| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 2000, | |
| "global_step": 16425, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0091324200913242, | |
| "grad_norm": 221.6422576904297, | |
| "learning_rate": 9.97564687975647e-06, | |
| "loss": 6.0077, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0182648401826484, | |
| "grad_norm": 53.356204986572266, | |
| "learning_rate": 9.945205479452056e-06, | |
| "loss": 4.1135, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 82.45592498779297, | |
| "learning_rate": 9.914764079147641e-06, | |
| "loss": 2.8067, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0365296803652968, | |
| "grad_norm": 55.64297866821289, | |
| "learning_rate": 9.884322678843227e-06, | |
| "loss": 2.3073, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.045662100456621, | |
| "grad_norm": 41.508663177490234, | |
| "learning_rate": 9.853881278538814e-06, | |
| "loss": 1.924, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 202.38238525390625, | |
| "learning_rate": 9.824048706240488e-06, | |
| "loss": 1.6561, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0639269406392694, | |
| "grad_norm": 40.34769821166992, | |
| "learning_rate": 9.793607305936074e-06, | |
| "loss": 1.7111, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0730593607305936, | |
| "grad_norm": 44.13899612426758, | |
| "learning_rate": 9.76316590563166e-06, | |
| "loss": 1.7346, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 33.424198150634766, | |
| "learning_rate": 9.732724505327246e-06, | |
| "loss": 1.5305, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.091324200913242, | |
| "grad_norm": 41.58274459838867, | |
| "learning_rate": 9.702283105022831e-06, | |
| "loss": 1.4748, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1004566210045662, | |
| "grad_norm": 24.718538284301758, | |
| "learning_rate": 9.671841704718417e-06, | |
| "loss": 1.3904, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 106.5848388671875, | |
| "learning_rate": 9.641400304414004e-06, | |
| "loss": 1.5063, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1187214611872146, | |
| "grad_norm": 30.068777084350586, | |
| "learning_rate": 9.61095890410959e-06, | |
| "loss": 1.4255, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.1278538812785388, | |
| "grad_norm": 40.35118103027344, | |
| "learning_rate": 9.580517503805176e-06, | |
| "loss": 1.3404, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 25.713661193847656, | |
| "learning_rate": 9.550076103500761e-06, | |
| "loss": 1.2721, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1461187214611872, | |
| "grad_norm": 37.896827697753906, | |
| "learning_rate": 9.519634703196347e-06, | |
| "loss": 1.2902, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.1552511415525114, | |
| "grad_norm": 28.79201316833496, | |
| "learning_rate": 9.489193302891934e-06, | |
| "loss": 1.3189, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 40.73591232299805, | |
| "learning_rate": 9.45875190258752e-06, | |
| "loss": 1.2969, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.1735159817351598, | |
| "grad_norm": 37.75580596923828, | |
| "learning_rate": 9.428310502283106e-06, | |
| "loss": 1.1992, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.182648401826484, | |
| "grad_norm": 23.1336727142334, | |
| "learning_rate": 9.397869101978691e-06, | |
| "loss": 1.2553, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 24.494401931762695, | |
| "learning_rate": 9.367427701674277e-06, | |
| "loss": 1.22, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.2009132420091324, | |
| "grad_norm": 25.258699417114258, | |
| "learning_rate": 9.336986301369864e-06, | |
| "loss": 1.2455, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2100456621004566, | |
| "grad_norm": 24.47820281982422, | |
| "learning_rate": 9.30654490106545e-06, | |
| "loss": 1.2092, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 21.027442932128906, | |
| "learning_rate": 9.276103500761036e-06, | |
| "loss": 1.1863, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.228310502283105, | |
| "grad_norm": 28.407264709472656, | |
| "learning_rate": 9.245662100456621e-06, | |
| "loss": 1.1964, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.2374429223744292, | |
| "grad_norm": 37.75485610961914, | |
| "learning_rate": 9.215220700152207e-06, | |
| "loss": 1.1217, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 36.54895782470703, | |
| "learning_rate": 9.184779299847793e-06, | |
| "loss": 1.1317, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.2557077625570776, | |
| "grad_norm": 20.005041122436523, | |
| "learning_rate": 9.15433789954338e-06, | |
| "loss": 1.0816, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.2648401826484018, | |
| "grad_norm": 30.079883575439453, | |
| "learning_rate": 9.123896499238966e-06, | |
| "loss": 1.1259, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 21.35468864440918, | |
| "learning_rate": 9.093455098934552e-06, | |
| "loss": 1.0557, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2831050228310502, | |
| "grad_norm": 35.95537185668945, | |
| "learning_rate": 9.063013698630137e-06, | |
| "loss": 1.2042, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2922374429223744, | |
| "grad_norm": 24.126571655273438, | |
| "learning_rate": 9.032572298325723e-06, | |
| "loss": 1.0724, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 37.01513671875, | |
| "learning_rate": 9.00213089802131e-06, | |
| "loss": 1.0762, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.3105022831050228, | |
| "grad_norm": 29.829130172729492, | |
| "learning_rate": 8.971689497716896e-06, | |
| "loss": 1.075, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.319634703196347, | |
| "grad_norm": 27.351768493652344, | |
| "learning_rate": 8.941248097412482e-06, | |
| "loss": 1.099, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 151.5540771484375, | |
| "learning_rate": 8.910806697108067e-06, | |
| "loss": 1.0758, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.3378995433789954, | |
| "grad_norm": 22.22905158996582, | |
| "learning_rate": 8.880365296803653e-06, | |
| "loss": 1.2048, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.3470319634703196, | |
| "grad_norm": 22.852474212646484, | |
| "learning_rate": 8.849923896499239e-06, | |
| "loss": 1.1506, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 25.591411590576172, | |
| "learning_rate": 8.819482496194826e-06, | |
| "loss": 1.0681, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.365296803652968, | |
| "grad_norm": 57.05115509033203, | |
| "learning_rate": 8.789041095890412e-06, | |
| "loss": 1.0501, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.365296803652968, | |
| "eval_exact_match": 78.10785241248817, | |
| "eval_f1": 86.60142116809823, | |
| "eval_runtime": 408.6372, | |
| "eval_samples_per_second": 25.866, | |
| "eval_steps_per_second": 1.618, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3744292237442922, | |
| "grad_norm": 18.742263793945312, | |
| "learning_rate": 8.758599695585997e-06, | |
| "loss": 1.0993, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 44.574337005615234, | |
| "learning_rate": 8.728158295281583e-06, | |
| "loss": 1.101, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.3926940639269406, | |
| "grad_norm": 13.44596004486084, | |
| "learning_rate": 8.697716894977169e-06, | |
| "loss": 1.0706, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.4018264840182648, | |
| "grad_norm": 35.56928253173828, | |
| "learning_rate": 8.667275494672756e-06, | |
| "loss": 1.0132, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 23.093910217285156, | |
| "learning_rate": 8.636834094368342e-06, | |
| "loss": 1.075, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.4200913242009132, | |
| "grad_norm": 53.623291015625, | |
| "learning_rate": 8.606392694063928e-06, | |
| "loss": 1.0114, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.4292237442922374, | |
| "grad_norm": 19.787992477416992, | |
| "learning_rate": 8.575951293759513e-06, | |
| "loss": 0.9579, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 85.31494903564453, | |
| "learning_rate": 8.545509893455099e-06, | |
| "loss": 1.0426, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.4474885844748858, | |
| "grad_norm": 14.577733993530273, | |
| "learning_rate": 8.515068493150686e-06, | |
| "loss": 1.0229, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.45662100456621, | |
| "grad_norm": 40.19715881347656, | |
| "learning_rate": 8.484627092846272e-06, | |
| "loss": 1.05, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 30.88080596923828, | |
| "learning_rate": 8.454185692541858e-06, | |
| "loss": 0.9825, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.4748858447488584, | |
| "grad_norm": 23.49712562561035, | |
| "learning_rate": 8.423744292237443e-06, | |
| "loss": 1.0332, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.4840182648401826, | |
| "grad_norm": 28.84528160095215, | |
| "learning_rate": 8.393302891933029e-06, | |
| "loss": 0.9683, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 93.11555480957031, | |
| "learning_rate": 8.362861491628615e-06, | |
| "loss": 0.9531, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.502283105022831, | |
| "grad_norm": 30.038545608520508, | |
| "learning_rate": 8.332420091324202e-06, | |
| "loss": 0.9505, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.5114155251141552, | |
| "grad_norm": 26.936176300048828, | |
| "learning_rate": 8.301978691019788e-06, | |
| "loss": 1.0348, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 47.295257568359375, | |
| "learning_rate": 8.271537290715373e-06, | |
| "loss": 0.9554, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.5296803652968036, | |
| "grad_norm": 18.67024803161621, | |
| "learning_rate": 8.241095890410959e-06, | |
| "loss": 0.981, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.5388127853881278, | |
| "grad_norm": 35.41249084472656, | |
| "learning_rate": 8.210654490106545e-06, | |
| "loss": 0.9714, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 89.86371612548828, | |
| "learning_rate": 8.180213089802132e-06, | |
| "loss": 0.969, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5570776255707762, | |
| "grad_norm": 17.144739151000977, | |
| "learning_rate": 8.149771689497718e-06, | |
| "loss": 0.9632, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.5662100456621004, | |
| "grad_norm": 34.07986068725586, | |
| "learning_rate": 8.119330289193304e-06, | |
| "loss": 0.9724, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 33.85867691040039, | |
| "learning_rate": 8.08888888888889e-06, | |
| "loss": 0.9273, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.5844748858447488, | |
| "grad_norm": 28.088581085205078, | |
| "learning_rate": 8.058447488584475e-06, | |
| "loss": 0.9382, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.593607305936073, | |
| "grad_norm": 21.896081924438477, | |
| "learning_rate": 8.028006088280062e-06, | |
| "loss": 0.8934, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 24.811033248901367, | |
| "learning_rate": 7.997564687975648e-06, | |
| "loss": 0.8633, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.6118721461187214, | |
| "grad_norm": 17.348302841186523, | |
| "learning_rate": 7.967123287671234e-06, | |
| "loss": 0.9522, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.6210045662100456, | |
| "grad_norm": 16.166751861572266, | |
| "learning_rate": 7.93668188736682e-06, | |
| "loss": 0.8775, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 17.120800018310547, | |
| "learning_rate": 7.906240487062405e-06, | |
| "loss": 0.915, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.639269406392694, | |
| "grad_norm": 24.14845848083496, | |
| "learning_rate": 7.87579908675799e-06, | |
| "loss": 0.8992, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6484018264840182, | |
| "grad_norm": 21.47500228881836, | |
| "learning_rate": 7.845357686453578e-06, | |
| "loss": 0.8837, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 893.0142822265625, | |
| "learning_rate": 7.814916286149164e-06, | |
| "loss": 0.7566, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 19.631641387939453, | |
| "learning_rate": 7.78447488584475e-06, | |
| "loss": 0.9036, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.6757990867579908, | |
| "grad_norm": 21.34396743774414, | |
| "learning_rate": 7.754033485540335e-06, | |
| "loss": 0.8583, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 41.54507064819336, | |
| "learning_rate": 7.72359208523592e-06, | |
| "loss": 0.8158, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.6940639269406392, | |
| "grad_norm": 18.237974166870117, | |
| "learning_rate": 7.693150684931508e-06, | |
| "loss": 0.8654, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.7031963470319634, | |
| "grad_norm": 8.820842742919922, | |
| "learning_rate": 7.662709284627094e-06, | |
| "loss": 0.8388, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 12.438421249389648, | |
| "learning_rate": 7.63226788432268e-06, | |
| "loss": 0.9076, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.7214611872146118, | |
| "grad_norm": 31.41686248779297, | |
| "learning_rate": 7.601826484018265e-06, | |
| "loss": 0.8863, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.730593607305936, | |
| "grad_norm": 54.51583480834961, | |
| "learning_rate": 7.571385083713852e-06, | |
| "loss": 0.865, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.730593607305936, | |
| "eval_exact_match": 83.91674550614948, | |
| "eval_f1": 91.06161273436486, | |
| "eval_runtime": 406.2293, | |
| "eval_samples_per_second": 26.02, | |
| "eval_steps_per_second": 1.627, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 23.26498031616211, | |
| "learning_rate": 7.540943683409437e-06, | |
| "loss": 0.875, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.7488584474885844, | |
| "grad_norm": 22.035991668701172, | |
| "learning_rate": 7.510502283105023e-06, | |
| "loss": 0.8573, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.7579908675799086, | |
| "grad_norm": 38.90880584716797, | |
| "learning_rate": 7.4800608828006096e-06, | |
| "loss": 0.9069, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 21.68418312072754, | |
| "learning_rate": 7.449619482496195e-06, | |
| "loss": 0.8559, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.776255707762557, | |
| "grad_norm": 94.90328979492188, | |
| "learning_rate": 7.419178082191782e-06, | |
| "loss": 0.876, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.7853881278538812, | |
| "grad_norm": 40.229251861572266, | |
| "learning_rate": 7.3887366818873674e-06, | |
| "loss": 0.8589, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 23.035062789916992, | |
| "learning_rate": 7.358295281582953e-06, | |
| "loss": 0.8919, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.8036529680365296, | |
| "grad_norm": 21.854232788085938, | |
| "learning_rate": 7.32785388127854e-06, | |
| "loss": 0.9156, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.8127853881278538, | |
| "grad_norm": 14.7982759475708, | |
| "learning_rate": 7.297412480974125e-06, | |
| "loss": 0.7823, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 11.709835052490234, | |
| "learning_rate": 7.266971080669711e-06, | |
| "loss": 0.8257, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8310502283105022, | |
| "grad_norm": 34.76528549194336, | |
| "learning_rate": 7.2365296803652975e-06, | |
| "loss": 0.794, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.8401826484018264, | |
| "grad_norm": 29.65485954284668, | |
| "learning_rate": 7.206088280060883e-06, | |
| "loss": 0.8861, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 17.04764747619629, | |
| "learning_rate": 7.17564687975647e-06, | |
| "loss": 0.8877, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.8584474885844748, | |
| "grad_norm": 16.863462448120117, | |
| "learning_rate": 7.145205479452055e-06, | |
| "loss": 0.8717, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.867579908675799, | |
| "grad_norm": 22.229736328125, | |
| "learning_rate": 7.114764079147641e-06, | |
| "loss": 0.8354, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 21.367616653442383, | |
| "learning_rate": 7.084322678843228e-06, | |
| "loss": 0.8083, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.8858447488584474, | |
| "grad_norm": 23.51657485961914, | |
| "learning_rate": 7.053881278538813e-06, | |
| "loss": 0.8566, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.8949771689497716, | |
| "grad_norm": 25.13926124572754, | |
| "learning_rate": 7.023439878234399e-06, | |
| "loss": 0.7737, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 8.506234169006348, | |
| "learning_rate": 6.9929984779299855e-06, | |
| "loss": 0.7892, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.91324200913242, | |
| "grad_norm": 18.09160041809082, | |
| "learning_rate": 6.962557077625571e-06, | |
| "loss": 0.7986, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9223744292237442, | |
| "grad_norm": 16.591453552246094, | |
| "learning_rate": 6.932115677321158e-06, | |
| "loss": 0.8058, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 12.483757972717285, | |
| "learning_rate": 6.901674277016743e-06, | |
| "loss": 0.7827, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.9406392694063926, | |
| "grad_norm": 16.847013473510742, | |
| "learning_rate": 6.871232876712329e-06, | |
| "loss": 0.8306, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.9497716894977168, | |
| "grad_norm": 27.550743103027344, | |
| "learning_rate": 6.840791476407916e-06, | |
| "loss": 0.7677, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 118.00872802734375, | |
| "learning_rate": 6.810350076103501e-06, | |
| "loss": 0.8312, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.9680365296803652, | |
| "grad_norm": 28.487119674682617, | |
| "learning_rate": 6.779908675799087e-06, | |
| "loss": 0.8193, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.9771689497716894, | |
| "grad_norm": 25.242734909057617, | |
| "learning_rate": 6.7494672754946735e-06, | |
| "loss": 0.8542, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 29.314556121826172, | |
| "learning_rate": 6.719025875190259e-06, | |
| "loss": 0.8617, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.9954337899543378, | |
| "grad_norm": 7.821478366851807, | |
| "learning_rate": 6.688584474885846e-06, | |
| "loss": 0.7687, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.004566210045662, | |
| "grad_norm": 8.230454444885254, | |
| "learning_rate": 6.658143074581431e-06, | |
| "loss": 0.7007, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 17.57550048828125, | |
| "learning_rate": 6.627701674277017e-06, | |
| "loss": 0.6705, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.0228310502283104, | |
| "grad_norm": 31.853958129882812, | |
| "learning_rate": 6.597260273972604e-06, | |
| "loss": 0.6468, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.0319634703196348, | |
| "grad_norm": 25.719881057739258, | |
| "learning_rate": 6.566818873668189e-06, | |
| "loss": 0.7003, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 7.838934421539307, | |
| "learning_rate": 6.536377473363775e-06, | |
| "loss": 0.6471, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.0502283105022832, | |
| "grad_norm": 39.390769958496094, | |
| "learning_rate": 6.5059360730593615e-06, | |
| "loss": 0.6483, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.0593607305936072, | |
| "grad_norm": 45.38280487060547, | |
| "learning_rate": 6.475494672754947e-06, | |
| "loss": 0.5945, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 9.456714630126953, | |
| "learning_rate": 6.445053272450533e-06, | |
| "loss": 0.6261, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.0776255707762556, | |
| "grad_norm": 16.920135498046875, | |
| "learning_rate": 6.414611872146119e-06, | |
| "loss": 0.6281, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.08675799086758, | |
| "grad_norm": 33.61515426635742, | |
| "learning_rate": 6.384170471841705e-06, | |
| "loss": 0.7094, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 14.23517894744873, | |
| "learning_rate": 6.3537290715372916e-06, | |
| "loss": 0.6322, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 85.34531693472091, | |
| "eval_f1": 92.10865701973381, | |
| "eval_runtime": 406.1815, | |
| "eval_samples_per_second": 26.023, | |
| "eval_steps_per_second": 1.627, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1050228310502284, | |
| "grad_norm": 12.242176055908203, | |
| "learning_rate": 6.323287671232877e-06, | |
| "loss": 0.6571, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.1141552511415524, | |
| "grad_norm": 22.35227394104004, | |
| "learning_rate": 6.292846270928463e-06, | |
| "loss": 0.7238, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 19.85144805908203, | |
| "learning_rate": 6.2624048706240495e-06, | |
| "loss": 0.6978, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.1324200913242009, | |
| "grad_norm": 37.92237854003906, | |
| "learning_rate": 6.231963470319635e-06, | |
| "loss": 0.6241, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.1415525114155252, | |
| "grad_norm": 21.54163932800293, | |
| "learning_rate": 6.201522070015221e-06, | |
| "loss": 0.6512, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 19.37819480895996, | |
| "learning_rate": 6.171080669710807e-06, | |
| "loss": 0.5888, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.1598173515981736, | |
| "grad_norm": 59.02834701538086, | |
| "learning_rate": 6.140639269406393e-06, | |
| "loss": 0.6524, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.1689497716894977, | |
| "grad_norm": 19.709335327148438, | |
| "learning_rate": 6.1101978691019796e-06, | |
| "loss": 0.6272, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 94.84259033203125, | |
| "learning_rate": 6.079756468797565e-06, | |
| "loss": 0.6506, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.187214611872146, | |
| "grad_norm": 27.324445724487305, | |
| "learning_rate": 6.049315068493151e-06, | |
| "loss": 0.7212, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1963470319634704, | |
| "grad_norm": 27.72054100036621, | |
| "learning_rate": 6.0188736681887374e-06, | |
| "loss": 0.6363, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 19.417390823364258, | |
| "learning_rate": 5.988432267884323e-06, | |
| "loss": 0.7211, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.2146118721461188, | |
| "grad_norm": 17.682470321655273, | |
| "learning_rate": 5.957990867579909e-06, | |
| "loss": 0.7258, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.2237442922374429, | |
| "grad_norm": 13.04556655883789, | |
| "learning_rate": 5.927549467275495e-06, | |
| "loss": 0.728, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 19.57162094116211, | |
| "learning_rate": 5.897108066971081e-06, | |
| "loss": 0.6404, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.2420091324200913, | |
| "grad_norm": 14.33209228515625, | |
| "learning_rate": 5.8666666666666675e-06, | |
| "loss": 0.6511, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.2511415525114156, | |
| "grad_norm": 11.948081970214844, | |
| "learning_rate": 5.836225266362253e-06, | |
| "loss": 0.6386, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 11.66781997680664, | |
| "learning_rate": 5.805783866057839e-06, | |
| "loss": 0.6319, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 1.269406392694064, | |
| "grad_norm": 40.73119354248047, | |
| "learning_rate": 5.775342465753425e-06, | |
| "loss": 0.6657, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 1.278538812785388, | |
| "grad_norm": 9.036286354064941, | |
| "learning_rate": 5.744901065449011e-06, | |
| "loss": 0.5752, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 23.672693252563477, | |
| "learning_rate": 5.715068493150685e-06, | |
| "loss": 0.7783, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 1.2968036529680365, | |
| "grad_norm": 19.77155303955078, | |
| "learning_rate": 5.684627092846271e-06, | |
| "loss": 0.6248, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 1.3059360730593608, | |
| "grad_norm": 11.635490417480469, | |
| "learning_rate": 5.654185692541857e-06, | |
| "loss": 0.6601, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 9.181543350219727, | |
| "learning_rate": 5.623744292237443e-06, | |
| "loss": 0.6029, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.3242009132420092, | |
| "grad_norm": 9.500978469848633, | |
| "learning_rate": 5.593302891933029e-06, | |
| "loss": 0.6676, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 16.34039878845215, | |
| "learning_rate": 5.562861491628615e-06, | |
| "loss": 0.6372, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 31.06061553955078, | |
| "learning_rate": 5.532420091324201e-06, | |
| "loss": 0.7154, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.3515981735159817, | |
| "grad_norm": 27.624256134033203, | |
| "learning_rate": 5.501978691019787e-06, | |
| "loss": 0.6578, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.360730593607306, | |
| "grad_norm": 15.686092376708984, | |
| "learning_rate": 5.471537290715373e-06, | |
| "loss": 0.6471, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 13.39659595489502, | |
| "learning_rate": 5.441095890410959e-06, | |
| "loss": 0.6481, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3789954337899544, | |
| "grad_norm": 8.473033905029297, | |
| "learning_rate": 5.410654490106545e-06, | |
| "loss": 0.6719, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.3881278538812785, | |
| "grad_norm": 20.205528259277344, | |
| "learning_rate": 5.380213089802131e-06, | |
| "loss": 0.5757, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 19.173370361328125, | |
| "learning_rate": 5.349771689497717e-06, | |
| "loss": 0.6266, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.4063926940639269, | |
| "grad_norm": 18.172975540161133, | |
| "learning_rate": 5.319330289193303e-06, | |
| "loss": 0.5782, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.4155251141552512, | |
| "grad_norm": 9.981927871704102, | |
| "learning_rate": 5.288888888888889e-06, | |
| "loss": 0.6568, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 10.33353328704834, | |
| "learning_rate": 5.2584474885844746e-06, | |
| "loss": 0.6683, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.4337899543378996, | |
| "grad_norm": 29.337627410888672, | |
| "learning_rate": 5.228006088280061e-06, | |
| "loss": 0.6609, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.4429223744292237, | |
| "grad_norm": 12.93662166595459, | |
| "learning_rate": 5.197564687975647e-06, | |
| "loss": 0.6648, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 29.355287551879883, | |
| "learning_rate": 5.167123287671233e-06, | |
| "loss": 0.6149, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.461187214611872, | |
| "grad_norm": 16.967021942138672, | |
| "learning_rate": 5.136681887366819e-06, | |
| "loss": 0.7242, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.461187214611872, | |
| "eval_exact_match": 86.1116367076632, | |
| "eval_f1": 92.4648953066692, | |
| "eval_runtime": 406.5019, | |
| "eval_samples_per_second": 26.002, | |
| "eval_steps_per_second": 1.626, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4703196347031964, | |
| "grad_norm": 11.783562660217285, | |
| "learning_rate": 5.106240487062405e-06, | |
| "loss": 0.6163, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 13.724154472351074, | |
| "learning_rate": 5.075799086757991e-06, | |
| "loss": 0.654, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.4885844748858448, | |
| "grad_norm": 10.402881622314453, | |
| "learning_rate": 5.045357686453577e-06, | |
| "loss": 0.5943, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.4977168949771689, | |
| "grad_norm": 20.88226318359375, | |
| "learning_rate": 5.0149162861491625e-06, | |
| "loss": 0.6884, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 10.194794654846191, | |
| "learning_rate": 4.984474885844749e-06, | |
| "loss": 0.6385, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.5159817351598175, | |
| "grad_norm": 5.215353488922119, | |
| "learning_rate": 4.954033485540336e-06, | |
| "loss": 0.6616, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.5251141552511416, | |
| "grad_norm": 12.108441352844238, | |
| "learning_rate": 4.923592085235921e-06, | |
| "loss": 0.6275, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 15.30664348602295, | |
| "learning_rate": 4.893150684931508e-06, | |
| "loss": 0.6444, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.54337899543379, | |
| "grad_norm": 18.936824798583984, | |
| "learning_rate": 4.8627092846270935e-06, | |
| "loss": 0.6434, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.5525114155251143, | |
| "grad_norm": 7.780643939971924, | |
| "learning_rate": 4.832267884322679e-06, | |
| "loss": 0.6749, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 19.338315963745117, | |
| "learning_rate": 4.801826484018266e-06, | |
| "loss": 0.7254, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.5707762557077625, | |
| "grad_norm": 15.236742973327637, | |
| "learning_rate": 4.771385083713851e-06, | |
| "loss": 0.6367, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.5799086757990868, | |
| "grad_norm": 14.18782901763916, | |
| "learning_rate": 4.740943683409437e-06, | |
| "loss": 0.6125, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 24.446998596191406, | |
| "learning_rate": 4.710502283105024e-06, | |
| "loss": 0.7166, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.5981735159817352, | |
| "grad_norm": 10.571455955505371, | |
| "learning_rate": 4.680060882800609e-06, | |
| "loss": 0.6405, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.6073059360730593, | |
| "grad_norm": 12.6207857131958, | |
| "learning_rate": 4.649619482496196e-06, | |
| "loss": 0.699, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 16.94082260131836, | |
| "learning_rate": 4.6191780821917815e-06, | |
| "loss": 0.6562, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.625570776255708, | |
| "grad_norm": 12.49853229522705, | |
| "learning_rate": 4.588736681887367e-06, | |
| "loss": 0.6067, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.634703196347032, | |
| "grad_norm": 31.45545196533203, | |
| "learning_rate": 4.558295281582954e-06, | |
| "loss": 0.5983, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 11.867836952209473, | |
| "learning_rate": 4.527853881278539e-06, | |
| "loss": 0.6285, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.6529680365296804, | |
| "grad_norm": 14.369145393371582, | |
| "learning_rate": 4.497412480974125e-06, | |
| "loss": 0.6224, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.6621004566210047, | |
| "grad_norm": 10.372947692871094, | |
| "learning_rate": 4.4669710806697116e-06, | |
| "loss": 0.6846, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 31.318424224853516, | |
| "learning_rate": 4.436529680365297e-06, | |
| "loss": 0.6266, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.6803652968036529, | |
| "grad_norm": 9.971720695495605, | |
| "learning_rate": 4.406088280060884e-06, | |
| "loss": 0.6116, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.6894977168949772, | |
| "grad_norm": 13.311767578125, | |
| "learning_rate": 4.3756468797564694e-06, | |
| "loss": 0.6772, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 10.240290641784668, | |
| "learning_rate": 4.345205479452055e-06, | |
| "loss": 0.6617, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.7077625570776256, | |
| "grad_norm": 13.63064956665039, | |
| "learning_rate": 4.314764079147642e-06, | |
| "loss": 0.6619, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.7168949771689497, | |
| "grad_norm": 10.325277328491211, | |
| "learning_rate": 4.284322678843227e-06, | |
| "loss": 0.593, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 11.242974281311035, | |
| "learning_rate": 4.253881278538813e-06, | |
| "loss": 0.5745, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.7351598173515983, | |
| "grad_norm": 9.956265449523926, | |
| "learning_rate": 4.2234398782343995e-06, | |
| "loss": 0.6093, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.7442922374429224, | |
| "grad_norm": 7.890584945678711, | |
| "learning_rate": 4.192998477929985e-06, | |
| "loss": 0.5413, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 27.36825942993164, | |
| "learning_rate": 4.162557077625572e-06, | |
| "loss": 0.6878, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.7625570776255708, | |
| "grad_norm": 27.449216842651367, | |
| "learning_rate": 4.1321156773211574e-06, | |
| "loss": 0.5506, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.771689497716895, | |
| "grad_norm": 13.601576805114746, | |
| "learning_rate": 4.101674277016743e-06, | |
| "loss": 0.6537, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 20.260234832763672, | |
| "learning_rate": 4.07123287671233e-06, | |
| "loss": 0.6504, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.7899543378995433, | |
| "grad_norm": 8.724568367004395, | |
| "learning_rate": 4.040791476407915e-06, | |
| "loss": 0.646, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.7990867579908676, | |
| "grad_norm": 12.175917625427246, | |
| "learning_rate": 4.010350076103501e-06, | |
| "loss": 0.6627, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 9.895513534545898, | |
| "learning_rate": 3.9799086757990875e-06, | |
| "loss": 0.665, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.817351598173516, | |
| "grad_norm": 9.971092224121094, | |
| "learning_rate": 3.949467275494673e-06, | |
| "loss": 0.6, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.82648401826484, | |
| "grad_norm": 8.606634140014648, | |
| "learning_rate": 3.91902587519026e-06, | |
| "loss": 0.6091, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.82648401826484, | |
| "eval_exact_match": 85.96026490066225, | |
| "eval_f1": 92.43954998062115, | |
| "eval_runtime": 406.4548, | |
| "eval_samples_per_second": 26.005, | |
| "eval_steps_per_second": 1.626, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 22.578126907348633, | |
| "learning_rate": 3.888584474885845e-06, | |
| "loss": 0.6935, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.8447488584474887, | |
| "grad_norm": 29.010652542114258, | |
| "learning_rate": 3.858143074581431e-06, | |
| "loss": 0.6552, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.8538812785388128, | |
| "grad_norm": 14.48583984375, | |
| "learning_rate": 3.827701674277018e-06, | |
| "loss": 0.6621, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 20.008892059326172, | |
| "learning_rate": 3.797260273972603e-06, | |
| "loss": 0.6703, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.8721461187214612, | |
| "grad_norm": 32.114173889160156, | |
| "learning_rate": 3.766818873668189e-06, | |
| "loss": 0.6611, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.8812785388127855, | |
| "grad_norm": 11.282811164855957, | |
| "learning_rate": 3.7363774733637747e-06, | |
| "loss": 0.6402, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 35.3563232421875, | |
| "learning_rate": 3.7059360730593608e-06, | |
| "loss": 0.5949, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.8995433789954337, | |
| "grad_norm": 13.916271209716797, | |
| "learning_rate": 3.675494672754947e-06, | |
| "loss": 0.6059, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.908675799086758, | |
| "grad_norm": 27.980178833007812, | |
| "learning_rate": 3.645053272450533e-06, | |
| "loss": 0.6022, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 10.44404411315918, | |
| "learning_rate": 3.6146118721461186e-06, | |
| "loss": 0.6091, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9269406392694064, | |
| "grad_norm": 9.538860321044922, | |
| "learning_rate": 3.5841704718417047e-06, | |
| "loss": 0.6522, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.9360730593607305, | |
| "grad_norm": 10.887898445129395, | |
| "learning_rate": 3.553729071537291e-06, | |
| "loss": 0.6392, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 40.29354476928711, | |
| "learning_rate": 3.5232876712328765e-06, | |
| "loss": 0.6288, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.954337899543379, | |
| "grad_norm": 29.277299880981445, | |
| "learning_rate": 3.4928462709284626e-06, | |
| "loss": 0.7117, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.9634703196347032, | |
| "grad_norm": 30.566862106323242, | |
| "learning_rate": 3.4624048706240487e-06, | |
| "loss": 0.6687, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 40.0589485168457, | |
| "learning_rate": 3.431963470319635e-06, | |
| "loss": 0.6745, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.9817351598173516, | |
| "grad_norm": 17.117198944091797, | |
| "learning_rate": 3.4015220700152205e-06, | |
| "loss": 0.6129, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.990867579908676, | |
| "grad_norm": 22.412439346313477, | |
| "learning_rate": 3.3710806697108066e-06, | |
| "loss": 0.6759, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 11.01193904876709, | |
| "learning_rate": 3.3406392694063927e-06, | |
| "loss": 0.6857, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 2.009132420091324, | |
| "grad_norm": 8.564952850341797, | |
| "learning_rate": 3.310197869101979e-06, | |
| "loss": 0.4644, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.018264840182648, | |
| "grad_norm": 13.085915565490723, | |
| "learning_rate": 3.2797564687975645e-06, | |
| "loss": 0.5165, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 14.992934226989746, | |
| "learning_rate": 3.2493150684931506e-06, | |
| "loss": 0.4602, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 2.036529680365297, | |
| "grad_norm": 15.46022891998291, | |
| "learning_rate": 3.2188736681887367e-06, | |
| "loss": 0.4808, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 2.045662100456621, | |
| "grad_norm": 18.564807891845703, | |
| "learning_rate": 3.188432267884323e-06, | |
| "loss": 0.5057, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 9.455687522888184, | |
| "learning_rate": 3.1579908675799085e-06, | |
| "loss": 0.5159, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 2.0639269406392695, | |
| "grad_norm": 24.577774047851562, | |
| "learning_rate": 3.1275494672754946e-06, | |
| "loss": 0.5139, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 2.0730593607305936, | |
| "grad_norm": 13.79776668548584, | |
| "learning_rate": 3.0971080669710807e-06, | |
| "loss": 0.4821, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 8.983718872070312, | |
| "learning_rate": 3.066666666666667e-06, | |
| "loss": 0.5006, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 2.091324200913242, | |
| "grad_norm": 5.1394171714782715, | |
| "learning_rate": 3.0362252663622525e-06, | |
| "loss": 0.4897, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 2.1004566210045663, | |
| "grad_norm": 20.14058494567871, | |
| "learning_rate": 3.0057838660578386e-06, | |
| "loss": 0.4637, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 12.311975479125977, | |
| "learning_rate": 2.9753424657534247e-06, | |
| "loss": 0.4711, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 2.1187214611872145, | |
| "grad_norm": 14.106832504272461, | |
| "learning_rate": 2.944901065449011e-06, | |
| "loss": 0.4794, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 2.127853881278539, | |
| "grad_norm": 13.880457878112793, | |
| "learning_rate": 2.9144596651445965e-06, | |
| "loss": 0.4539, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 13.430336952209473, | |
| "learning_rate": 2.884627092846271e-06, | |
| "loss": 0.5021, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 2.146118721461187, | |
| "grad_norm": 9.724105834960938, | |
| "learning_rate": 2.854185692541857e-06, | |
| "loss": 0.4958, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 2.1552511415525113, | |
| "grad_norm": 9.325925827026367, | |
| "learning_rate": 2.823744292237443e-06, | |
| "loss": 0.5021, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 8.804998397827148, | |
| "learning_rate": 2.7933028919330292e-06, | |
| "loss": 0.491, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 2.17351598173516, | |
| "grad_norm": 29.120508193969727, | |
| "learning_rate": 2.762861491628615e-06, | |
| "loss": 0.4636, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 2.182648401826484, | |
| "grad_norm": 7.916499137878418, | |
| "learning_rate": 2.732420091324201e-06, | |
| "loss": 0.417, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 4.8723297119140625, | |
| "learning_rate": 2.701978691019787e-06, | |
| "loss": 0.5017, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 85.71428571428571, | |
| "eval_f1": 92.4631336628407, | |
| "eval_runtime": 406.3845, | |
| "eval_samples_per_second": 26.01, | |
| "eval_steps_per_second": 1.627, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.2009132420091326, | |
| "grad_norm": 10.724153518676758, | |
| "learning_rate": 2.6715372907153732e-06, | |
| "loss": 0.4026, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 2.2100456621004567, | |
| "grad_norm": 24.394311904907227, | |
| "learning_rate": 2.641095890410959e-06, | |
| "loss": 0.4777, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 24.320796966552734, | |
| "learning_rate": 2.610654490106545e-06, | |
| "loss": 0.5017, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 2.228310502283105, | |
| "grad_norm": 16.416059494018555, | |
| "learning_rate": 2.580213089802131e-06, | |
| "loss": 0.457, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 2.237442922374429, | |
| "grad_norm": 12.136763572692871, | |
| "learning_rate": 2.5497716894977172e-06, | |
| "loss": 0.5128, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 11.742807388305664, | |
| "learning_rate": 2.519330289193303e-06, | |
| "loss": 0.4782, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 2.2557077625570776, | |
| "grad_norm": 18.11116600036621, | |
| "learning_rate": 2.488888888888889e-06, | |
| "loss": 0.5777, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 2.2648401826484017, | |
| "grad_norm": 19.796613693237305, | |
| "learning_rate": 2.458447488584475e-06, | |
| "loss": 0.4448, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 21.082096099853516, | |
| "learning_rate": 2.4280060882800612e-06, | |
| "loss": 0.5114, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 2.2831050228310503, | |
| "grad_norm": 45.51653289794922, | |
| "learning_rate": 2.397564687975647e-06, | |
| "loss": 0.483, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.2922374429223744, | |
| "grad_norm": 19.373531341552734, | |
| "learning_rate": 2.367123287671233e-06, | |
| "loss": 0.5448, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 13.89834976196289, | |
| "learning_rate": 2.336681887366819e-06, | |
| "loss": 0.4464, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 2.3105022831050226, | |
| "grad_norm": 7.9293317794799805, | |
| "learning_rate": 2.306240487062405e-06, | |
| "loss": 0.5133, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 2.319634703196347, | |
| "grad_norm": 12.511297225952148, | |
| "learning_rate": 2.275799086757991e-06, | |
| "loss": 0.5569, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 11.701761245727539, | |
| "learning_rate": 2.245357686453577e-06, | |
| "loss": 0.449, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 2.3378995433789953, | |
| "grad_norm": 9.876680374145508, | |
| "learning_rate": 2.214916286149163e-06, | |
| "loss": 0.4601, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 2.34703196347032, | |
| "grad_norm": 16.884044647216797, | |
| "learning_rate": 2.1844748858447488e-06, | |
| "loss": 0.4786, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 7.96138858795166, | |
| "learning_rate": 2.154033485540335e-06, | |
| "loss": 0.4868, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 2.365296803652968, | |
| "grad_norm": 20.283720016479492, | |
| "learning_rate": 2.123592085235921e-06, | |
| "loss": 0.5288, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 2.374429223744292, | |
| "grad_norm": 14.561513900756836, | |
| "learning_rate": 2.093150684931507e-06, | |
| "loss": 0.4397, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 31.751888275146484, | |
| "learning_rate": 2.0627092846270928e-06, | |
| "loss": 0.451, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 2.3926940639269407, | |
| "grad_norm": 20.009138107299805, | |
| "learning_rate": 2.032267884322679e-06, | |
| "loss": 0.4765, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 2.401826484018265, | |
| "grad_norm": 15.038084030151367, | |
| "learning_rate": 2.001826484018265e-06, | |
| "loss": 0.4864, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 7.947664737701416, | |
| "learning_rate": 1.971385083713851e-06, | |
| "loss": 0.4893, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 2.4200913242009134, | |
| "grad_norm": 32.7381591796875, | |
| "learning_rate": 1.9409436834094368e-06, | |
| "loss": 0.5467, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 2.4292237442922375, | |
| "grad_norm": 13.605596542358398, | |
| "learning_rate": 1.910502283105023e-06, | |
| "loss": 0.5008, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 15.449544906616211, | |
| "learning_rate": 1.880060882800609e-06, | |
| "loss": 0.4838, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 2.4474885844748857, | |
| "grad_norm": 19.095609664916992, | |
| "learning_rate": 1.8496194824961949e-06, | |
| "loss": 0.4473, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 2.45662100456621, | |
| "grad_norm": 28.19283676147461, | |
| "learning_rate": 1.819178082191781e-06, | |
| "loss": 0.4662, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 9.908361434936523, | |
| "learning_rate": 1.7887366818873668e-06, | |
| "loss": 0.5119, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4748858447488584, | |
| "grad_norm": 11.031998634338379, | |
| "learning_rate": 1.758295281582953e-06, | |
| "loss": 0.5061, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 2.4840182648401825, | |
| "grad_norm": 16.79950714111328, | |
| "learning_rate": 1.7278538812785388e-06, | |
| "loss": 0.4602, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 73.04358673095703, | |
| "learning_rate": 1.697412480974125e-06, | |
| "loss": 0.4397, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 2.502283105022831, | |
| "grad_norm": 9.9924955368042, | |
| "learning_rate": 1.6669710806697108e-06, | |
| "loss": 0.5315, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 2.5114155251141552, | |
| "grad_norm": 10.066008567810059, | |
| "learning_rate": 1.636529680365297e-06, | |
| "loss": 0.4622, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 14.022153854370117, | |
| "learning_rate": 1.6060882800608828e-06, | |
| "loss": 0.545, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 2.5296803652968034, | |
| "grad_norm": 9.877713203430176, | |
| "learning_rate": 1.575646879756469e-06, | |
| "loss": 0.4607, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 2.538812785388128, | |
| "grad_norm": 9.370101928710938, | |
| "learning_rate": 1.5452054794520548e-06, | |
| "loss": 0.4565, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 18.39552879333496, | |
| "learning_rate": 1.514764079147641e-06, | |
| "loss": 0.5645, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 2.557077625570776, | |
| "grad_norm": 17.700393676757812, | |
| "learning_rate": 1.4843226788432268e-06, | |
| "loss": 0.4371, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.557077625570776, | |
| "eval_exact_match": 86.20624408703878, | |
| "eval_f1": 92.64669797483194, | |
| "eval_runtime": 406.4221, | |
| "eval_samples_per_second": 26.007, | |
| "eval_steps_per_second": 1.626, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5662100456621006, | |
| "grad_norm": 12.384415626525879, | |
| "learning_rate": 1.453881278538813e-06, | |
| "loss": 0.462, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 12.213122367858887, | |
| "learning_rate": 1.4240487062404874e-06, | |
| "loss": 0.4749, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 2.584474885844749, | |
| "grad_norm": 13.337528228759766, | |
| "learning_rate": 1.3936073059360733e-06, | |
| "loss": 0.4622, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 2.593607305936073, | |
| "grad_norm": 247.74070739746094, | |
| "learning_rate": 1.3631659056316594e-06, | |
| "loss": 0.4973, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 19.168542861938477, | |
| "learning_rate": 1.3327245053272453e-06, | |
| "loss": 0.4236, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 2.6118721461187215, | |
| "grad_norm": 38.19758224487305, | |
| "learning_rate": 1.3022831050228314e-06, | |
| "loss": 0.4858, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 2.6210045662100456, | |
| "grad_norm": 19.262054443359375, | |
| "learning_rate": 1.2718417047184173e-06, | |
| "loss": 0.4157, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 8.72314739227295, | |
| "learning_rate": 1.2414003044140032e-06, | |
| "loss": 0.4958, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 2.6392694063926943, | |
| "grad_norm": 14.77987289428711, | |
| "learning_rate": 1.210958904109589e-06, | |
| "loss": 0.4949, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 2.6484018264840183, | |
| "grad_norm": 26.928768157958984, | |
| "learning_rate": 1.1805175038051752e-06, | |
| "loss": 0.4543, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 26.89804458618164, | |
| "learning_rate": 1.150076103500761e-06, | |
| "loss": 0.4866, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 24.923315048217773, | |
| "learning_rate": 1.1196347031963471e-06, | |
| "loss": 0.5296, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.6757990867579906, | |
| "grad_norm": 22.595211029052734, | |
| "learning_rate": 1.089193302891933e-06, | |
| "loss": 0.5194, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 6.4614949226379395, | |
| "learning_rate": 1.0587519025875191e-06, | |
| "loss": 0.5242, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.6940639269406392, | |
| "grad_norm": 12.488426208496094, | |
| "learning_rate": 1.028310502283105e-06, | |
| "loss": 0.4829, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.7031963470319633, | |
| "grad_norm": 11.035359382629395, | |
| "learning_rate": 9.978691019786911e-07, | |
| "loss": 0.4688, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 15.173184394836426, | |
| "learning_rate": 9.67427701674277e-07, | |
| "loss": 0.4491, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.721461187214612, | |
| "grad_norm": 10.253326416015625, | |
| "learning_rate": 9.369863013698631e-07, | |
| "loss": 0.4675, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.730593607305936, | |
| "grad_norm": 21.999897003173828, | |
| "learning_rate": 9.065449010654491e-07, | |
| "loss": 0.4482, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 20.603681564331055, | |
| "learning_rate": 8.761035007610351e-07, | |
| "loss": 0.4908, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.748858447488584, | |
| "grad_norm": 28.629119873046875, | |
| "learning_rate": 8.456621004566211e-07, | |
| "loss": 0.51, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.7579908675799087, | |
| "grad_norm": 10.979697227478027, | |
| "learning_rate": 8.152207001522071e-07, | |
| "loss": 0.4738, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 23.009090423583984, | |
| "learning_rate": 7.847792998477931e-07, | |
| "loss": 0.4778, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.776255707762557, | |
| "grad_norm": 13.957650184631348, | |
| "learning_rate": 7.543378995433791e-07, | |
| "loss": 0.5011, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.7853881278538815, | |
| "grad_norm": 51.35847091674805, | |
| "learning_rate": 7.238964992389651e-07, | |
| "loss": 0.4852, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 13.39809799194336, | |
| "learning_rate": 6.934550989345511e-07, | |
| "loss": 0.4967, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.8036529680365296, | |
| "grad_norm": 16.116273880004883, | |
| "learning_rate": 6.630136986301371e-07, | |
| "loss": 0.4922, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.8127853881278537, | |
| "grad_norm": 11.181424140930176, | |
| "learning_rate": 6.325722983257231e-07, | |
| "loss": 0.4525, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 11.645832061767578, | |
| "learning_rate": 6.021308980213091e-07, | |
| "loss": 0.4547, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.8310502283105023, | |
| "grad_norm": 37.076683044433594, | |
| "learning_rate": 5.716894977168951e-07, | |
| "loss": 0.4633, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.8401826484018264, | |
| "grad_norm": 13.449100494384766, | |
| "learning_rate": 5.412480974124811e-07, | |
| "loss": 0.4936, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 20.85655975341797, | |
| "learning_rate": 5.10806697108067e-07, | |
| "loss": 0.4931, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.858447488584475, | |
| "grad_norm": 19.300870895385742, | |
| "learning_rate": 4.80365296803653e-07, | |
| "loss": 0.4486, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.867579908675799, | |
| "grad_norm": 30.8187255859375, | |
| "learning_rate": 4.49923896499239e-07, | |
| "loss": 0.4809, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 15.742734909057617, | |
| "learning_rate": 4.19482496194825e-07, | |
| "loss": 0.4727, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.8858447488584473, | |
| "grad_norm": 11.21142578125, | |
| "learning_rate": 3.89041095890411e-07, | |
| "loss": 0.4315, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.8949771689497714, | |
| "grad_norm": 19.95684242248535, | |
| "learning_rate": 3.58599695585997e-07, | |
| "loss": 0.4594, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 10.31857681274414, | |
| "learning_rate": 3.28158295281583e-07, | |
| "loss": 0.4698, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.91324200913242, | |
| "grad_norm": 11.127820014953613, | |
| "learning_rate": 2.97716894977169e-07, | |
| "loss": 0.4376, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.922374429223744, | |
| "grad_norm": 11.543863296508789, | |
| "learning_rate": 2.6727549467275497e-07, | |
| "loss": 0.4423, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.922374429223744, | |
| "eval_exact_match": 86.14001892147587, | |
| "eval_f1": 92.68818723743551, | |
| "eval_runtime": 406.3458, | |
| "eval_samples_per_second": 26.012, | |
| "eval_steps_per_second": 1.627, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 6.806540012359619, | |
| "learning_rate": 2.3683409436834097e-07, | |
| "loss": 0.4837, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.9406392694063928, | |
| "grad_norm": 11.885407447814941, | |
| "learning_rate": 2.0639269406392697e-07, | |
| "loss": 0.4772, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.949771689497717, | |
| "grad_norm": 14.364296913146973, | |
| "learning_rate": 1.7595129375951297e-07, | |
| "loss": 0.4339, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 30.527639389038086, | |
| "learning_rate": 1.461187214611872e-07, | |
| "loss": 0.5007, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.968036529680365, | |
| "grad_norm": 31.71573829650879, | |
| "learning_rate": 1.1567732115677322e-07, | |
| "loss": 0.4726, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.9771689497716896, | |
| "grad_norm": 12.93132209777832, | |
| "learning_rate": 8.523592085235922e-08, | |
| "loss": 0.4475, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 5.9728569984436035, | |
| "learning_rate": 5.479452054794521e-08, | |
| "loss": 0.4457, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.9954337899543377, | |
| "grad_norm": 9.14331340789795, | |
| "learning_rate": 2.4353120243531205e-08, | |
| "loss": 0.4563, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 16425, | |
| "total_flos": 2.0062622083669033e+18, | |
| "train_loss": 0.7534013637850455, | |
| "train_runtime": 50383.896, | |
| "train_samples_per_second": 5.216, | |
| "train_steps_per_second": 0.326 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 16425, | |
| "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": 2.0062622083669033e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |