1.2457, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 19.11823081970215, | |
| "learning_rate": 2.8645205479452058e-05, | |
| "loss": 1.2022, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.14383561643835616, | |
| "grad_norm": 24.3985538482666, | |
| "learning_rate": 2.8576712328767124e-05, | |
| "loss": 1.2132, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1506849315068493, | |
| "grad_norm": 29.999114990234375, | |
| "learning_rate": 2.8508219178082194e-05, | |
| "loss": 1.4006, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.15753424657534246, | |
| "grad_norm": 61.78759002685547, | |
| "learning_rate": 2.8439726027397263e-05, | |
| "loss": 1.4269, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 28.08926773071289, | |
| "learning_rate": 2.8371232876712332e-05, | |
| "loss": 1.2884, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.17123287671232876, | |
| "grad_norm": 31.31874656677246, | |
| "learning_rate": 2.8302739726027395e-05, | |
| "loss": 1.1787, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1780821917808219, | |
| "grad_norm": 27.130922317504883, | |
| "learning_rate": 2.8234246575342465e-05, | |
| "loss": 1.1857, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.18493150684931506, | |
| "grad_norm": 18.046369552612305, | |
| "learning_rate": 2.8165753424657534e-05, | |
| "loss": 1.1351, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 71.45783996582031, | |
| "learning_rate": 2.8097260273972604e-05, | |
| "loss": 1.196, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.19863013698630136, | |
| "grad_norm": 12.047041893005371, | |
| "learning_rate": 2.803013698630137e-05, | |
| "loss": 1.3001, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.2054794520547945, | |
| "grad_norm": 15.405315399169922, | |
| "learning_rate": 2.796164383561644e-05, | |
| "loss": 1.147, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21232876712328766, | |
| "grad_norm": 25.710603713989258, | |
| "learning_rate": 2.7893150684931508e-05, | |
| "loss": 1.1912, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 52.69472885131836, | |
| "learning_rate": 2.7824657534246574e-05, | |
| "loss": 1.0702, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.22602739726027396, | |
| "grad_norm": 19.084321975708008, | |
| "learning_rate": 2.7756164383561643e-05, | |
| "loss": 1.2096, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.2328767123287671, | |
| "grad_norm": 18.86844253540039, | |
| "learning_rate": 2.7687671232876713e-05, | |
| "loss": 1.0732, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.23972602739726026, | |
| "grad_norm": 20.677623748779297, | |
| "learning_rate": 2.7619178082191782e-05, | |
| "loss": 1.0515, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 22.844743728637695, | |
| "learning_rate": 2.755068493150685e-05, | |
| "loss": 1.0897, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.2534246575342466, | |
| "grad_norm": 22.20302963256836, | |
| "learning_rate": 2.7482191780821918e-05, | |
| "loss": 1.0234, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.2602739726027397, | |
| "grad_norm": 21.382713317871094, | |
| "learning_rate": 2.7413698630136987e-05, | |
| "loss": 1.0679, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2671232876712329, | |
| "grad_norm": 38.08824920654297, | |
| "learning_rate": 2.7345205479452057e-05, | |
| "loss": 1.0237, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 64.6864013671875, | |
| "learning_rate": 2.7276712328767123e-05, | |
| "loss": 0.9552, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "eval_exact_match": 81.20151371807, | |
| "eval_f1": 89.11978143319536, | |
| "eval_runtime": 373.3956, | |
| "eval_samples_per_second": 28.308, | |
| "eval_steps_per_second": 1.77, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2808219178082192, | |
| "grad_norm": 35.17247009277344, | |
| "learning_rate": 2.7208219178082192e-05, | |
| "loss": 1.0625, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2876712328767123, | |
| "grad_norm": 14.59727954864502, | |
| "learning_rate": 2.7139726027397262e-05, | |
| "loss": 0.896, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2945205479452055, | |
| "grad_norm": 17.56501579284668, | |
| "learning_rate": 2.707123287671233e-05, | |
| "loss": 0.9751, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 39.82829666137695, | |
| "learning_rate": 2.7002739726027397e-05, | |
| "loss": 0.9483, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3082191780821918, | |
| "grad_norm": 52.49971008300781, | |
| "learning_rate": 2.6934246575342467e-05, | |
| "loss": 0.9356, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3150684931506849, | |
| "grad_norm": 12.524872779846191, | |
| "learning_rate": 2.6865753424657536e-05, | |
| "loss": 1.0543, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3219178082191781, | |
| "grad_norm": 26.332027435302734, | |
| "learning_rate": 2.6797260273972606e-05, | |
| "loss": 0.9407, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 56.26033020019531, | |
| "learning_rate": 2.6728767123287672e-05, | |
| "loss": 1.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3356164383561644, | |
| "grad_norm": 33.63979721069336, | |
| "learning_rate": 2.666027397260274e-05, | |
| "loss": 1.0453, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 11.63798713684082, | |
| "learning_rate": 2.659178082191781e-05, | |
| "loss": 1.0912, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3493150684931507, | |
| "grad_norm": 48.529258728027344, | |
| "learning_rate": 2.652328767123288e-05, | |
| "loss": 1.0498, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 11.467239379882812, | |
| "learning_rate": 2.6454794520547943e-05, | |
| "loss": 0.9941, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.363013698630137, | |
| "grad_norm": 18.566442489624023, | |
| "learning_rate": 2.6386301369863013e-05, | |
| "loss": 0.9262, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.3698630136986301, | |
| "grad_norm": 6.612974166870117, | |
| "learning_rate": 2.6317808219178082e-05, | |
| "loss": 0.9856, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3767123287671233, | |
| "grad_norm": 14.778943061828613, | |
| "learning_rate": 2.624931506849315e-05, | |
| "loss": 0.9819, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 23.16988754272461, | |
| "learning_rate": 2.6180821917808218e-05, | |
| "loss": 1.0082, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.3904109589041096, | |
| "grad_norm": 18.585800170898438, | |
| "learning_rate": 2.6112328767123287e-05, | |
| "loss": 0.9695, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3972602739726027, | |
| "grad_norm": 28.559650421142578, | |
| "learning_rate": 2.6043835616438357e-05, | |
| "loss": 0.961, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4041095890410959, | |
| "grad_norm": 53.306453704833984, | |
| "learning_rate": 2.5975342465753426e-05, | |
| "loss": 0.9382, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 21.294361114501953, | |
| "learning_rate": 2.5906849315068492e-05, | |
| "loss": 0.9482, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4178082191780822, | |
| "grad_norm": 74.29987335205078, | |
| "learning_rate": 2.583835616438356e-05, | |
| "loss": 0.969, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.4246575342465753, | |
| "grad_norm": 38.378807067871094, | |
| "learning_rate": 2.576986301369863e-05, | |
| "loss": 0.869, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4315068493150685, | |
| "grad_norm": 20.193374633789062, | |
| "learning_rate": 2.57013698630137e-05, | |
| "loss": 0.9143, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 22.409589767456055, | |
| "learning_rate": 2.5632876712328767e-05, | |
| "loss": 0.993, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.4452054794520548, | |
| "grad_norm": 11.131108283996582, | |
| "learning_rate": 2.5564383561643836e-05, | |
| "loss": 1.0915, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.4520547945205479, | |
| "grad_norm": 20.399301528930664, | |
| "learning_rate": 2.5495890410958906e-05, | |
| "loss": 0.9514, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4589041095890411, | |
| "grad_norm": 19.632505416870117, | |
| "learning_rate": 2.5427397260273975e-05, | |
| "loss": 0.9821, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 9.592618942260742, | |
| "learning_rate": 2.535890410958904e-05, | |
| "loss": 0.9956, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.4726027397260274, | |
| "grad_norm": 34.0085334777832, | |
| "learning_rate": 2.529041095890411e-05, | |
| "loss": 0.9384, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.4794520547945205, | |
| "grad_norm": 29.698089599609375, | |
| "learning_rate": 2.522191780821918e-05, | |
| "loss": 0.8727, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4863013698630137, | |
| "grad_norm": 4.650140762329102, | |
| "learning_rate": 2.515342465753425e-05, | |
| "loss": 0.9278, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 47.80159378051758, | |
| "learning_rate": 2.5084931506849316e-05, | |
| "loss": 0.9, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 27.069276809692383, | |
| "learning_rate": 2.5016438356164385e-05, | |
| "loss": 0.91, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 21.940160751342773, | |
| "learning_rate": 2.4947945205479455e-05, | |
| "loss": 1.159, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5136986301369864, | |
| "grad_norm": 34.344852447509766, | |
| "learning_rate": 2.4879452054794524e-05, | |
| "loss": 0.9166, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 45.67475128173828, | |
| "learning_rate": 2.481095890410959e-05, | |
| "loss": 0.947, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5273972602739726, | |
| "grad_norm": 45.106163024902344, | |
| "learning_rate": 2.474246575342466e-05, | |
| "loss": 0.9908, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.5342465753424658, | |
| "grad_norm": 20.590816497802734, | |
| "learning_rate": 2.4673972602739726e-05, | |
| "loss": 0.9051, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.541095890410959, | |
| "grad_norm": 27.564393997192383, | |
| "learning_rate": 2.4605479452054795e-05, | |
| "loss": 0.9555, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 7.4184956550598145, | |
| "learning_rate": 2.453698630136986e-05, | |
| "loss": 0.9574, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "eval_exact_match": 82.3841059602649, | |
| "eval_f1": 90.0771836682692, | |
| "eval_runtime": 374.3076, | |
| "eval_samples_per_second": 28.239, | |
| "eval_steps_per_second": 1.766, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5547945205479452, | |
| "grad_norm": 17.022262573242188, | |
| "learning_rate": 2.446849315068493e-05, | |
| "loss": 0.986, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.5616438356164384, | |
| "grad_norm": 22.404705047607422, | |
| "learning_rate": 2.44e-05, | |
| "loss": 0.956, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.5684931506849316, | |
| "grad_norm": 22.185340881347656, | |
| "learning_rate": 2.433150684931507e-05, | |
| "loss": 0.8592, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 38.17957305908203, | |
| "learning_rate": 2.4263013698630136e-05, | |
| "loss": 0.9033, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5821917808219178, | |
| "grad_norm": 19.095006942749023, | |
| "learning_rate": 2.4194520547945205e-05, | |
| "loss": 0.9732, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.589041095890411, | |
| "grad_norm": 7.960810661315918, | |
| "learning_rate": 2.4126027397260275e-05, | |
| "loss": 0.9306, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5958904109589042, | |
| "grad_norm": 14.839760780334473, | |
| "learning_rate": 2.4057534246575344e-05, | |
| "loss": 0.8492, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 14.839119911193848, | |
| "learning_rate": 2.398904109589041e-05, | |
| "loss": 0.8939, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6095890410958904, | |
| "grad_norm": 9.723929405212402, | |
| "learning_rate": 2.392054794520548e-05, | |
| "loss": 1.0003, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.6164383561643836, | |
| "grad_norm": 26.83091163635254, | |
| "learning_rate": 2.385205479452055e-05, | |
| "loss": 0.8949, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6232876712328768, | |
| "grad_norm": 12.107084274291992, | |
| "learning_rate": 2.378356164383562e-05, | |
| "loss": 0.9106, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 8.986040115356445, | |
| "learning_rate": 2.3715068493150685e-05, | |
| "loss": 0.9329, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.636986301369863, | |
| "grad_norm": 6.278331756591797, | |
| "learning_rate": 2.3646575342465754e-05, | |
| "loss": 0.9002, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.6438356164383562, | |
| "grad_norm": 5.5899434089660645, | |
| "learning_rate": 2.3578082191780824e-05, | |
| "loss": 0.8955, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.6506849315068494, | |
| "grad_norm": 37.710784912109375, | |
| "learning_rate": 2.3509589041095893e-05, | |
| "loss": 0.9641, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 18.405412673950195, | |
| "learning_rate": 2.344109589041096e-05, | |
| "loss": 0.7766, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.6643835616438356, | |
| "grad_norm": 18.17953872680664, | |
| "learning_rate": 2.337260273972603e-05, | |
| "loss": 1.0796, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.6712328767123288, | |
| "grad_norm": 20.769283294677734, | |
| "learning_rate": 2.3304109589041098e-05, | |
| "loss": 0.9752, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.678082191780822, | |
| "grad_norm": 19.938074111938477, | |
| "learning_rate": 2.3235616438356168e-05, | |
| "loss": 0.875, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 26.178194046020508, | |
| "learning_rate": 2.3167123287671234e-05, | |
| "loss": 0.9007, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6917808219178082, | |
| "grad_norm": 21.135181427001953, | |
| "learning_rate": 2.3098630136986303e-05, | |
| "loss": 0.8955, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.6986301369863014, | |
| "grad_norm": 24.87206268310547, | |
| "learning_rate": 2.3030136986301373e-05, | |
| "loss": 0.9886, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7054794520547946, | |
| "grad_norm": 16.213960647583008, | |
| "learning_rate": 2.296164383561644e-05, | |
| "loss": 1.0165, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 12.362586975097656, | |
| "learning_rate": 2.2893150684931505e-05, | |
| "loss": 1.0008, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.7191780821917808, | |
| "grad_norm": 1259.0946044921875, | |
| "learning_rate": 2.2824657534246574e-05, | |
| "loss": 0.9962, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.726027397260274, | |
| "grad_norm": 16.600799560546875, | |
| "learning_rate": 2.2756164383561644e-05, | |
| "loss": 0.8839, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.7328767123287672, | |
| "grad_norm": 33.82933044433594, | |
| "learning_rate": 2.2687671232876713e-05, | |
| "loss": 0.9854, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 15.420291900634766, | |
| "learning_rate": 2.261917808219178e-05, | |
| "loss": 0.9803, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.7465753424657534, | |
| "grad_norm": 32.19168472290039, | |
| "learning_rate": 2.255068493150685e-05, | |
| "loss": 0.9143, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "grad_norm": 32.13875198364258, | |
| "learning_rate": 2.248219178082192e-05, | |
| "loss": 1.0092, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7602739726027398, | |
| "grad_norm": 12.64287281036377, | |
| "learning_rate": 2.2413698630136988e-05, | |
| "loss": 0.9426, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 12.699542999267578, | |
| "learning_rate": 2.2345205479452054e-05, | |
| "loss": 0.8253, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.773972602739726, | |
| "grad_norm": 8.844592094421387, | |
| "learning_rate": 2.2276712328767123e-05, | |
| "loss": 0.8658, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.7808219178082192, | |
| "grad_norm": 28.156341552734375, | |
| "learning_rate": 2.2208219178082193e-05, | |
| "loss": 0.9257, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.7876712328767124, | |
| "grad_norm": 12.614666938781738, | |
| "learning_rate": 2.2139726027397262e-05, | |
| "loss": 0.8974, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 16.726621627807617, | |
| "learning_rate": 2.207123287671233e-05, | |
| "loss": 0.9291, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.8013698630136986, | |
| "grad_norm": 10.39370346069336, | |
| "learning_rate": 2.2002739726027398e-05, | |
| "loss": 0.9179, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.8082191780821918, | |
| "grad_norm": 21.894121170043945, | |
| "learning_rate": 2.1934246575342467e-05, | |
| "loss": 0.8794, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.815068493150685, | |
| "grad_norm": 24.440290451049805, | |
| "learning_rate": 2.1865753424657537e-05, | |
| "loss": 0.8238, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 12.202630996704102, | |
| "learning_rate": 2.1797260273972603e-05, | |
| "loss": 0.8661, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "eval_exact_match": 83.85052034058657, | |
| "eval_f1": 90.9800111426571, | |
| "eval_runtime": 373.5608, | |
| "eval_samples_per_second": 28.295, | |
| "eval_steps_per_second": 1.769, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8287671232876712, | |
| "grad_norm": 32.982906341552734, | |
| "learning_rate": 2.1728767123287672e-05, | |
| "loss": 0.8246, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.8356164383561644, | |
| "grad_norm": 18.63083839416504, | |
| "learning_rate": 2.1660273972602742e-05, | |
| "loss": 0.7897, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.8424657534246576, | |
| "grad_norm": 41.52643585205078, | |
| "learning_rate": 2.159178082191781e-05, | |
| "loss": 0.9554, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 7.247984409332275, | |
| "learning_rate": 2.1523287671232877e-05, | |
| "loss": 0.8986, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.8561643835616438, | |
| "grad_norm": 35.788936614990234, | |
| "learning_rate": 2.1454794520547947e-05, | |
| "loss": 0.8827, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.863013698630137, | |
| "grad_norm": 14.294588088989258, | |
| "learning_rate": 2.1386301369863016e-05, | |
| "loss": 0.8467, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.8698630136986302, | |
| "grad_norm": 7.7929253578186035, | |
| "learning_rate": 2.1317808219178086e-05, | |
| "loss": 0.8478, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 22.97696876525879, | |
| "learning_rate": 2.1249315068493152e-05, | |
| "loss": 0.828, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.8835616438356164, | |
| "grad_norm": 9.756913185119629, | |
| "learning_rate": 2.1180821917808218e-05, | |
| "loss": 0.8868, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.8904109589041096, | |
| "grad_norm": 10.705896377563477, | |
| "learning_rate": 2.1112328767123287e-05, | |
| "loss": 0.8307, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8972602739726028, | |
| "grad_norm": 10.845613479614258, | |
| "learning_rate": 2.1043835616438357e-05, | |
| "loss": 0.7624, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 19.542699813842773, | |
| "learning_rate": 2.0975342465753423e-05, | |
| "loss": 0.8325, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.910958904109589, | |
| "grad_norm": 13.544864654541016, | |
| "learning_rate": 2.0906849315068493e-05, | |
| "loss": 0.851, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.9178082191780822, | |
| "grad_norm": 6.318390369415283, | |
| "learning_rate": 2.0838356164383562e-05, | |
| "loss": 0.8091, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.9246575342465754, | |
| "grad_norm": 9.816180229187012, | |
| "learning_rate": 2.076986301369863e-05, | |
| "loss": 0.8832, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 15.709238052368164, | |
| "learning_rate": 2.0701369863013698e-05, | |
| "loss": 0.808, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.9383561643835616, | |
| "grad_norm": 14.45374870300293, | |
| "learning_rate": 2.0632876712328767e-05, | |
| "loss": 0.8885, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.9452054794520548, | |
| "grad_norm": 17.111806869506836, | |
| "learning_rate": 2.0564383561643836e-05, | |
| "loss": 0.7972, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.952054794520548, | |
| "grad_norm": 20.28083610534668, | |
| "learning_rate": 2.0495890410958906e-05, | |
| "loss": 0.8979, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 27.66487693786621, | |
| "learning_rate": 2.0427397260273972e-05, | |
| "loss": 0.8672, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9657534246575342, | |
| "grad_norm": 12.894664764404297, | |
| "learning_rate": 2.035890410958904e-05, | |
| "loss": 0.9069, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.9726027397260274, | |
| "grad_norm": 30.51468849182129, | |
| "learning_rate": 2.029041095890411e-05, | |
| "loss": 0.8747, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.9794520547945206, | |
| "grad_norm": 39.097129821777344, | |
| "learning_rate": 2.022191780821918e-05, | |
| "loss": 0.8729, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 32.58139419555664, | |
| "learning_rate": 2.0153424657534247e-05, | |
| "loss": 0.8815, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.9931506849315068, | |
| "grad_norm": 56.9384651184082, | |
| "learning_rate": 2.0084931506849316e-05, | |
| "loss": 0.8372, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 41.42207717895508, | |
| "learning_rate": 2.0016438356164386e-05, | |
| "loss": 0.7201, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.0068493150684932, | |
| "grad_norm": 12.120720863342285, | |
| "learning_rate": 1.9947945205479455e-05, | |
| "loss": 0.6208, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 13.257436752319336, | |
| "learning_rate": 1.987945205479452e-05, | |
| "loss": 0.6044, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.0205479452054795, | |
| "grad_norm": 13.67995548248291, | |
| "learning_rate": 1.981095890410959e-05, | |
| "loss": 0.6329, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.0273972602739727, | |
| "grad_norm": 9.850571632385254, | |
| "learning_rate": 1.974246575342466e-05, | |
| "loss": 0.6957, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0342465753424657, | |
| "grad_norm": 10.688873291015625, | |
| "learning_rate": 1.9673972602739726e-05, | |
| "loss": 0.6464, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 8.531394004821777, | |
| "learning_rate": 1.9605479452054796e-05, | |
| "loss": 0.6109, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.047945205479452, | |
| "grad_norm": 10.75566577911377, | |
| "learning_rate": 1.9536986301369865e-05, | |
| "loss": 0.6616, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.0547945205479452, | |
| "grad_norm": 8.011566162109375, | |
| "learning_rate": 1.9468493150684935e-05, | |
| "loss": 0.5976, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.0616438356164384, | |
| "grad_norm": 20.438060760498047, | |
| "learning_rate": 1.9399999999999997e-05, | |
| "loss": 0.6787, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 7.955737113952637, | |
| "learning_rate": 1.9331506849315067e-05, | |
| "loss": 0.5945, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.0753424657534247, | |
| "grad_norm": 11.127422332763672, | |
| "learning_rate": 1.9263013698630136e-05, | |
| "loss": 0.6308, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.0821917808219177, | |
| "grad_norm": 38.58836364746094, | |
| "learning_rate": 1.9194520547945206e-05, | |
| "loss": 0.6206, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.0890410958904109, | |
| "grad_norm": 32.406192779541016, | |
| "learning_rate": 1.9126027397260272e-05, | |
| "loss": 0.701, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 10.007853507995605, | |
| "learning_rate": 1.905753424657534e-05, | |
| "loss": 0.5769, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 85.44938505203406, | |
| "eval_f1": 92.14833349343365, | |
| "eval_runtime": 374.398, | |
| "eval_samples_per_second": 28.232, | |
| "eval_steps_per_second": 1.766, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.1027397260273972, | |
| "grad_norm": 28.756309509277344, | |
| "learning_rate": 1.898904109589041e-05, | |
| "loss": 0.6004, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.1095890410958904, | |
| "grad_norm": 7.047467231750488, | |
| "learning_rate": 1.892054794520548e-05, | |
| "loss": 0.6181, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.1164383561643836, | |
| "grad_norm": 8.658169746398926, | |
| "learning_rate": 1.8852054794520546e-05, | |
| "loss": 0.7291, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 13.879542350769043, | |
| "learning_rate": 1.8783561643835616e-05, | |
| "loss": 0.6838, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.13013698630137, | |
| "grad_norm": 22.860614776611328, | |
| "learning_rate": 1.8715068493150685e-05, | |
| "loss": 0.608, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.1369863013698631, | |
| "grad_norm": 16.130298614501953, | |
| "learning_rate": 1.8646575342465755e-05, | |
| "loss": 0.6373, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.143835616438356, | |
| "grad_norm": 10.36241340637207, | |
| "learning_rate": 1.857808219178082e-05, | |
| "loss": 0.5687, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 16.204782485961914, | |
| "learning_rate": 1.850958904109589e-05, | |
| "loss": 0.5777, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.1575342465753424, | |
| "grad_norm": 12.540966033935547, | |
| "learning_rate": 1.844109589041096e-05, | |
| "loss": 0.6184, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.1643835616438356, | |
| "grad_norm": 10.892577171325684, | |
| "learning_rate": 1.837260273972603e-05, | |
| "loss": 0.5981, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.1712328767123288, | |
| "grad_norm": 35.12110137939453, | |
| "learning_rate": 1.8304109589041095e-05, | |
| "loss": 0.6206, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 59.20553970336914, | |
| "learning_rate": 1.8235616438356165e-05, | |
| "loss": 0.6139, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.1849315068493151, | |
| "grad_norm": 10.697203636169434, | |
| "learning_rate": 1.8167123287671234e-05, | |
| "loss": 0.6646, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.191780821917808, | |
| "grad_norm": 14.472150802612305, | |
| "learning_rate": 1.8098630136986304e-05, | |
| "loss": 0.6595, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.1986301369863013, | |
| "grad_norm": 72.91890716552734, | |
| "learning_rate": 1.803013698630137e-05, | |
| "loss": 0.6285, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 17.10064697265625, | |
| "learning_rate": 1.796164383561644e-05, | |
| "loss": 0.6663, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.2123287671232876, | |
| "grad_norm": 20.469289779663086, | |
| "learning_rate": 1.789315068493151e-05, | |
| "loss": 0.7494, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.2191780821917808, | |
| "grad_norm": 26.936960220336914, | |
| "learning_rate": 1.7824657534246578e-05, | |
| "loss": 0.709, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.226027397260274, | |
| "grad_norm": 46.11455535888672, | |
| "learning_rate": 1.7756164383561644e-05, | |
| "loss": 0.7078, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 19.011287689208984, | |
| "learning_rate": 1.7687671232876714e-05, | |
| "loss": 0.6223, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2397260273972603, | |
| "grad_norm": 14.254953384399414, | |
| "learning_rate": 1.7620547945205482e-05, | |
| "loss": 0.7151, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.2465753424657535, | |
| "grad_norm": 13.175114631652832, | |
| "learning_rate": 1.7552054794520545e-05, | |
| "loss": 0.6855, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.2534246575342465, | |
| "grad_norm": 7.520362854003906, | |
| "learning_rate": 1.7483561643835615e-05, | |
| "loss": 0.6629, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 12.215356826782227, | |
| "learning_rate": 1.7415068493150684e-05, | |
| "loss": 0.5965, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.2671232876712328, | |
| "grad_norm": 12.193058013916016, | |
| "learning_rate": 1.7346575342465754e-05, | |
| "loss": 0.6565, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.273972602739726, | |
| "grad_norm": 32.37642288208008, | |
| "learning_rate": 1.727808219178082e-05, | |
| "loss": 0.5647, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.2808219178082192, | |
| "grad_norm": 31.779830932617188, | |
| "learning_rate": 1.720958904109589e-05, | |
| "loss": 0.6011, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 21.49200439453125, | |
| "learning_rate": 1.714109589041096e-05, | |
| "loss": 0.6918, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.2945205479452055, | |
| "grad_norm": 5.468780517578125, | |
| "learning_rate": 1.7072602739726028e-05, | |
| "loss": 0.5157, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.3013698630136985, | |
| "grad_norm": 28.108797073364258, | |
| "learning_rate": 1.7004109589041094e-05, | |
| "loss": 0.6383, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.308219178082192, | |
| "grad_norm": 23.633098602294922, | |
| "learning_rate": 1.6935616438356164e-05, | |
| "loss": 0.6621, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 22.421751022338867, | |
| "learning_rate": 1.6867123287671233e-05, | |
| "loss": 0.5817, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.321917808219178, | |
| "grad_norm": 15.235234260559082, | |
| "learning_rate": 1.6798630136986303e-05, | |
| "loss": 0.6597, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.3287671232876712, | |
| "grad_norm": 35.48225402832031, | |
| "learning_rate": 1.673013698630137e-05, | |
| "loss": 0.579, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.3356164383561644, | |
| "grad_norm": 15.125396728515625, | |
| "learning_rate": 1.6661643835616438e-05, | |
| "loss": 0.6055, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 18.926666259765625, | |
| "learning_rate": 1.6593150684931508e-05, | |
| "loss": 0.6323, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.3493150684931507, | |
| "grad_norm": 8.361820220947266, | |
| "learning_rate": 1.6524657534246577e-05, | |
| "loss": 0.6317, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.356164383561644, | |
| "grad_norm": 26.376338958740234, | |
| "learning_rate": 1.6456164383561643e-05, | |
| "loss": 0.5951, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.3630136986301369, | |
| "grad_norm": 27.7407169342041, | |
| "learning_rate": 1.6387671232876713e-05, | |
| "loss": 0.6338, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 26.390640258789062, | |
| "learning_rate": 1.6319178082191782e-05, | |
| "loss": 0.6335, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "eval_exact_match": 84.70198675496688, | |
| "eval_f1": 91.62789345682066, | |
| "eval_runtime": 374.2131, | |
| "eval_samples_per_second": 28.246, | |
| "eval_steps_per_second": 1.766, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.3767123287671232, | |
| "grad_norm": 5.783433437347412, | |
| "learning_rate": 1.625068493150685e-05, | |
| "loss": 0.6771, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.3835616438356164, | |
| "grad_norm": 9.566642761230469, | |
| "learning_rate": 1.6182191780821918e-05, | |
| "loss": 0.498, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.3904109589041096, | |
| "grad_norm": 9.79062271118164, | |
| "learning_rate": 1.6113698630136987e-05, | |
| "loss": 0.6378, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 12.324602127075195, | |
| "learning_rate": 1.6045205479452057e-05, | |
| "loss": 0.5836, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.404109589041096, | |
| "grad_norm": 20.825387954711914, | |
| "learning_rate": 1.5976712328767126e-05, | |
| "loss": 0.5341, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.410958904109589, | |
| "grad_norm": 5.5393829345703125, | |
| "learning_rate": 1.5908219178082192e-05, | |
| "loss": 0.4899, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.4178082191780823, | |
| "grad_norm": 16.400829315185547, | |
| "learning_rate": 1.5839726027397258e-05, | |
| "loss": 0.7078, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 4.527399063110352, | |
| "learning_rate": 1.5771232876712328e-05, | |
| "loss": 0.6155, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.4315068493150684, | |
| "grad_norm": 27.31574058532715, | |
| "learning_rate": 1.5702739726027397e-05, | |
| "loss": 0.6125, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.4383561643835616, | |
| "grad_norm": 10.297114372253418, | |
| "learning_rate": 1.5634246575342463e-05, | |
| "loss": 0.6478, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.4452054794520548, | |
| "grad_norm": 15.242220878601074, | |
| "learning_rate": 1.5565753424657533e-05, | |
| "loss": 0.5599, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 16.776491165161133, | |
| "learning_rate": 1.5497260273972602e-05, | |
| "loss": 0.6219, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.4589041095890412, | |
| "grad_norm": 37.072898864746094, | |
| "learning_rate": 1.5428767123287672e-05, | |
| "loss": 0.693, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.4657534246575343, | |
| "grad_norm": 14.27625560760498, | |
| "learning_rate": 1.5360273972602738e-05, | |
| "loss": 0.6377, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.4726027397260273, | |
| "grad_norm": 88.88594055175781, | |
| "learning_rate": 1.5291780821917807e-05, | |
| "loss": 0.5319, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 32.19773483276367, | |
| "learning_rate": 1.5223287671232877e-05, | |
| "loss": 0.6918, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.4863013698630136, | |
| "grad_norm": 11.865937232971191, | |
| "learning_rate": 1.5154794520547946e-05, | |
| "loss": 0.5923, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.4931506849315068, | |
| "grad_norm": 7.134413719177246, | |
| "learning_rate": 1.5086301369863012e-05, | |
| "loss": 0.639, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 23.235034942626953, | |
| "learning_rate": 1.5017808219178082e-05, | |
| "loss": 0.5687, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 9.480228424072266, | |
| "learning_rate": 1.4949315068493151e-05, | |
| "loss": 0.6373, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.5136986301369864, | |
| "grad_norm": 14.618119239807129, | |
| "learning_rate": 1.4880821917808219e-05, | |
| "loss": 0.6153, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 1.5205479452054793, | |
| "grad_norm": 17.39520263671875, | |
| "learning_rate": 1.4812328767123289e-05, | |
| "loss": 0.6147, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.5273972602739727, | |
| "grad_norm": 18.27179527282715, | |
| "learning_rate": 1.4743835616438356e-05, | |
| "loss": 0.638, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 29.771127700805664, | |
| "learning_rate": 1.4675342465753426e-05, | |
| "loss": 0.5823, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 1.541095890410959, | |
| "grad_norm": 10.72756290435791, | |
| "learning_rate": 1.4606849315068494e-05, | |
| "loss": 0.6456, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.547945205479452, | |
| "grad_norm": 7.729266166687012, | |
| "learning_rate": 1.4538356164383563e-05, | |
| "loss": 0.5763, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 1.5547945205479452, | |
| "grad_norm": 10.376163482666016, | |
| "learning_rate": 1.4469863013698629e-05, | |
| "loss": 0.6476, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 24.55621337890625, | |
| "learning_rate": 1.4401369863013699e-05, | |
| "loss": 0.694, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.5684931506849316, | |
| "grad_norm": 27.07891845703125, | |
| "learning_rate": 1.4332876712328766e-05, | |
| "loss": 0.6128, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 1.5753424657534247, | |
| "grad_norm": 11.695100784301758, | |
| "learning_rate": 1.4264383561643836e-05, | |
| "loss": 0.5408, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.5821917808219177, | |
| "grad_norm": 9.359724998474121, | |
| "learning_rate": 1.4195890410958904e-05, | |
| "loss": 0.6425, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 5.442782878875732, | |
| "learning_rate": 1.4127397260273973e-05, | |
| "loss": 0.6208, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 1.595890410958904, | |
| "grad_norm": 12.379018783569336, | |
| "learning_rate": 1.4058904109589041e-05, | |
| "loss": 0.6293, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 1.6027397260273972, | |
| "grad_norm": 7.264742851257324, | |
| "learning_rate": 1.399041095890411e-05, | |
| "loss": 0.5549, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.6095890410958904, | |
| "grad_norm": 9.013050079345703, | |
| "learning_rate": 1.3921917808219178e-05, | |
| "loss": 0.6517, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 8.451722145080566, | |
| "learning_rate": 1.3853424657534248e-05, | |
| "loss": 0.6393, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 1.6232876712328768, | |
| "grad_norm": 28.516525268554688, | |
| "learning_rate": 1.3784931506849315e-05, | |
| "loss": 0.5393, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 1.6301369863013697, | |
| "grad_norm": 12.159906387329102, | |
| "learning_rate": 1.3716438356164385e-05, | |
| "loss": 0.5287, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 1.6369863013698631, | |
| "grad_norm": 12.206666946411133, | |
| "learning_rate": 1.3647945205479453e-05, | |
| "loss": 0.5779, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 11.194659233093262, | |
| "learning_rate": 1.357945205479452e-05, | |
| "loss": 0.5891, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "eval_exact_match": 86.26300851466415, | |
| "eval_f1": 92.63117973024548, | |
| "eval_runtime": 373.981, | |
| "eval_samples_per_second": 28.263, | |
| "eval_steps_per_second": 1.767, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.6506849315068495, | |
| "grad_norm": 9.522333145141602, | |
| "learning_rate": 1.3510958904109588e-05, | |
| "loss": 0.5278, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 1.6575342465753424, | |
| "grad_norm": 6.968531608581543, | |
| "learning_rate": 1.3442465753424658e-05, | |
| "loss": 0.6286, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 1.6643835616438356, | |
| "grad_norm": 14.288814544677734, | |
| "learning_rate": 1.3373972602739725e-05, | |
| "loss": 0.6047, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 9.370450019836426, | |
| "learning_rate": 1.3305479452054795e-05, | |
| "loss": 0.6241, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 1.678082191780822, | |
| "grad_norm": 15.267229080200195, | |
| "learning_rate": 1.3236986301369863e-05, | |
| "loss": 0.6022, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.6849315068493151, | |
| "grad_norm": 4.616830348968506, | |
| "learning_rate": 1.3168493150684932e-05, | |
| "loss": 0.556, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.691780821917808, | |
| "grad_norm": 34.63910675048828, | |
| "learning_rate": 1.31e-05, | |
| "loss": 0.6593, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 38.46183776855469, | |
| "learning_rate": 1.303150684931507e-05, | |
| "loss": 0.6686, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 1.7054794520547945, | |
| "grad_norm": 8.761494636535645, | |
| "learning_rate": 1.2963013698630137e-05, | |
| "loss": 0.6229, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 1.7123287671232876, | |
| "grad_norm": 8.143457412719727, | |
| "learning_rate": 1.2894520547945207e-05, | |
| "loss": 0.5573, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.7191780821917808, | |
| "grad_norm": 9.30538272857666, | |
| "learning_rate": 1.2826027397260274e-05, | |
| "loss": 0.5749, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 23.49202537536621, | |
| "learning_rate": 1.2757534246575342e-05, | |
| "loss": 0.4989, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.7328767123287672, | |
| "grad_norm": 64.99909210205078, | |
| "learning_rate": 1.268904109589041e-05, | |
| "loss": 0.5611, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 1.7397260273972601, | |
| "grad_norm": 8.76978588104248, | |
| "learning_rate": 1.262054794520548e-05, | |
| "loss": 0.5026, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 1.7465753424657535, | |
| "grad_norm": 18.690654754638672, | |
| "learning_rate": 1.2552054794520547e-05, | |
| "loss": 0.5583, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 13.614333152770996, | |
| "learning_rate": 1.2483561643835617e-05, | |
| "loss": 0.664, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 1.7602739726027399, | |
| "grad_norm": 4.357183933258057, | |
| "learning_rate": 1.2415068493150685e-05, | |
| "loss": 0.5702, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 1.7671232876712328, | |
| "grad_norm": 17.687780380249023, | |
| "learning_rate": 1.2346575342465754e-05, | |
| "loss": 0.5633, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.773972602739726, | |
| "grad_norm": 18.73659896850586, | |
| "learning_rate": 1.2278082191780822e-05, | |
| "loss": 0.5673, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 26.50644874572754, | |
| "learning_rate": 1.2209589041095891e-05, | |
| "loss": 0.6397, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7876712328767124, | |
| "grad_norm": 18.648330688476562, | |
| "learning_rate": 1.2141095890410959e-05, | |
| "loss": 0.623, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 1.7945205479452055, | |
| "grad_norm": 9.746988296508789, | |
| "learning_rate": 1.2072602739726028e-05, | |
| "loss": 0.605, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 1.8013698630136985, | |
| "grad_norm": 18.111595153808594, | |
| "learning_rate": 1.2004109589041096e-05, | |
| "loss": 0.6267, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 4.0696120262146, | |
| "learning_rate": 1.1935616438356166e-05, | |
| "loss": 0.5834, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.8150684931506849, | |
| "grad_norm": 6.453619480133057, | |
| "learning_rate": 1.1867123287671232e-05, | |
| "loss": 0.5391, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 1.821917808219178, | |
| "grad_norm": 13.60524845123291, | |
| "learning_rate": 1.1798630136986301e-05, | |
| "loss": 0.5457, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 1.8287671232876712, | |
| "grad_norm": 15.015334129333496, | |
| "learning_rate": 1.1730136986301369e-05, | |
| "loss": 0.5734, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 37.3073844909668, | |
| "learning_rate": 1.1661643835616439e-05, | |
| "loss": 0.6094, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 1.8424657534246576, | |
| "grad_norm": 18.685808181762695, | |
| "learning_rate": 1.1593150684931506e-05, | |
| "loss": 0.5903, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 1.8493150684931505, | |
| "grad_norm": 14.505097389221191, | |
| "learning_rate": 1.1524657534246576e-05, | |
| "loss": 0.6347, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.856164383561644, | |
| "grad_norm": 9.019182205200195, | |
| "learning_rate": 1.1456164383561644e-05, | |
| "loss": 0.6303, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 10.008683204650879, | |
| "learning_rate": 1.1387671232876713e-05, | |
| "loss": 0.6534, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 1.8698630136986303, | |
| "grad_norm": 41.62886047363281, | |
| "learning_rate": 1.131917808219178e-05, | |
| "loss": 0.6067, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 1.8767123287671232, | |
| "grad_norm": 16.496919631958008, | |
| "learning_rate": 1.125068493150685e-05, | |
| "loss": 0.6346, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 1.8835616438356164, | |
| "grad_norm": 9.691319465637207, | |
| "learning_rate": 1.1182191780821918e-05, | |
| "loss": 0.5597, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 8.646857261657715, | |
| "learning_rate": 1.1113698630136988e-05, | |
| "loss": 0.5628, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.8972602739726028, | |
| "grad_norm": 11.704099655151367, | |
| "learning_rate": 1.1045205479452055e-05, | |
| "loss": 0.5685, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 1.904109589041096, | |
| "grad_norm": 8.259442329406738, | |
| "learning_rate": 1.0976712328767123e-05, | |
| "loss": 0.5124, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 1.910958904109589, | |
| "grad_norm": 11.258965492248535, | |
| "learning_rate": 1.0908219178082191e-05, | |
| "loss": 0.6225, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 6.926393985748291, | |
| "learning_rate": 1.083972602739726e-05, | |
| "loss": 0.5896, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "eval_exact_match": 86.45222327341533, | |
| "eval_f1": 92.99265968366515, | |
| "eval_runtime": 374.4115, | |
| "eval_samples_per_second": 28.231, | |
| "eval_steps_per_second": 1.765, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.9246575342465753, | |
| "grad_norm": 29.349151611328125, | |
| "learning_rate": 1.0771232876712328e-05, | |
| "loss": 0.5955, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 1.9315068493150684, | |
| "grad_norm": 17.9901123046875, | |
| "learning_rate": 1.0702739726027398e-05, | |
| "loss": 0.6424, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.9383561643835616, | |
| "grad_norm": 28.261518478393555, | |
| "learning_rate": 1.0634246575342465e-05, | |
| "loss": 0.5522, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 27.7220401763916, | |
| "learning_rate": 1.0565753424657535e-05, | |
| "loss": 0.5946, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 1.952054794520548, | |
| "grad_norm": 29.411664962768555, | |
| "learning_rate": 1.0497260273972603e-05, | |
| "loss": 0.6488, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 1.958904109589041, | |
| "grad_norm": 18.539878845214844, | |
| "learning_rate": 1.0428767123287672e-05, | |
| "loss": 0.5892, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 1.9657534246575343, | |
| "grad_norm": 40.02704620361328, | |
| "learning_rate": 1.036027397260274e-05, | |
| "loss": 0.6774, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 10.566826820373535, | |
| "learning_rate": 1.029178082191781e-05, | |
| "loss": 0.584, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.9794520547945207, | |
| "grad_norm": 119.12139129638672, | |
| "learning_rate": 1.0223287671232877e-05, | |
| "loss": 0.5327, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 1.9863013698630136, | |
| "grad_norm": 7.952827453613281, | |
| "learning_rate": 1.0154794520547947e-05, | |
| "loss": 0.6381, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.9931506849315068, | |
| "grad_norm": 7.785494327545166, | |
| "learning_rate": 1.0086301369863013e-05, | |
| "loss": 0.5943, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 14.003286361694336, | |
| "learning_rate": 1.0017808219178082e-05, | |
| "loss": 0.6213, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.006849315068493, | |
| "grad_norm": 13.985429763793945, | |
| "learning_rate": 9.94931506849315e-06, | |
| "loss": 0.3419, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.0136986301369864, | |
| "grad_norm": 9.347712516784668, | |
| "learning_rate": 9.88082191780822e-06, | |
| "loss": 0.3652, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.0205479452054793, | |
| "grad_norm": 10.267210006713867, | |
| "learning_rate": 9.812328767123287e-06, | |
| "loss": 0.3994, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 49.33255386352539, | |
| "learning_rate": 9.743835616438357e-06, | |
| "loss": 0.3548, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.0342465753424657, | |
| "grad_norm": 6.08221435546875, | |
| "learning_rate": 9.675342465753424e-06, | |
| "loss": 0.3511, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.041095890410959, | |
| "grad_norm": 3.6202051639556885, | |
| "learning_rate": 9.606849315068494e-06, | |
| "loss": 0.3856, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.047945205479452, | |
| "grad_norm": 30.515443801879883, | |
| "learning_rate": 9.538356164383562e-06, | |
| "loss": 0.3352, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 15.059028625488281, | |
| "learning_rate": 9.469863013698631e-06, | |
| "loss": 0.3396, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.0616438356164384, | |
| "grad_norm": 16.944942474365234, | |
| "learning_rate": 9.401369863013699e-06, | |
| "loss": 0.4043, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.0684931506849313, | |
| "grad_norm": 12.575927734375, | |
| "learning_rate": 9.332876712328768e-06, | |
| "loss": 0.3958, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.0753424657534247, | |
| "grad_norm": 5.4868574142456055, | |
| "learning_rate": 9.264383561643836e-06, | |
| "loss": 0.3482, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 6.648189544677734, | |
| "learning_rate": 9.195890410958904e-06, | |
| "loss": 0.3973, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.089041095890411, | |
| "grad_norm": 14.415315628051758, | |
| "learning_rate": 9.127397260273972e-06, | |
| "loss": 0.3581, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.095890410958904, | |
| "grad_norm": 16.45214080810547, | |
| "learning_rate": 9.058904109589041e-06, | |
| "loss": 0.3387, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.1027397260273974, | |
| "grad_norm": 21.584300994873047, | |
| "learning_rate": 8.990410958904109e-06, | |
| "loss": 0.3634, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 17.63618278503418, | |
| "learning_rate": 8.923287671232876e-06, | |
| "loss": 0.3903, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.1164383561643834, | |
| "grad_norm": 11.995582580566406, | |
| "learning_rate": 8.854794520547946e-06, | |
| "loss": 0.3552, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.1232876712328768, | |
| "grad_norm": 8.57909870147705, | |
| "learning_rate": 8.786301369863013e-06, | |
| "loss": 0.3398, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.1301369863013697, | |
| "grad_norm": 37.580360412597656, | |
| "learning_rate": 8.717808219178083e-06, | |
| "loss": 0.3554, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 10.59188461303711, | |
| "learning_rate": 8.64931506849315e-06, | |
| "loss": 0.374, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.143835616438356, | |
| "grad_norm": 9.80274486541748, | |
| "learning_rate": 8.58082191780822e-06, | |
| "loss": 0.3954, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.1506849315068495, | |
| "grad_norm": 12.254220008850098, | |
| "learning_rate": 8.512328767123288e-06, | |
| "loss": 0.3623, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.1575342465753424, | |
| "grad_norm": 52.04378128051758, | |
| "learning_rate": 8.443835616438357e-06, | |
| "loss": 0.3648, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 11.195460319519043, | |
| "learning_rate": 8.375342465753425e-06, | |
| "loss": 0.3791, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.171232876712329, | |
| "grad_norm": 24.721216201782227, | |
| "learning_rate": 8.306849315068495e-06, | |
| "loss": 0.3589, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.1780821917808217, | |
| "grad_norm": 14.407581329345703, | |
| "learning_rate": 8.23835616438356e-06, | |
| "loss": 0.3242, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.184931506849315, | |
| "grad_norm": 3.953585147857666, | |
| "learning_rate": 8.16986301369863e-06, | |
| "loss": 0.3214, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 2.40246319770813, | |
| "learning_rate": 8.101369863013698e-06, | |
| "loss": 0.4225, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 85.84673604541155, | |
| "eval_f1": 92.62008664474187, | |
| "eval_runtime": 373.8017, | |
| "eval_samples_per_second": 28.277, | |
| "eval_steps_per_second": 1.768, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.1986301369863015, | |
| "grad_norm": 1.9170202016830444, | |
| "learning_rate": 8.032876712328767e-06, | |
| "loss": 0.3246, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.2054794520547945, | |
| "grad_norm": 11.83944034576416, | |
| "learning_rate": 7.964383561643835e-06, | |
| "loss": 0.3439, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.212328767123288, | |
| "grad_norm": 19.029552459716797, | |
| "learning_rate": 7.895890410958905e-06, | |
| "loss": 0.3921, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 8.53359317779541, | |
| "learning_rate": 7.827397260273972e-06, | |
| "loss": 0.3652, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.2260273972602738, | |
| "grad_norm": 14.722824096679688, | |
| "learning_rate": 7.758904109589042e-06, | |
| "loss": 0.3759, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.232876712328767, | |
| "grad_norm": 13.773172378540039, | |
| "learning_rate": 7.69041095890411e-06, | |
| "loss": 0.3888, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.23972602739726, | |
| "grad_norm": 57.0293083190918, | |
| "learning_rate": 7.621917808219179e-06, | |
| "loss": 0.4037, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 12.52489948272705, | |
| "learning_rate": 7.553424657534246e-06, | |
| "loss": 0.3284, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 2.2534246575342465, | |
| "grad_norm": 7.030215263366699, | |
| "learning_rate": 7.484931506849315e-06, | |
| "loss": 0.4432, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 2.26027397260274, | |
| "grad_norm": 16.76194190979004, | |
| "learning_rate": 7.416438356164383e-06, | |
| "loss": 0.3826, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.267123287671233, | |
| "grad_norm": 18.368928909301758, | |
| "learning_rate": 7.347945205479452e-06, | |
| "loss": 0.3668, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 9.03032398223877, | |
| "learning_rate": 7.2794520547945206e-06, | |
| "loss": 0.3559, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 2.280821917808219, | |
| "grad_norm": 20.22943878173828, | |
| "learning_rate": 7.210958904109589e-06, | |
| "loss": 0.375, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 2.287671232876712, | |
| "grad_norm": 82.04647827148438, | |
| "learning_rate": 7.142465753424657e-06, | |
| "loss": 0.4216, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 2.2945205479452055, | |
| "grad_norm": 30.627729415893555, | |
| "learning_rate": 7.073972602739726e-06, | |
| "loss": 0.3863, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 29.063501358032227, | |
| "learning_rate": 7.005479452054794e-06, | |
| "loss": 0.3599, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 2.308219178082192, | |
| "grad_norm": 22.352022171020508, | |
| "learning_rate": 6.936986301369863e-06, | |
| "loss": 0.4223, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 2.315068493150685, | |
| "grad_norm": 9.427127838134766, | |
| "learning_rate": 6.8684931506849315e-06, | |
| "loss": 0.3762, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 2.3219178082191783, | |
| "grad_norm": 82.33016204833984, | |
| "learning_rate": 6.8e-06, | |
| "loss": 0.4001, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 11.738672256469727, | |
| "learning_rate": 6.731506849315069e-06, | |
| "loss": 0.334, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.3356164383561646, | |
| "grad_norm": 7.739569187164307, | |
| "learning_rate": 6.6630136986301365e-06, | |
| "loss": 0.3344, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 2.3424657534246576, | |
| "grad_norm": 17.13344383239746, | |
| "learning_rate": 6.594520547945205e-06, | |
| "loss": 0.3847, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 2.3493150684931505, | |
| "grad_norm": 19.908540725708008, | |
| "learning_rate": 6.526027397260274e-06, | |
| "loss": 0.3276, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 13.234891891479492, | |
| "learning_rate": 6.457534246575342e-06, | |
| "loss": 0.3489, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 2.363013698630137, | |
| "grad_norm": 14.126317024230957, | |
| "learning_rate": 6.389041095890411e-06, | |
| "loss": 0.3892, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.3698630136986303, | |
| "grad_norm": 1.4812201261520386, | |
| "learning_rate": 6.32054794520548e-06, | |
| "loss": 0.3717, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 2.3767123287671232, | |
| "grad_norm": 8.576751708984375, | |
| "learning_rate": 6.2520547945205474e-06, | |
| "loss": 0.3515, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 13.8392333984375, | |
| "learning_rate": 6.183561643835616e-06, | |
| "loss": 0.3399, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.3904109589041096, | |
| "grad_norm": 8.925323486328125, | |
| "learning_rate": 6.115068493150685e-06, | |
| "loss": 0.3598, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 2.3972602739726026, | |
| "grad_norm": 9.674339294433594, | |
| "learning_rate": 6.046575342465753e-06, | |
| "loss": 0.3573, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.404109589041096, | |
| "grad_norm": 4.276260852813721, | |
| "learning_rate": 5.978082191780822e-06, | |
| "loss": 0.3596, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 8.640230178833008, | |
| "learning_rate": 5.9095890410958906e-06, | |
| "loss": 0.3591, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 2.4178082191780823, | |
| "grad_norm": 24.30743408203125, | |
| "learning_rate": 5.841095890410958e-06, | |
| "loss": 0.3721, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 2.4246575342465753, | |
| "grad_norm": 14.108036041259766, | |
| "learning_rate": 5.772602739726027e-06, | |
| "loss": 0.4469, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.4315068493150687, | |
| "grad_norm": 6.523674488067627, | |
| "learning_rate": 5.704109589041096e-06, | |
| "loss": 0.356, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 14.432558059692383, | |
| "learning_rate": 5.635616438356164e-06, | |
| "loss": 0.3651, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 2.4452054794520546, | |
| "grad_norm": 42.44483947753906, | |
| "learning_rate": 5.567123287671233e-06, | |
| "loss": 0.3245, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 2.452054794520548, | |
| "grad_norm": 32.770179748535156, | |
| "learning_rate": 5.4986301369863015e-06, | |
| "loss": 0.3644, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 2.458904109589041, | |
| "grad_norm": 15.61273193359375, | |
| "learning_rate": 5.43013698630137e-06, | |
| "loss": 0.4216, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 6.411821365356445, | |
| "learning_rate": 5.361643835616438e-06, | |
| "loss": 0.3787, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "eval_exact_match": 86.03595080416272, | |
| "eval_f1": 92.59598379751915, | |
| "eval_runtime": 374.3115, | |
| "eval_samples_per_second": 28.239, | |
| "eval_steps_per_second": 1.766, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.4726027397260273, | |
| "grad_norm": 8.469440460205078, | |
| "learning_rate": 5.2931506849315065e-06, | |
| "loss": 0.3794, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 2.4794520547945207, | |
| "grad_norm": 34.662200927734375, | |
| "learning_rate": 5.224657534246575e-06, | |
| "loss": 0.3383, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 2.4863013698630136, | |
| "grad_norm": 9.564376831054688, | |
| "learning_rate": 5.156164383561644e-06, | |
| "loss": 0.3725, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 21.028474807739258, | |
| "learning_rate": 5.087671232876712e-06, | |
| "loss": 0.3401, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 30.721982955932617, | |
| "learning_rate": 5.019178082191781e-06, | |
| "loss": 0.4297, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.506849315068493, | |
| "grad_norm": 13.996516227722168, | |
| "learning_rate": 4.950684931506849e-06, | |
| "loss": 0.3357, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.5136986301369864, | |
| "grad_norm": 17.068370819091797, | |
| "learning_rate": 4.8821917808219174e-06, | |
| "loss": 0.3447, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 17.016204833984375, | |
| "learning_rate": 4.813698630136986e-06, | |
| "loss": 0.4224, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 2.5273972602739727, | |
| "grad_norm": 24.968732833862305, | |
| "learning_rate": 4.746575342465753e-06, | |
| "loss": 0.3144, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 2.5342465753424657, | |
| "grad_norm": 9.2473783493042, | |
| "learning_rate": 4.678082191780822e-06, | |
| "loss": 0.3584, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.541095890410959, | |
| "grad_norm": 4.206846714019775, | |
| "learning_rate": 4.60958904109589e-06, | |
| "loss": 0.4063, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 23.909751892089844, | |
| "learning_rate": 4.541095890410959e-06, | |
| "loss": 0.4532, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.5547945205479454, | |
| "grad_norm": 7.862274169921875, | |
| "learning_rate": 4.4726027397260276e-06, | |
| "loss": 0.3294, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 2.5616438356164384, | |
| "grad_norm": 11.439240455627441, | |
| "learning_rate": 4.404109589041095e-06, | |
| "loss": 0.35, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 2.5684931506849313, | |
| "grad_norm": 88.37039184570312, | |
| "learning_rate": 4.335616438356164e-06, | |
| "loss": 0.3066, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 11.26742172241211, | |
| "learning_rate": 4.267123287671233e-06, | |
| "loss": 0.3356, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 2.5821917808219177, | |
| "grad_norm": 6.460461616516113, | |
| "learning_rate": 4.198630136986301e-06, | |
| "loss": 0.3016, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 2.589041095890411, | |
| "grad_norm": 16.87274742126465, | |
| "learning_rate": 4.13013698630137e-06, | |
| "loss": 0.3711, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.595890410958904, | |
| "grad_norm": 9.084839820861816, | |
| "learning_rate": 4.0616438356164385e-06, | |
| "loss": 0.368, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 14.624054908752441, | |
| "learning_rate": 3.993150684931506e-06, | |
| "loss": 0.3217, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.6095890410958904, | |
| "grad_norm": 9.953332901000977, | |
| "learning_rate": 3.924657534246575e-06, | |
| "loss": 0.3739, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 2.616438356164384, | |
| "grad_norm": 23.690486907958984, | |
| "learning_rate": 3.8561643835616435e-06, | |
| "loss": 0.3504, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 2.6232876712328768, | |
| "grad_norm": 15.352545738220215, | |
| "learning_rate": 3.787671232876712e-06, | |
| "loss": 0.3248, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 9.884965896606445, | |
| "learning_rate": 3.7205479452054796e-06, | |
| "loss": 0.4217, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.636986301369863, | |
| "grad_norm": 6.916888236999512, | |
| "learning_rate": 3.652054794520548e-06, | |
| "loss": 0.3578, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.643835616438356, | |
| "grad_norm": 31.198816299438477, | |
| "learning_rate": 3.5835616438356164e-06, | |
| "loss": 0.354, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 2.6506849315068495, | |
| "grad_norm": 19.81708335876465, | |
| "learning_rate": 3.515068493150685e-06, | |
| "loss": 0.3645, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 13.156720161437988, | |
| "learning_rate": 3.4465753424657536e-06, | |
| "loss": 0.3673, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 2.6643835616438354, | |
| "grad_norm": 2.532683849334717, | |
| "learning_rate": 3.378082191780822e-06, | |
| "loss": 0.3605, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 2.671232876712329, | |
| "grad_norm": 21.692689895629883, | |
| "learning_rate": 3.3095890410958905e-06, | |
| "loss": 0.4003, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.678082191780822, | |
| "grad_norm": 12.005087852478027, | |
| "learning_rate": 3.241095890410959e-06, | |
| "loss": 0.3284, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 11.013394355773926, | |
| "learning_rate": 3.1726027397260277e-06, | |
| "loss": 0.3844, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 2.691780821917808, | |
| "grad_norm": 4.333644866943359, | |
| "learning_rate": 3.104109589041096e-06, | |
| "loss": 0.3349, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 2.6986301369863015, | |
| "grad_norm": 19.21991729736328, | |
| "learning_rate": 3.0356164383561646e-06, | |
| "loss": 0.4117, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 2.7054794520547945, | |
| "grad_norm": 4.89330530166626, | |
| "learning_rate": 2.967123287671233e-06, | |
| "loss": 0.3275, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 18.977140426635742, | |
| "learning_rate": 2.8986301369863014e-06, | |
| "loss": 0.2834, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.719178082191781, | |
| "grad_norm": 12.3119535446167, | |
| "learning_rate": 2.83013698630137e-06, | |
| "loss": 0.3581, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 2.7260273972602738, | |
| "grad_norm": 17.003293991088867, | |
| "learning_rate": 2.7616438356164386e-06, | |
| "loss": 0.3172, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 2.732876712328767, | |
| "grad_norm": 11.635333061218262, | |
| "learning_rate": 2.693150684931507e-06, | |
| "loss": 0.3026, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 69.9380874633789, | |
| "learning_rate": 2.6246575342465755e-06, | |
| "loss": 0.3742, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "eval_exact_match": 86.20624408703878, | |
| "eval_f1": 92.65137657035267, | |
| "eval_runtime": 374.5596, | |
| "eval_samples_per_second": 28.22, | |
| "eval_steps_per_second": 1.765, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.7465753424657535, | |
| "grad_norm": 9.264750480651855, | |
| "learning_rate": 2.556164383561644e-06, | |
| "loss": 0.3425, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 2.7534246575342465, | |
| "grad_norm": 16.502531051635742, | |
| "learning_rate": 2.4876712328767123e-06, | |
| "loss": 0.3694, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.76027397260274, | |
| "grad_norm": 8.838066101074219, | |
| "learning_rate": 2.419178082191781e-06, | |
| "loss": 0.3469, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 19.364974975585938, | |
| "learning_rate": 2.3506849315068496e-06, | |
| "loss": 0.349, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 2.7739726027397262, | |
| "grad_norm": 5.768221855163574, | |
| "learning_rate": 2.2821917808219178e-06, | |
| "loss": 0.346, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 2.780821917808219, | |
| "grad_norm": 14.016840934753418, | |
| "learning_rate": 2.2136986301369864e-06, | |
| "loss": 0.3712, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 2.787671232876712, | |
| "grad_norm": 11.506481170654297, | |
| "learning_rate": 2.1465753424657534e-06, | |
| "loss": 0.3534, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 10.090023040771484, | |
| "learning_rate": 2.078082191780822e-06, | |
| "loss": 0.3401, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.8013698630136985, | |
| "grad_norm": 13.737336158752441, | |
| "learning_rate": 2.0095890410958906e-06, | |
| "loss": 0.354, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 2.808219178082192, | |
| "grad_norm": 12.264633178710938, | |
| "learning_rate": 1.941095890410959e-06, | |
| "loss": 0.3419, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.815068493150685, | |
| "grad_norm": 13.820444107055664, | |
| "learning_rate": 1.8726027397260275e-06, | |
| "loss": 0.3269, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 21.030885696411133, | |
| "learning_rate": 1.8041095890410959e-06, | |
| "loss": 0.3196, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 2.828767123287671, | |
| "grad_norm": 57.07254409790039, | |
| "learning_rate": 1.7356164383561645e-06, | |
| "loss": 0.3293, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 2.8356164383561646, | |
| "grad_norm": 12.856677055358887, | |
| "learning_rate": 1.667123287671233e-06, | |
| "loss": 0.3315, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.8424657534246576, | |
| "grad_norm": 12.119853019714355, | |
| "learning_rate": 1.5986301369863013e-06, | |
| "loss": 0.3586, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 14.653212547302246, | |
| "learning_rate": 1.53013698630137e-06, | |
| "loss": 0.3567, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 2.856164383561644, | |
| "grad_norm": 73.28644561767578, | |
| "learning_rate": 1.4616438356164384e-06, | |
| "loss": 0.3455, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 2.863013698630137, | |
| "grad_norm": 8.310192108154297, | |
| "learning_rate": 1.393150684931507e-06, | |
| "loss": 0.3657, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 2.8698630136986303, | |
| "grad_norm": 7.967892169952393, | |
| "learning_rate": 1.3246575342465754e-06, | |
| "loss": 0.3304, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 17.915605545043945, | |
| "learning_rate": 1.2561643835616438e-06, | |
| "loss": 0.366, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.883561643835616, | |
| "grad_norm": 25.84489631652832, | |
| "learning_rate": 1.1876712328767125e-06, | |
| "loss": 0.3038, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 2.8904109589041096, | |
| "grad_norm": 16.457599639892578, | |
| "learning_rate": 1.1191780821917809e-06, | |
| "loss": 0.355, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 2.897260273972603, | |
| "grad_norm": 6.114046573638916, | |
| "learning_rate": 1.0506849315068493e-06, | |
| "loss": 0.3422, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 5.753522872924805, | |
| "learning_rate": 9.82191780821918e-07, | |
| "loss": 0.3282, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 2.910958904109589, | |
| "grad_norm": 10.73289966583252, | |
| "learning_rate": 9.136986301369863e-07, | |
| "loss": 0.3226, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 2.9178082191780823, | |
| "grad_norm": 6.898795127868652, | |
| "learning_rate": 8.452054794520548e-07, | |
| "loss": 0.3129, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.9246575342465753, | |
| "grad_norm": 27.09336280822754, | |
| "learning_rate": 7.767123287671233e-07, | |
| "loss": 0.2969, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 5.278433799743652, | |
| "learning_rate": 7.082191780821918e-07, | |
| "loss": 0.3627, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 2.9383561643835616, | |
| "grad_norm": 19.34138298034668, | |
| "learning_rate": 6.397260273972602e-07, | |
| "loss": 0.3751, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 2.9452054794520546, | |
| "grad_norm": 6.466626167297363, | |
| "learning_rate": 5.712328767123287e-07, | |
| "loss": 0.3006, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.952054794520548, | |
| "grad_norm": 12.599709510803223, | |
| "learning_rate": 5.027397260273973e-07, | |
| "loss": 0.3446, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 12.089503288269043, | |
| "learning_rate": 4.342465753424658e-07, | |
| "loss": 0.3768, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.9657534246575343, | |
| "grad_norm": 36.18865203857422, | |
| "learning_rate": 3.6575342465753424e-07, | |
| "loss": 0.311, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 2.9726027397260273, | |
| "grad_norm": 3.4642269611358643, | |
| "learning_rate": 2.972602739726027e-07, | |
| "loss": 0.3611, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 2.9794520547945207, | |
| "grad_norm": 19.14399528503418, | |
| "learning_rate": 2.2876712328767126e-07, | |
| "loss": 0.3402, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 7.870481967926025, | |
| "learning_rate": 1.6027397260273975e-07, | |
| "loss": 0.3231, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 2.993150684931507, | |
| "grad_norm": 23.812009811401367, | |
| "learning_rate": 9.178082191780822e-08, | |
| "loss": 0.3451, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 54.19838333129883, | |
| "learning_rate": 2.3287671232876715e-08, | |
| "loss": 0.3079, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 21900, | |
| "total_flos": 2.006738209660207e+18, | |
| "train_loss": 0.6803066000655361, | |
| "train_runtime": 54119.8344, | |
| "train_samples_per_second": 4.856, | |
| "train_steps_per_second": 0.405 | |
| } | |
| ], | |
| "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 | |
| } | |