| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 2000, | |
| "global_step": 21900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0091324200913242, | |
| "grad_norm": 141.86129760742188, | |
| "learning_rate": 9.982648401826485e-06, | |
| "loss": 6.156, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0182648401826484, | |
| "grad_norm": 83.82975769042969, | |
| "learning_rate": 9.959817351598175e-06, | |
| "loss": 4.1566, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 171.21966552734375, | |
| "learning_rate": 9.936986301369864e-06, | |
| "loss": 3.1539, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0365296803652968, | |
| "grad_norm": 86.14761352539062, | |
| "learning_rate": 9.914155251141553e-06, | |
| "loss": 2.7091, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.045662100456621, | |
| "grad_norm": 155.69285583496094, | |
| "learning_rate": 9.891324200913242e-06, | |
| "loss": 2.3896, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 3496.547607421875, | |
| "learning_rate": 9.868493150684932e-06, | |
| "loss": 2.186, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0639269406392694, | |
| "grad_norm": 53.159969329833984, | |
| "learning_rate": 9.845662100456623e-06, | |
| "loss": 2.3572, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0730593607305936, | |
| "grad_norm": 3373.076416015625, | |
| "learning_rate": 9.82283105022831e-06, | |
| "loss": 2.2758, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 105.64806365966797, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 2.0534, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.091324200913242, | |
| "grad_norm": 77.49466705322266, | |
| "learning_rate": 9.77716894977169e-06, | |
| "loss": 2.0835, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1004566210045662, | |
| "grad_norm": 111.51605224609375, | |
| "learning_rate": 9.75433789954338e-06, | |
| "loss": 1.9298, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 43.85201644897461, | |
| "learning_rate": 9.731506849315069e-06, | |
| "loss": 1.9758, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1187214611872146, | |
| "grad_norm": 55.28047180175781, | |
| "learning_rate": 9.708675799086758e-06, | |
| "loss": 1.9304, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.1278538812785388, | |
| "grad_norm": 50.79043197631836, | |
| "learning_rate": 9.685844748858449e-06, | |
| "loss": 1.9333, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 26.703508377075195, | |
| "learning_rate": 9.663013698630138e-06, | |
| "loss": 1.817, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1461187214611872, | |
| "grad_norm": 36.239036560058594, | |
| "learning_rate": 9.640182648401828e-06, | |
| "loss": 1.6908, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.1552511415525114, | |
| "grad_norm": 207.3531494140625, | |
| "learning_rate": 9.617351598173517e-06, | |
| "loss": 1.727, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 49.595481872558594, | |
| "learning_rate": 9.594520547945206e-06, | |
| "loss": 1.6568, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.1735159817351598, | |
| "grad_norm": 49.645843505859375, | |
| "learning_rate": 9.571689497716895e-06, | |
| "loss": 1.6459, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.182648401826484, | |
| "grad_norm": 28.5211181640625, | |
| "learning_rate": 9.548858447488585e-06, | |
| "loss": 1.5953, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 43.102718353271484, | |
| "learning_rate": 9.526027397260274e-06, | |
| "loss": 1.507, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.2009132420091324, | |
| "grad_norm": 26.59748077392578, | |
| "learning_rate": 9.503196347031965e-06, | |
| "loss": 1.4658, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2100456621004566, | |
| "grad_norm": 33.241085052490234, | |
| "learning_rate": 9.480365296803654e-06, | |
| "loss": 1.5042, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 31.159038543701172, | |
| "learning_rate": 9.457534246575343e-06, | |
| "loss": 1.4255, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.228310502283105, | |
| "grad_norm": 45.05030059814453, | |
| "learning_rate": 9.434703196347033e-06, | |
| "loss": 1.4393, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.2374429223744292, | |
| "grad_norm": 25.268295288085938, | |
| "learning_rate": 9.411872146118722e-06, | |
| "loss": 1.4258, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 24.656484603881836, | |
| "learning_rate": 9.389041095890413e-06, | |
| "loss": 1.3565, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.2557077625570776, | |
| "grad_norm": 23.01006507873535, | |
| "learning_rate": 9.3662100456621e-06, | |
| "loss": 1.3265, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.2648401826484018, | |
| "grad_norm": 30.002086639404297, | |
| "learning_rate": 9.343378995433791e-06, | |
| "loss": 1.3825, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 27.9675350189209, | |
| "learning_rate": 9.32054794520548e-06, | |
| "loss": 1.305, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2831050228310502, | |
| "grad_norm": 40.30059814453125, | |
| "learning_rate": 9.29771689497717e-06, | |
| "loss": 1.3627, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2922374429223744, | |
| "grad_norm": 29.09018898010254, | |
| "learning_rate": 9.274885844748859e-06, | |
| "loss": 1.3124, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 49.747859954833984, | |
| "learning_rate": 9.252054794520548e-06, | |
| "loss": 1.3085, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.3105022831050228, | |
| "grad_norm": 34.45816421508789, | |
| "learning_rate": 9.229223744292238e-06, | |
| "loss": 1.2565, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.319634703196347, | |
| "grad_norm": 25.871952056884766, | |
| "learning_rate": 9.206392694063929e-06, | |
| "loss": 1.3172, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 40.712486267089844, | |
| "learning_rate": 9.183561643835616e-06, | |
| "loss": 1.3525, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.3378995433789954, | |
| "grad_norm": 41.31215286254883, | |
| "learning_rate": 9.160730593607307e-06, | |
| "loss": 1.3379, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.3470319634703196, | |
| "grad_norm": 34.724308013916016, | |
| "learning_rate": 9.137899543378996e-06, | |
| "loss": 1.3613, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 32.3390007019043, | |
| "learning_rate": 9.115068493150686e-06, | |
| "loss": 1.2672, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.365296803652968, | |
| "grad_norm": 41.678157806396484, | |
| "learning_rate": 9.092237442922375e-06, | |
| "loss": 1.2615, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.365296803652968, | |
| "eval_exact_match": 71.52317880794702, | |
| "eval_f1": 80.78902012347558, | |
| "eval_runtime": 396.1868, | |
| "eval_samples_per_second": 26.679, | |
| "eval_steps_per_second": 1.668, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3744292237442922, | |
| "grad_norm": 25.548471450805664, | |
| "learning_rate": 9.069406392694064e-06, | |
| "loss": 1.2814, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 50.78987503051758, | |
| "learning_rate": 9.046575342465755e-06, | |
| "loss": 1.258, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.3926940639269406, | |
| "grad_norm": 27.325096130371094, | |
| "learning_rate": 9.023744292237444e-06, | |
| "loss": 1.2259, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.4018264840182648, | |
| "grad_norm": 45.96411895751953, | |
| "learning_rate": 9.000913242009132e-06, | |
| "loss": 1.1546, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 32.445091247558594, | |
| "learning_rate": 8.978082191780823e-06, | |
| "loss": 1.2579, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.4200913242009132, | |
| "grad_norm": 23.029979705810547, | |
| "learning_rate": 8.955251141552512e-06, | |
| "loss": 1.169, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.4292237442922374, | |
| "grad_norm": 30.386211395263672, | |
| "learning_rate": 8.932420091324201e-06, | |
| "loss": 1.1398, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 17.771953582763672, | |
| "learning_rate": 8.90958904109589e-06, | |
| "loss": 1.1934, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.4474885844748858, | |
| "grad_norm": 12.856189727783203, | |
| "learning_rate": 8.88675799086758e-06, | |
| "loss": 1.1591, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.45662100456621, | |
| "grad_norm": 29.3809871673584, | |
| "learning_rate": 8.863926940639271e-06, | |
| "loss": 1.199, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 14.796148300170898, | |
| "learning_rate": 8.84109589041096e-06, | |
| "loss": 1.1566, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.4748858447488584, | |
| "grad_norm": 32.102691650390625, | |
| "learning_rate": 8.81826484018265e-06, | |
| "loss": 1.2326, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.4840182648401826, | |
| "grad_norm": 27.475114822387695, | |
| "learning_rate": 8.795433789954339e-06, | |
| "loss": 1.1362, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 20.31199073791504, | |
| "learning_rate": 8.772602739726028e-06, | |
| "loss": 1.1284, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.502283105022831, | |
| "grad_norm": 34.34742736816406, | |
| "learning_rate": 8.749771689497719e-06, | |
| "loss": 1.1765, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.5114155251141552, | |
| "grad_norm": 38.820980072021484, | |
| "learning_rate": 8.726940639269406e-06, | |
| "loss": 1.2815, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 39.47126388549805, | |
| "learning_rate": 8.704109589041096e-06, | |
| "loss": 1.2214, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.5296803652968036, | |
| "grad_norm": 23.623933792114258, | |
| "learning_rate": 8.681278538812787e-06, | |
| "loss": 1.2233, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.5388127853881278, | |
| "grad_norm": 28.29753875732422, | |
| "learning_rate": 8.658447488584476e-06, | |
| "loss": 1.2184, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 11.690893173217773, | |
| "learning_rate": 8.635616438356165e-06, | |
| "loss": 1.2078, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5570776255707762, | |
| "grad_norm": 22.935726165771484, | |
| "learning_rate": 8.612785388127854e-06, | |
| "loss": 1.1662, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.5662100456621004, | |
| "grad_norm": 41.53331756591797, | |
| "learning_rate": 8.589954337899544e-06, | |
| "loss": 1.1703, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 39.44691467285156, | |
| "learning_rate": 8.567123287671235e-06, | |
| "loss": 1.0978, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.5844748858447488, | |
| "grad_norm": 24.385053634643555, | |
| "learning_rate": 8.544292237442922e-06, | |
| "loss": 1.164, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.593607305936073, | |
| "grad_norm": 25.12101173400879, | |
| "learning_rate": 8.521461187214613e-06, | |
| "loss": 1.0858, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 26.9351806640625, | |
| "learning_rate": 8.498630136986302e-06, | |
| "loss": 1.119, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.6118721461187214, | |
| "grad_norm": 17.235322952270508, | |
| "learning_rate": 8.475799086757992e-06, | |
| "loss": 1.1505, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.6210045662100456, | |
| "grad_norm": 27.588298797607422, | |
| "learning_rate": 8.452968036529681e-06, | |
| "loss": 1.0935, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 27.88182258605957, | |
| "learning_rate": 8.43013698630137e-06, | |
| "loss": 1.138, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.639269406392694, | |
| "grad_norm": 38.748836517333984, | |
| "learning_rate": 8.40730593607306e-06, | |
| "loss": 1.0731, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6484018264840182, | |
| "grad_norm": 38.5189094543457, | |
| "learning_rate": 8.38447488584475e-06, | |
| "loss": 1.0474, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 33.10224151611328, | |
| "learning_rate": 8.361643835616438e-06, | |
| "loss": 0.8802, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 45.91608810424805, | |
| "learning_rate": 8.338812785388129e-06, | |
| "loss": 1.0351, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.6757990867579908, | |
| "grad_norm": 59.66465759277344, | |
| "learning_rate": 8.315981735159818e-06, | |
| "loss": 1.0275, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 20.09067726135254, | |
| "learning_rate": 8.293150684931507e-06, | |
| "loss": 1.0122, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.6940639269406392, | |
| "grad_norm": 17.678356170654297, | |
| "learning_rate": 8.270319634703197e-06, | |
| "loss": 1.0117, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.7031963470319634, | |
| "grad_norm": 15.248215675354004, | |
| "learning_rate": 8.247488584474886e-06, | |
| "loss": 1.0043, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 15.972944259643555, | |
| "learning_rate": 8.224657534246577e-06, | |
| "loss": 1.075, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.7214611872146118, | |
| "grad_norm": 35.40599822998047, | |
| "learning_rate": 8.201826484018266e-06, | |
| "loss": 1.0646, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.730593607305936, | |
| "grad_norm": 50.43526840209961, | |
| "learning_rate": 8.178995433789955e-06, | |
| "loss": 1.04, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.730593607305936, | |
| "eval_exact_match": 78.41059602649007, | |
| "eval_f1": 86.58121239815173, | |
| "eval_runtime": 396.4641, | |
| "eval_samples_per_second": 26.661, | |
| "eval_steps_per_second": 1.667, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 38.34847640991211, | |
| "learning_rate": 8.156164383561645e-06, | |
| "loss": 1.0143, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.7488584474885844, | |
| "grad_norm": 24.275487899780273, | |
| "learning_rate": 8.133333333333334e-06, | |
| "loss": 1.0638, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.7579908675799086, | |
| "grad_norm": 57.73759078979492, | |
| "learning_rate": 8.110502283105023e-06, | |
| "loss": 1.0726, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 26.460296630859375, | |
| "learning_rate": 8.087671232876712e-06, | |
| "loss": 0.9874, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.776255707762557, | |
| "grad_norm": 34.39996337890625, | |
| "learning_rate": 8.064840182648402e-06, | |
| "loss": 1.0369, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.7853881278538812, | |
| "grad_norm": 31.487375259399414, | |
| "learning_rate": 8.042009132420093e-06, | |
| "loss": 1.033, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 26.804845809936523, | |
| "learning_rate": 8.019178082191782e-06, | |
| "loss": 1.0823, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.8036529680365296, | |
| "grad_norm": 17.534334182739258, | |
| "learning_rate": 7.996347031963471e-06, | |
| "loss": 1.0338, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.8127853881278538, | |
| "grad_norm": 16.264385223388672, | |
| "learning_rate": 7.97351598173516e-06, | |
| "loss": 0.9379, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 15.467246055603027, | |
| "learning_rate": 7.95068493150685e-06, | |
| "loss": 0.9996, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8310502283105022, | |
| "grad_norm": 38.00102615356445, | |
| "learning_rate": 7.92785388127854e-06, | |
| "loss": 0.9505, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.8401826484018264, | |
| "grad_norm": 25.31998634338379, | |
| "learning_rate": 7.905022831050228e-06, | |
| "loss": 1.0487, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 14.750944137573242, | |
| "learning_rate": 7.88219178082192e-06, | |
| "loss": 1.0197, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.8584474885844748, | |
| "grad_norm": 34.27394485473633, | |
| "learning_rate": 7.859360730593608e-06, | |
| "loss": 0.9946, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.867579908675799, | |
| "grad_norm": 48.27435302734375, | |
| "learning_rate": 7.836529680365298e-06, | |
| "loss": 0.9748, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 26.16417121887207, | |
| "learning_rate": 7.813698630136987e-06, | |
| "loss": 0.9385, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.8858447488584474, | |
| "grad_norm": 28.09895896911621, | |
| "learning_rate": 7.790867579908676e-06, | |
| "loss": 0.9434, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.8949771689497716, | |
| "grad_norm": 31.583642959594727, | |
| "learning_rate": 7.768036529680366e-06, | |
| "loss": 0.9019, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 11.045271873474121, | |
| "learning_rate": 7.745205479452056e-06, | |
| "loss": 0.9086, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.91324200913242, | |
| "grad_norm": 33.99599838256836, | |
| "learning_rate": 7.722374429223744e-06, | |
| "loss": 0.9362, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9223744292237442, | |
| "grad_norm": 20.473243713378906, | |
| "learning_rate": 7.699543378995435e-06, | |
| "loss": 0.9518, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 22.298221588134766, | |
| "learning_rate": 7.676712328767124e-06, | |
| "loss": 0.8972, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.9406392694063926, | |
| "grad_norm": 23.762950897216797, | |
| "learning_rate": 7.653881278538813e-06, | |
| "loss": 0.9735, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.9497716894977168, | |
| "grad_norm": 31.01708221435547, | |
| "learning_rate": 7.631050228310503e-06, | |
| "loss": 0.9931, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 27.428834915161133, | |
| "learning_rate": 7.608219178082192e-06, | |
| "loss": 0.9892, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.9680365296803652, | |
| "grad_norm": 23.573469161987305, | |
| "learning_rate": 7.585388127853882e-06, | |
| "loss": 0.9576, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.9771689497716894, | |
| "grad_norm": 30.324687957763672, | |
| "learning_rate": 7.562557077625572e-06, | |
| "loss": 0.9869, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 35.9589958190918, | |
| "learning_rate": 7.539726027397261e-06, | |
| "loss": 0.9965, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.9954337899543378, | |
| "grad_norm": 13.001126289367676, | |
| "learning_rate": 7.516894977168951e-06, | |
| "loss": 0.9539, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.004566210045662, | |
| "grad_norm": 10.314200401306152, | |
| "learning_rate": 7.49406392694064e-06, | |
| "loss": 0.8357, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0136986301369864, | |
| "grad_norm": 22.927391052246094, | |
| "learning_rate": 7.47123287671233e-06, | |
| "loss": 0.7973, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.0228310502283104, | |
| "grad_norm": 18.42430305480957, | |
| "learning_rate": 7.4484018264840185e-06, | |
| "loss": 0.7766, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.0319634703196348, | |
| "grad_norm": 35.02809524536133, | |
| "learning_rate": 7.425570776255709e-06, | |
| "loss": 0.8057, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.0410958904109588, | |
| "grad_norm": 9.175677299499512, | |
| "learning_rate": 7.402739726027398e-06, | |
| "loss": 0.7897, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.0502283105022832, | |
| "grad_norm": 72.41110229492188, | |
| "learning_rate": 7.379908675799088e-06, | |
| "loss": 0.766, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.0593607305936072, | |
| "grad_norm": 25.297029495239258, | |
| "learning_rate": 7.357077625570776e-06, | |
| "loss": 0.6862, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.0684931506849316, | |
| "grad_norm": 17.325456619262695, | |
| "learning_rate": 7.3342465753424665e-06, | |
| "loss": 0.7479, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.0776255707762556, | |
| "grad_norm": 21.178491592407227, | |
| "learning_rate": 7.311415525114156e-06, | |
| "loss": 0.715, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.08675799086758, | |
| "grad_norm": 30.497148513793945, | |
| "learning_rate": 7.288584474885846e-06, | |
| "loss": 0.7737, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "grad_norm": 30.200098037719727, | |
| "learning_rate": 7.265753424657534e-06, | |
| "loss": 0.7385, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.095890410958904, | |
| "eval_exact_match": 80.99337748344371, | |
| "eval_f1": 88.52122863303049, | |
| "eval_runtime": 395.8625, | |
| "eval_samples_per_second": 26.701, | |
| "eval_steps_per_second": 1.67, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1050228310502284, | |
| "grad_norm": 13.1331205368042, | |
| "learning_rate": 7.242922374429224e-06, | |
| "loss": 0.7753, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 1.1141552511415524, | |
| "grad_norm": 17.219268798828125, | |
| "learning_rate": 7.220091324200914e-06, | |
| "loss": 0.8186, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.1232876712328768, | |
| "grad_norm": 21.086429595947266, | |
| "learning_rate": 7.197260273972604e-06, | |
| "loss": 0.823, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 1.1324200913242009, | |
| "grad_norm": 36.774166107177734, | |
| "learning_rate": 7.174429223744292e-06, | |
| "loss": 0.8048, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.1415525114155252, | |
| "grad_norm": 14.021933555603027, | |
| "learning_rate": 7.151598173515982e-06, | |
| "loss": 0.771, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 1.1506849315068493, | |
| "grad_norm": 16.90314292907715, | |
| "learning_rate": 7.128767123287672e-06, | |
| "loss": 0.6957, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.1598173515981736, | |
| "grad_norm": 29.586990356445312, | |
| "learning_rate": 7.105936073059362e-06, | |
| "loss": 0.7806, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 1.1689497716894977, | |
| "grad_norm": 13.954606056213379, | |
| "learning_rate": 7.083105022831051e-06, | |
| "loss": 0.7764, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.178082191780822, | |
| "grad_norm": 27.69614601135254, | |
| "learning_rate": 7.06027397260274e-06, | |
| "loss": 0.7411, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.187214611872146, | |
| "grad_norm": 14.492104530334473, | |
| "learning_rate": 7.03744292237443e-06, | |
| "loss": 0.821, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1963470319634704, | |
| "grad_norm": 27.99992561340332, | |
| "learning_rate": 7.0146118721461195e-06, | |
| "loss": 0.7502, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.2054794520547945, | |
| "grad_norm": 26.75726890563965, | |
| "learning_rate": 6.991780821917809e-06, | |
| "loss": 0.8265, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.2146118721461188, | |
| "grad_norm": 20.587421417236328, | |
| "learning_rate": 6.968949771689498e-06, | |
| "loss": 0.8594, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.2237442922374429, | |
| "grad_norm": 15.334819793701172, | |
| "learning_rate": 6.946118721461188e-06, | |
| "loss": 0.8162, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.2328767123287672, | |
| "grad_norm": 18.709548950195312, | |
| "learning_rate": 6.9232876712328774e-06, | |
| "loss": 0.7414, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.2420091324200913, | |
| "grad_norm": 23.877904891967773, | |
| "learning_rate": 6.900456621004567e-06, | |
| "loss": 0.7779, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.2511415525114156, | |
| "grad_norm": 15.881481170654297, | |
| "learning_rate": 6.877625570776256e-06, | |
| "loss": 0.7659, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 1.2602739726027397, | |
| "grad_norm": 16.046165466308594, | |
| "learning_rate": 6.854794520547946e-06, | |
| "loss": 0.7246, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 1.269406392694064, | |
| "grad_norm": 19.089651107788086, | |
| "learning_rate": 6.831963470319636e-06, | |
| "loss": 0.7905, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 1.278538812785388, | |
| "grad_norm": 23.612180709838867, | |
| "learning_rate": 6.809132420091325e-06, | |
| "loss": 0.671, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2876712328767124, | |
| "grad_norm": 33.00151443481445, | |
| "learning_rate": 6.786301369863014e-06, | |
| "loss": 0.8364, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 1.2968036529680365, | |
| "grad_norm": 21.602069854736328, | |
| "learning_rate": 6.763470319634704e-06, | |
| "loss": 0.7281, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 1.3059360730593608, | |
| "grad_norm": 9.662558555603027, | |
| "learning_rate": 6.740639269406394e-06, | |
| "loss": 0.8133, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 1.3150684931506849, | |
| "grad_norm": 101.09040069580078, | |
| "learning_rate": 6.7178082191780825e-06, | |
| "loss": 0.7267, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.3242009132420092, | |
| "grad_norm": 29.580852508544922, | |
| "learning_rate": 6.6949771689497726e-06, | |
| "loss": 0.7759, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 16.52943992614746, | |
| "learning_rate": 6.672146118721462e-06, | |
| "loss": 0.74, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.3424657534246576, | |
| "grad_norm": 26.262174606323242, | |
| "learning_rate": 6.649315068493152e-06, | |
| "loss": 0.8074, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.3515981735159817, | |
| "grad_norm": 28.12957191467285, | |
| "learning_rate": 6.62648401826484e-06, | |
| "loss": 0.7235, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.360730593607306, | |
| "grad_norm": 29.391780853271484, | |
| "learning_rate": 6.6036529680365305e-06, | |
| "loss": 0.7643, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.36986301369863, | |
| "grad_norm": 15.40013313293457, | |
| "learning_rate": 6.58082191780822e-06, | |
| "loss": 0.7499, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3789954337899544, | |
| "grad_norm": 10.625802993774414, | |
| "learning_rate": 6.557990867579909e-06, | |
| "loss": 0.8129, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.3881278538812785, | |
| "grad_norm": 21.512968063354492, | |
| "learning_rate": 6.535159817351598e-06, | |
| "loss": 0.7053, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.3972602739726028, | |
| "grad_norm": 43.220603942871094, | |
| "learning_rate": 6.512328767123288e-06, | |
| "loss": 0.7325, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.4063926940639269, | |
| "grad_norm": 20.33064842224121, | |
| "learning_rate": 6.489497716894978e-06, | |
| "loss": 0.6695, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.4155251141552512, | |
| "grad_norm": 12.011942863464355, | |
| "learning_rate": 6.466666666666667e-06, | |
| "loss": 0.7438, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.4246575342465753, | |
| "grad_norm": 17.656845092773438, | |
| "learning_rate": 6.443835616438356e-06, | |
| "loss": 0.7761, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.4337899543378996, | |
| "grad_norm": 19.555429458618164, | |
| "learning_rate": 6.421004566210046e-06, | |
| "loss": 0.7815, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.4429223744292237, | |
| "grad_norm": 15.19387149810791, | |
| "learning_rate": 6.398173515981736e-06, | |
| "loss": 0.7897, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.452054794520548, | |
| "grad_norm": 48.02432632446289, | |
| "learning_rate": 6.375342465753425e-06, | |
| "loss": 0.7725, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.461187214611872, | |
| "grad_norm": 18.34738540649414, | |
| "learning_rate": 6.352511415525114e-06, | |
| "loss": 0.8278, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.461187214611872, | |
| "eval_exact_match": 81.57994323557237, | |
| "eval_f1": 88.68935584305143, | |
| "eval_runtime": 395.554, | |
| "eval_samples_per_second": 26.722, | |
| "eval_steps_per_second": 1.671, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4703196347031964, | |
| "grad_norm": 13.668383598327637, | |
| "learning_rate": 6.329680365296804e-06, | |
| "loss": 0.7063, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.4794520547945205, | |
| "grad_norm": 21.045486450195312, | |
| "learning_rate": 6.306849315068494e-06, | |
| "loss": 0.7451, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.4885844748858448, | |
| "grad_norm": 24.719881057739258, | |
| "learning_rate": 6.284018264840183e-06, | |
| "loss": 0.7129, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.4977168949771689, | |
| "grad_norm": 8.752976417541504, | |
| "learning_rate": 6.261187214611873e-06, | |
| "loss": 0.7768, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.5068493150684932, | |
| "grad_norm": 12.758523941040039, | |
| "learning_rate": 6.238356164383562e-06, | |
| "loss": 0.7322, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.5159817351598175, | |
| "grad_norm": 12.468632698059082, | |
| "learning_rate": 6.215525114155252e-06, | |
| "loss": 0.7916, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.5251141552511416, | |
| "grad_norm": 20.91657257080078, | |
| "learning_rate": 6.1926940639269405e-06, | |
| "loss": 0.7525, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.5342465753424657, | |
| "grad_norm": 38.200714111328125, | |
| "learning_rate": 6.169863013698631e-06, | |
| "loss": 0.7487, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.54337899543379, | |
| "grad_norm": 28.027910232543945, | |
| "learning_rate": 6.14703196347032e-06, | |
| "loss": 0.783, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.5525114155251143, | |
| "grad_norm": 9.391847610473633, | |
| "learning_rate": 6.12420091324201e-06, | |
| "loss": 0.7869, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.5616438356164384, | |
| "grad_norm": 122.94676971435547, | |
| "learning_rate": 6.101369863013698e-06, | |
| "loss": 0.8702, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.5707762557077625, | |
| "grad_norm": 32.576805114746094, | |
| "learning_rate": 6.0785388127853885e-06, | |
| "loss": 0.7748, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.5799086757990868, | |
| "grad_norm": 17.054048538208008, | |
| "learning_rate": 6.055707762557078e-06, | |
| "loss": 0.7431, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.589041095890411, | |
| "grad_norm": 17.969711303710938, | |
| "learning_rate": 6.032876712328768e-06, | |
| "loss": 0.8278, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.5981735159817352, | |
| "grad_norm": 33.43639373779297, | |
| "learning_rate": 6.010045662100456e-06, | |
| "loss": 0.7789, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.6073059360730593, | |
| "grad_norm": 28.036745071411133, | |
| "learning_rate": 5.987214611872146e-06, | |
| "loss": 0.7694, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.6164383561643836, | |
| "grad_norm": 17.829118728637695, | |
| "learning_rate": 5.9643835616438365e-06, | |
| "loss": 0.7598, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.625570776255708, | |
| "grad_norm": 19.825571060180664, | |
| "learning_rate": 5.941552511415526e-06, | |
| "loss": 0.6907, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.634703196347032, | |
| "grad_norm": 29.223722457885742, | |
| "learning_rate": 5.918721461187214e-06, | |
| "loss": 0.7, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 9.084179878234863, | |
| "learning_rate": 5.895890410958904e-06, | |
| "loss": 0.7237, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.6529680365296804, | |
| "grad_norm": 18.593826293945312, | |
| "learning_rate": 5.873059360730594e-06, | |
| "loss": 0.7623, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.6621004566210047, | |
| "grad_norm": 21.121200561523438, | |
| "learning_rate": 5.850228310502284e-06, | |
| "loss": 0.7438, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.6712328767123288, | |
| "grad_norm": 22.33557891845703, | |
| "learning_rate": 5.827397260273973e-06, | |
| "loss": 0.7657, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.6803652968036529, | |
| "grad_norm": 19.565675735473633, | |
| "learning_rate": 5.804566210045662e-06, | |
| "loss": 0.7604, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.6894977168949772, | |
| "grad_norm": 24.26027488708496, | |
| "learning_rate": 5.781735159817352e-06, | |
| "loss": 0.8224, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.6986301369863015, | |
| "grad_norm": 13.819406509399414, | |
| "learning_rate": 5.7589041095890415e-06, | |
| "loss": 0.7744, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.7077625570776256, | |
| "grad_norm": 20.518844604492188, | |
| "learning_rate": 5.736073059360731e-06, | |
| "loss": 0.7758, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.7168949771689497, | |
| "grad_norm": 13.40528678894043, | |
| "learning_rate": 5.71324200913242e-06, | |
| "loss": 0.7066, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.726027397260274, | |
| "grad_norm": 8.37540054321289, | |
| "learning_rate": 5.69041095890411e-06, | |
| "loss": 0.6852, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.7351598173515983, | |
| "grad_norm": 21.453828811645508, | |
| "learning_rate": 5.6675799086758e-06, | |
| "loss": 0.675, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.7442922374429224, | |
| "grad_norm": 12.05217456817627, | |
| "learning_rate": 5.644748858447489e-06, | |
| "loss": 0.6603, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.7534246575342465, | |
| "grad_norm": 26.813020706176758, | |
| "learning_rate": 5.621917808219178e-06, | |
| "loss": 0.8166, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.7625570776255708, | |
| "grad_norm": 35.790435791015625, | |
| "learning_rate": 5.599086757990868e-06, | |
| "loss": 0.671, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.771689497716895, | |
| "grad_norm": 21.900400161743164, | |
| "learning_rate": 5.576255707762558e-06, | |
| "loss": 0.7431, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.7808219178082192, | |
| "grad_norm": 31.134845733642578, | |
| "learning_rate": 5.553424657534247e-06, | |
| "loss": 0.7594, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.7899543378995433, | |
| "grad_norm": 14.03296947479248, | |
| "learning_rate": 5.530593607305937e-06, | |
| "loss": 0.7507, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.7990867579908676, | |
| "grad_norm": 94.07928466796875, | |
| "learning_rate": 5.507762557077626e-06, | |
| "loss": 0.8151, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.808219178082192, | |
| "grad_norm": 11.925488471984863, | |
| "learning_rate": 5.484931506849316e-06, | |
| "loss": 0.7601, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.817351598173516, | |
| "grad_norm": 13.049914360046387, | |
| "learning_rate": 5.4621004566210045e-06, | |
| "loss": 0.7119, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.82648401826484, | |
| "grad_norm": 16.289236068725586, | |
| "learning_rate": 5.4392694063926946e-06, | |
| "loss": 0.7411, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.82648401826484, | |
| "eval_exact_match": 81.82592242194892, | |
| "eval_f1": 89.05234914493408, | |
| "eval_runtime": 395.3505, | |
| "eval_samples_per_second": 26.736, | |
| "eval_steps_per_second": 1.672, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8356164383561644, | |
| "grad_norm": 25.6737117767334, | |
| "learning_rate": 5.416438356164384e-06, | |
| "loss": 0.7937, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.8447488584474887, | |
| "grad_norm": 17.59447479248047, | |
| "learning_rate": 5.393607305936074e-06, | |
| "loss": 0.7665, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.8538812785388128, | |
| "grad_norm": 11.786163330078125, | |
| "learning_rate": 5.370776255707762e-06, | |
| "loss": 0.811, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.8630136986301369, | |
| "grad_norm": 19.147621154785156, | |
| "learning_rate": 5.3479452054794525e-06, | |
| "loss": 0.7869, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.8721461187214612, | |
| "grad_norm": 25.873458862304688, | |
| "learning_rate": 5.325114155251142e-06, | |
| "loss": 0.8007, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.8812785388127855, | |
| "grad_norm": 21.098203659057617, | |
| "learning_rate": 5.302283105022832e-06, | |
| "loss": 0.7293, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.8904109589041096, | |
| "grad_norm": 12.302563667297363, | |
| "learning_rate": 5.27945205479452e-06, | |
| "loss": 0.712, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.8995433789954337, | |
| "grad_norm": 15.331055641174316, | |
| "learning_rate": 5.25662100456621e-06, | |
| "loss": 0.6805, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.908675799086758, | |
| "grad_norm": 43.43006134033203, | |
| "learning_rate": 5.2337899543379005e-06, | |
| "loss": 0.7232, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.9178082191780823, | |
| "grad_norm": 73.6401596069336, | |
| "learning_rate": 5.21095890410959e-06, | |
| "loss": 0.7803, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9269406392694064, | |
| "grad_norm": 7.5964436531066895, | |
| "learning_rate": 5.188127853881278e-06, | |
| "loss": 0.7418, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.9360730593607305, | |
| "grad_norm": 18.22124671936035, | |
| "learning_rate": 5.165296803652968e-06, | |
| "loss": 0.8099, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.9452054794520548, | |
| "grad_norm": 27.608963012695312, | |
| "learning_rate": 5.142465753424658e-06, | |
| "loss": 0.7369, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.954337899543379, | |
| "grad_norm": 45.187503814697266, | |
| "learning_rate": 5.119634703196348e-06, | |
| "loss": 0.8291, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.9634703196347032, | |
| "grad_norm": 23.95409393310547, | |
| "learning_rate": 5.096803652968037e-06, | |
| "loss": 0.7422, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.9726027397260273, | |
| "grad_norm": 14.64238452911377, | |
| "learning_rate": 5.073972602739726e-06, | |
| "loss": 0.7852, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.9817351598173516, | |
| "grad_norm": 16.381633758544922, | |
| "learning_rate": 5.051141552511416e-06, | |
| "loss": 0.7173, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.990867579908676, | |
| "grad_norm": 16.627071380615234, | |
| "learning_rate": 5.0283105022831055e-06, | |
| "loss": 0.8126, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 23.54231071472168, | |
| "learning_rate": 5.005479452054795e-06, | |
| "loss": 0.7654, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 2.009132420091324, | |
| "grad_norm": 16.47466278076172, | |
| "learning_rate": 4.982648401826484e-06, | |
| "loss": 0.5292, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.018264840182648, | |
| "grad_norm": 20.614046096801758, | |
| "learning_rate": 4.959817351598174e-06, | |
| "loss": 0.5507, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 2.0273972602739727, | |
| "grad_norm": 22.253305435180664, | |
| "learning_rate": 4.93744292237443e-06, | |
| "loss": 0.506, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 2.036529680365297, | |
| "grad_norm": 38.29316711425781, | |
| "learning_rate": 4.9146118721461185e-06, | |
| "loss": 0.5594, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 2.045662100456621, | |
| "grad_norm": 21.850147247314453, | |
| "learning_rate": 4.891780821917809e-06, | |
| "loss": 0.5626, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 2.0547945205479454, | |
| "grad_norm": 13.217841148376465, | |
| "learning_rate": 4.868949771689498e-06, | |
| "loss": 0.5738, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 2.0639269406392695, | |
| "grad_norm": 18.664854049682617, | |
| "learning_rate": 4.846118721461188e-06, | |
| "loss": 0.6301, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 2.0730593607305936, | |
| "grad_norm": 11.781886100769043, | |
| "learning_rate": 4.823287671232877e-06, | |
| "loss": 0.534, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 2.0821917808219177, | |
| "grad_norm": 19.580331802368164, | |
| "learning_rate": 4.8004566210045665e-06, | |
| "loss": 0.5434, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 2.091324200913242, | |
| "grad_norm": 15.862279891967773, | |
| "learning_rate": 4.777625570776256e-06, | |
| "loss": 0.5809, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 2.1004566210045663, | |
| "grad_norm": 35.96525955200195, | |
| "learning_rate": 4.754794520547946e-06, | |
| "loss": 0.5169, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1095890410958904, | |
| "grad_norm": 14.418628692626953, | |
| "learning_rate": 4.731963470319635e-06, | |
| "loss": 0.5428, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 2.1187214611872145, | |
| "grad_norm": 19.833417892456055, | |
| "learning_rate": 4.709132420091324e-06, | |
| "loss": 0.554, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 2.127853881278539, | |
| "grad_norm": 7.7394118309021, | |
| "learning_rate": 4.686301369863014e-06, | |
| "loss": 0.5387, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 2.136986301369863, | |
| "grad_norm": 13.092580795288086, | |
| "learning_rate": 4.663470319634704e-06, | |
| "loss": 0.5263, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 2.146118721461187, | |
| "grad_norm": 11.311351776123047, | |
| "learning_rate": 4.640639269406393e-06, | |
| "loss": 0.5652, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 2.1552511415525113, | |
| "grad_norm": 8.475380897521973, | |
| "learning_rate": 4.617808219178082e-06, | |
| "loss": 0.5572, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 2.1643835616438354, | |
| "grad_norm": 9.318852424621582, | |
| "learning_rate": 4.5949771689497715e-06, | |
| "loss": 0.5704, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 2.17351598173516, | |
| "grad_norm": 17.163570404052734, | |
| "learning_rate": 4.572146118721462e-06, | |
| "loss": 0.5533, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 2.182648401826484, | |
| "grad_norm": 11.318022727966309, | |
| "learning_rate": 4.549315068493151e-06, | |
| "loss": 0.4844, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "grad_norm": 25.644908905029297, | |
| "learning_rate": 4.526484018264841e-06, | |
| "loss": 0.5628, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.191780821917808, | |
| "eval_exact_match": 82.33680227057711, | |
| "eval_f1": 89.23466367243802, | |
| "eval_runtime": 395.5495, | |
| "eval_samples_per_second": 26.722, | |
| "eval_steps_per_second": 1.671, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.2009132420091326, | |
| "grad_norm": 22.05693817138672, | |
| "learning_rate": 4.50365296803653e-06, | |
| "loss": 0.4721, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 2.2100456621004567, | |
| "grad_norm": 48.57001495361328, | |
| "learning_rate": 4.4808219178082195e-06, | |
| "loss": 0.5225, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 2.219178082191781, | |
| "grad_norm": 18.33832359313965, | |
| "learning_rate": 4.457990867579909e-06, | |
| "loss": 0.5864, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 2.228310502283105, | |
| "grad_norm": 17.940277099609375, | |
| "learning_rate": 4.435159817351599e-06, | |
| "loss": 0.5558, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 2.237442922374429, | |
| "grad_norm": 15.108060836791992, | |
| "learning_rate": 4.412328767123288e-06, | |
| "loss": 0.5825, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 2.2465753424657535, | |
| "grad_norm": 87.06202697753906, | |
| "learning_rate": 4.389497716894977e-06, | |
| "loss": 0.5755, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 2.2557077625570776, | |
| "grad_norm": 25.141014099121094, | |
| "learning_rate": 4.366666666666667e-06, | |
| "loss": 0.662, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 2.2648401826484017, | |
| "grad_norm": 17.8304386138916, | |
| "learning_rate": 4.343835616438357e-06, | |
| "loss": 0.5106, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 2.2739726027397262, | |
| "grad_norm": 30.3131160736084, | |
| "learning_rate": 4.321004566210046e-06, | |
| "loss": 0.5553, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 2.2831050228310503, | |
| "grad_norm": 16.719921112060547, | |
| "learning_rate": 4.298173515981735e-06, | |
| "loss": 0.5601, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.2922374429223744, | |
| "grad_norm": 17.55047035217285, | |
| "learning_rate": 4.2753424657534245e-06, | |
| "loss": 0.5909, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 2.3013698630136985, | |
| "grad_norm": 25.325210571289062, | |
| "learning_rate": 4.252511415525115e-06, | |
| "loss": 0.5292, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 2.3105022831050226, | |
| "grad_norm": 7.231600284576416, | |
| "learning_rate": 4.229680365296804e-06, | |
| "loss": 0.5618, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 2.319634703196347, | |
| "grad_norm": 45.177276611328125, | |
| "learning_rate": 4.206849315068494e-06, | |
| "loss": 0.6545, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 2.328767123287671, | |
| "grad_norm": 14.02021312713623, | |
| "learning_rate": 4.184018264840182e-06, | |
| "loss": 0.5223, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 2.3378995433789953, | |
| "grad_norm": 15.296303749084473, | |
| "learning_rate": 4.1611872146118725e-06, | |
| "loss": 0.5214, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 2.34703196347032, | |
| "grad_norm": 12.6821870803833, | |
| "learning_rate": 4.138356164383562e-06, | |
| "loss": 0.5621, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 2.356164383561644, | |
| "grad_norm": 11.630056381225586, | |
| "learning_rate": 4.115525114155252e-06, | |
| "loss": 0.5318, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 2.365296803652968, | |
| "grad_norm": 18.717533111572266, | |
| "learning_rate": 4.092694063926941e-06, | |
| "loss": 0.6217, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 2.374429223744292, | |
| "grad_norm": 18.65459442138672, | |
| "learning_rate": 4.06986301369863e-06, | |
| "loss": 0.5254, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.383561643835616, | |
| "grad_norm": 51.29315948486328, | |
| "learning_rate": 4.04703196347032e-06, | |
| "loss": 0.5516, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 2.3926940639269407, | |
| "grad_norm": 28.213186264038086, | |
| "learning_rate": 4.02420091324201e-06, | |
| "loss": 0.5661, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 2.401826484018265, | |
| "grad_norm": 25.394929885864258, | |
| "learning_rate": 4.001369863013699e-06, | |
| "loss": 0.5811, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 2.410958904109589, | |
| "grad_norm": 11.663569450378418, | |
| "learning_rate": 3.978538812785388e-06, | |
| "loss": 0.5785, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 2.4200913242009134, | |
| "grad_norm": 23.070152282714844, | |
| "learning_rate": 3.9557077625570776e-06, | |
| "loss": 0.5997, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 2.4292237442922375, | |
| "grad_norm": 9.592535018920898, | |
| "learning_rate": 3.932876712328768e-06, | |
| "loss": 0.5743, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 2.4383561643835616, | |
| "grad_norm": 23.151607513427734, | |
| "learning_rate": 3.910045662100457e-06, | |
| "loss": 0.5721, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 2.4474885844748857, | |
| "grad_norm": 22.872838973999023, | |
| "learning_rate": 3.887214611872146e-06, | |
| "loss": 0.5103, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 2.45662100456621, | |
| "grad_norm": 49.325408935546875, | |
| "learning_rate": 3.8643835616438354e-06, | |
| "loss": 0.552, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 13.299301147460938, | |
| "learning_rate": 3.8415525114155256e-06, | |
| "loss": 0.6012, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4748858447488584, | |
| "grad_norm": 10.96218204498291, | |
| "learning_rate": 3.818721461187215e-06, | |
| "loss": 0.5902, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 2.4840182648401825, | |
| "grad_norm": 25.126304626464844, | |
| "learning_rate": 3.7958904109589045e-06, | |
| "loss": 0.5583, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 2.493150684931507, | |
| "grad_norm": 88.76774597167969, | |
| "learning_rate": 3.7730593607305938e-06, | |
| "loss": 0.513, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 2.502283105022831, | |
| "grad_norm": 19.230432510375977, | |
| "learning_rate": 3.7502283105022834e-06, | |
| "loss": 0.5879, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 2.5114155251141552, | |
| "grad_norm": 9.016436576843262, | |
| "learning_rate": 3.7273972602739727e-06, | |
| "loss": 0.5277, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 2.5205479452054793, | |
| "grad_norm": 9.230749130249023, | |
| "learning_rate": 3.7045662100456624e-06, | |
| "loss": 0.6139, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 2.5296803652968034, | |
| "grad_norm": 10.221979141235352, | |
| "learning_rate": 3.6817351598173516e-06, | |
| "loss": 0.5702, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 2.538812785388128, | |
| "grad_norm": 26.807771682739258, | |
| "learning_rate": 3.6589041095890413e-06, | |
| "loss": 0.594, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 2.547945205479452, | |
| "grad_norm": 15.471819877624512, | |
| "learning_rate": 3.6360730593607306e-06, | |
| "loss": 0.665, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 2.557077625570776, | |
| "grad_norm": 29.882976531982422, | |
| "learning_rate": 3.6132420091324207e-06, | |
| "loss": 0.5556, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.557077625570776, | |
| "eval_exact_match": 82.63008514664143, | |
| "eval_f1": 89.62653441826156, | |
| "eval_runtime": 395.4302, | |
| "eval_samples_per_second": 26.73, | |
| "eval_steps_per_second": 1.672, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5662100456621006, | |
| "grad_norm": 20.645360946655273, | |
| "learning_rate": 3.5904109589041095e-06, | |
| "loss": 0.5454, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 2.5753424657534247, | |
| "grad_norm": 22.412696838378906, | |
| "learning_rate": 3.5675799086757996e-06, | |
| "loss": 0.5218, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 2.584474885844749, | |
| "grad_norm": 22.36018180847168, | |
| "learning_rate": 3.544748858447489e-06, | |
| "loss": 0.5433, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 2.593607305936073, | |
| "grad_norm": 30.506481170654297, | |
| "learning_rate": 3.5219178082191786e-06, | |
| "loss": 0.5584, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 2.602739726027397, | |
| "grad_norm": 18.273223876953125, | |
| "learning_rate": 3.499086757990868e-06, | |
| "loss": 0.4913, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 2.6118721461187215, | |
| "grad_norm": 29.807708740234375, | |
| "learning_rate": 3.4762557077625575e-06, | |
| "loss": 0.5721, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 2.6210045662100456, | |
| "grad_norm": 59.90718078613281, | |
| "learning_rate": 3.453881278538813e-06, | |
| "loss": 0.4875, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 2.6301369863013697, | |
| "grad_norm": 8.444619178771973, | |
| "learning_rate": 3.4310502283105023e-06, | |
| "loss": 0.5851, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 2.6392694063926943, | |
| "grad_norm": 119.5136947631836, | |
| "learning_rate": 3.408219178082192e-06, | |
| "loss": 0.5803, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 2.6484018264840183, | |
| "grad_norm": 14.274751663208008, | |
| "learning_rate": 3.3853881278538813e-06, | |
| "loss": 0.5416, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.6575342465753424, | |
| "grad_norm": 23.54309844970703, | |
| "learning_rate": 3.3625570776255714e-06, | |
| "loss": 0.5252, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 17.312332153320312, | |
| "learning_rate": 3.33972602739726e-06, | |
| "loss": 0.5931, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.6757990867579906, | |
| "grad_norm": 23.978649139404297, | |
| "learning_rate": 3.3168949771689503e-06, | |
| "loss": 0.5912, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.684931506849315, | |
| "grad_norm": 27.63211441040039, | |
| "learning_rate": 3.2940639269406396e-06, | |
| "loss": 0.6038, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.6940639269406392, | |
| "grad_norm": 9.839302062988281, | |
| "learning_rate": 3.2712328767123292e-06, | |
| "loss": 0.565, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.7031963470319633, | |
| "grad_norm": 12.635977745056152, | |
| "learning_rate": 3.2484018264840185e-06, | |
| "loss": 0.5678, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.712328767123288, | |
| "grad_norm": 21.576257705688477, | |
| "learning_rate": 3.225570776255708e-06, | |
| "loss": 0.4886, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.721461187214612, | |
| "grad_norm": 9.049654960632324, | |
| "learning_rate": 3.2027397260273974e-06, | |
| "loss": 0.5295, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.730593607305936, | |
| "grad_norm": 54.493343353271484, | |
| "learning_rate": 3.179908675799087e-06, | |
| "loss": 0.5178, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.73972602739726, | |
| "grad_norm": 21.172515869140625, | |
| "learning_rate": 3.1575342465753427e-06, | |
| "loss": 0.6128, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.748858447488584, | |
| "grad_norm": 23.262418746948242, | |
| "learning_rate": 3.134703196347032e-06, | |
| "loss": 0.5992, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.7579908675799087, | |
| "grad_norm": 6.960672855377197, | |
| "learning_rate": 3.1118721461187216e-06, | |
| "loss": 0.5634, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.767123287671233, | |
| "grad_norm": 53.032554626464844, | |
| "learning_rate": 3.089041095890411e-06, | |
| "loss": 0.5313, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.776255707762557, | |
| "grad_norm": 17.943723678588867, | |
| "learning_rate": 3.066210045662101e-06, | |
| "loss": 0.5871, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.7853881278538815, | |
| "grad_norm": 26.733182907104492, | |
| "learning_rate": 3.04337899543379e-06, | |
| "loss": 0.5628, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.7945205479452055, | |
| "grad_norm": 13.597286224365234, | |
| "learning_rate": 3.02054794520548e-06, | |
| "loss": 0.563, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.8036529680365296, | |
| "grad_norm": 23.258501052856445, | |
| "learning_rate": 2.997716894977169e-06, | |
| "loss": 0.5781, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.8127853881278537, | |
| "grad_norm": 32.529056549072266, | |
| "learning_rate": 2.974885844748859e-06, | |
| "loss": 0.5328, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.821917808219178, | |
| "grad_norm": 44.435787200927734, | |
| "learning_rate": 2.952054794520548e-06, | |
| "loss": 0.5254, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.8310502283105023, | |
| "grad_norm": 10.251275062561035, | |
| "learning_rate": 2.929223744292238e-06, | |
| "loss": 0.5194, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.8401826484018264, | |
| "grad_norm": 24.379085540771484, | |
| "learning_rate": 2.906392694063927e-06, | |
| "loss": 0.5232, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.8493150684931505, | |
| "grad_norm": 32.740516662597656, | |
| "learning_rate": 2.8835616438356167e-06, | |
| "loss": 0.5731, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.858447488584475, | |
| "grad_norm": 24.7740478515625, | |
| "learning_rate": 2.860730593607306e-06, | |
| "loss": 0.5251, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.867579908675799, | |
| "grad_norm": 29.77228355407715, | |
| "learning_rate": 2.8378995433789957e-06, | |
| "loss": 0.5457, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.8767123287671232, | |
| "grad_norm": 24.371246337890625, | |
| "learning_rate": 2.815068493150685e-06, | |
| "loss": 0.574, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.8858447488584473, | |
| "grad_norm": 10.619139671325684, | |
| "learning_rate": 2.7922374429223746e-06, | |
| "loss": 0.5337, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.8949771689497714, | |
| "grad_norm": 31.601932525634766, | |
| "learning_rate": 2.769406392694064e-06, | |
| "loss": 0.5693, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.904109589041096, | |
| "grad_norm": 28.158267974853516, | |
| "learning_rate": 2.7465753424657536e-06, | |
| "loss": 0.561, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.91324200913242, | |
| "grad_norm": 23.712093353271484, | |
| "learning_rate": 2.723744292237443e-06, | |
| "loss": 0.5312, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.922374429223744, | |
| "grad_norm": 16.564476013183594, | |
| "learning_rate": 2.700913242009133e-06, | |
| "loss": 0.4938, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.922374429223744, | |
| "eval_exact_match": 82.58278145695364, | |
| "eval_f1": 89.75516423119765, | |
| "eval_runtime": 395.7187, | |
| "eval_samples_per_second": 26.711, | |
| "eval_steps_per_second": 1.67, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9315068493150687, | |
| "grad_norm": 10.199018478393555, | |
| "learning_rate": 2.6780821917808218e-06, | |
| "loss": 0.5761, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.9406392694063928, | |
| "grad_norm": 12.591775894165039, | |
| "learning_rate": 2.655251141552512e-06, | |
| "loss": 0.5681, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.949771689497717, | |
| "grad_norm": 20.624439239501953, | |
| "learning_rate": 2.632420091324201e-06, | |
| "loss": 0.4934, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.958904109589041, | |
| "grad_norm": 38.65128707885742, | |
| "learning_rate": 2.609589041095891e-06, | |
| "loss": 0.5956, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.968036529680365, | |
| "grad_norm": 24.270273208618164, | |
| "learning_rate": 2.58675799086758e-06, | |
| "loss": 0.5219, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.9771689497716896, | |
| "grad_norm": 24.47715950012207, | |
| "learning_rate": 2.5639269406392698e-06, | |
| "loss": 0.5067, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.9863013698630136, | |
| "grad_norm": 7.162631988525391, | |
| "learning_rate": 2.541095890410959e-06, | |
| "loss": 0.5377, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.9954337899543377, | |
| "grad_norm": 9.583894729614258, | |
| "learning_rate": 2.5182648401826487e-06, | |
| "loss": 0.5518, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 3.0045662100456623, | |
| "grad_norm": 25.51879119873047, | |
| "learning_rate": 2.4958904109589042e-06, | |
| "loss": 0.4965, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 3.0136986301369864, | |
| "grad_norm": 142.86093139648438, | |
| "learning_rate": 2.473059360730594e-06, | |
| "loss": 0.4489, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.0228310502283104, | |
| "grad_norm": 13.863718032836914, | |
| "learning_rate": 2.450228310502283e-06, | |
| "loss": 0.4735, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 3.0319634703196345, | |
| "grad_norm": 12.609000205993652, | |
| "learning_rate": 2.427397260273973e-06, | |
| "loss": 0.4123, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 3.041095890410959, | |
| "grad_norm": 12.495716094970703, | |
| "learning_rate": 2.4045662100456625e-06, | |
| "loss": 0.4467, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 3.050228310502283, | |
| "grad_norm": 23.454723358154297, | |
| "learning_rate": 2.381735159817352e-06, | |
| "loss": 0.4544, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 3.0593607305936072, | |
| "grad_norm": 21.551776885986328, | |
| "learning_rate": 2.3589041095890415e-06, | |
| "loss": 0.4199, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 3.0684931506849313, | |
| "grad_norm": 29.826936721801758, | |
| "learning_rate": 2.3360730593607308e-06, | |
| "loss": 0.4447, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 3.077625570776256, | |
| "grad_norm": 24.2338809967041, | |
| "learning_rate": 2.3132420091324204e-06, | |
| "loss": 0.4281, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 3.08675799086758, | |
| "grad_norm": 12.561247825622559, | |
| "learning_rate": 2.2904109589041097e-06, | |
| "loss": 0.4572, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 3.095890410958904, | |
| "grad_norm": 26.721580505371094, | |
| "learning_rate": 2.2675799086757994e-06, | |
| "loss": 0.407, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 3.105022831050228, | |
| "grad_norm": 53.11214828491211, | |
| "learning_rate": 2.244748858447489e-06, | |
| "loss": 0.4183, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.1141552511415527, | |
| "grad_norm": 13.374695777893066, | |
| "learning_rate": 2.2219178082191783e-06, | |
| "loss": 0.4651, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 3.1232876712328768, | |
| "grad_norm": 17.132118225097656, | |
| "learning_rate": 2.199086757990868e-06, | |
| "loss": 0.3833, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 3.132420091324201, | |
| "grad_norm": 12.751568794250488, | |
| "learning_rate": 2.1762557077625573e-06, | |
| "loss": 0.4563, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 3.141552511415525, | |
| "grad_norm": 11.56992244720459, | |
| "learning_rate": 2.153424657534247e-06, | |
| "loss": 0.4315, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 3.1506849315068495, | |
| "grad_norm": 18.173213958740234, | |
| "learning_rate": 2.130593607305936e-06, | |
| "loss": 0.4013, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 3.1598173515981736, | |
| "grad_norm": 15.345487594604492, | |
| "learning_rate": 2.107762557077626e-06, | |
| "loss": 0.4176, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 3.1689497716894977, | |
| "grad_norm": 11.01826286315918, | |
| "learning_rate": 2.084931506849315e-06, | |
| "loss": 0.4426, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 3.1780821917808217, | |
| "grad_norm": 27.82256507873535, | |
| "learning_rate": 2.062100456621005e-06, | |
| "loss": 0.4511, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 3.1872146118721463, | |
| "grad_norm": 23.4956111907959, | |
| "learning_rate": 2.0392694063926945e-06, | |
| "loss": 0.4195, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 3.1963470319634704, | |
| "grad_norm": 15.364335060119629, | |
| "learning_rate": 2.0164383561643838e-06, | |
| "loss": 0.423, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.2054794520547945, | |
| "grad_norm": 6.171648025512695, | |
| "learning_rate": 1.9936073059360735e-06, | |
| "loss": 0.4323, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 3.2146118721461185, | |
| "grad_norm": 10.051441192626953, | |
| "learning_rate": 1.9707762557077627e-06, | |
| "loss": 0.4307, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 3.223744292237443, | |
| "grad_norm": 15.94997501373291, | |
| "learning_rate": 1.9479452054794524e-06, | |
| "loss": 0.4034, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 3.232876712328767, | |
| "grad_norm": 43.07652282714844, | |
| "learning_rate": 1.9251141552511417e-06, | |
| "loss": 0.4106, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 3.2420091324200913, | |
| "grad_norm": 17.467111587524414, | |
| "learning_rate": 1.9022831050228313e-06, | |
| "loss": 0.4648, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 3.2511415525114153, | |
| "grad_norm": 10.493595123291016, | |
| "learning_rate": 1.8794520547945208e-06, | |
| "loss": 0.4619, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 3.26027397260274, | |
| "grad_norm": 10.744498252868652, | |
| "learning_rate": 1.8566210045662103e-06, | |
| "loss": 0.3861, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 3.269406392694064, | |
| "grad_norm": 12.96605110168457, | |
| "learning_rate": 1.8337899543378998e-06, | |
| "loss": 0.4077, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 3.278538812785388, | |
| "grad_norm": 6.190354347229004, | |
| "learning_rate": 1.8109589041095892e-06, | |
| "loss": 0.3937, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 3.287671232876712, | |
| "grad_norm": 15.968260765075684, | |
| "learning_rate": 1.7881278538812787e-06, | |
| "loss": 0.4852, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.287671232876712, | |
| "eval_exact_match": 82.42194891201514, | |
| "eval_f1": 89.61372740496692, | |
| "eval_runtime": 413.8098, | |
| "eval_samples_per_second": 25.543, | |
| "eval_steps_per_second": 1.597, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.2968036529680367, | |
| "grad_norm": 39.12847900390625, | |
| "learning_rate": 1.7652968036529684e-06, | |
| "loss": 0.3748, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 3.3059360730593608, | |
| "grad_norm": 19.359882354736328, | |
| "learning_rate": 1.7424657534246579e-06, | |
| "loss": 0.4165, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 3.315068493150685, | |
| "grad_norm": 6.587915897369385, | |
| "learning_rate": 1.7196347031963473e-06, | |
| "loss": 0.4489, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 3.324200913242009, | |
| "grad_norm": 10.52927017211914, | |
| "learning_rate": 1.6968036529680368e-06, | |
| "loss": 0.4074, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 25.514862060546875, | |
| "learning_rate": 1.6739726027397263e-06, | |
| "loss": 0.4552, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 3.3424657534246576, | |
| "grad_norm": 12.100693702697754, | |
| "learning_rate": 1.6511415525114157e-06, | |
| "loss": 0.4463, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 3.3515981735159817, | |
| "grad_norm": 31.30626106262207, | |
| "learning_rate": 1.6283105022831052e-06, | |
| "loss": 0.393, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 3.3607305936073057, | |
| "grad_norm": 19.230247497558594, | |
| "learning_rate": 1.6054794520547947e-06, | |
| "loss": 0.4318, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 3.3698630136986303, | |
| "grad_norm": 10.644558906555176, | |
| "learning_rate": 1.5826484018264844e-06, | |
| "loss": 0.453, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 3.3789954337899544, | |
| "grad_norm": 9.888466835021973, | |
| "learning_rate": 1.5598173515981738e-06, | |
| "loss": 0.4205, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.3881278538812785, | |
| "grad_norm": 14.341341972351074, | |
| "learning_rate": 1.5369863013698633e-06, | |
| "loss": 0.4418, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 3.3972602739726026, | |
| "grad_norm": 26.137784957885742, | |
| "learning_rate": 1.5141552511415528e-06, | |
| "loss": 0.421, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 3.406392694063927, | |
| "grad_norm": 8.944993019104004, | |
| "learning_rate": 1.4913242009132423e-06, | |
| "loss": 0.4003, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 3.415525114155251, | |
| "grad_norm": 49.0703125, | |
| "learning_rate": 1.4684931506849317e-06, | |
| "loss": 0.4567, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 3.4246575342465753, | |
| "grad_norm": 23.05057716369629, | |
| "learning_rate": 1.4456621004566212e-06, | |
| "loss": 0.4215, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 3.4337899543378994, | |
| "grad_norm": 21.53809356689453, | |
| "learning_rate": 1.4228310502283107e-06, | |
| "loss": 0.3586, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 3.442922374429224, | |
| "grad_norm": 13.685781478881836, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.3776, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 3.452054794520548, | |
| "grad_norm": 20.39994239807129, | |
| "learning_rate": 1.3771689497716898e-06, | |
| "loss": 0.4484, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 3.461187214611872, | |
| "grad_norm": 35.2899169921875, | |
| "learning_rate": 1.3543378995433793e-06, | |
| "loss": 0.4426, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 3.470319634703196, | |
| "grad_norm": 54.691349029541016, | |
| "learning_rate": 1.3315068493150688e-06, | |
| "loss": 0.4736, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.4794520547945207, | |
| "grad_norm": 13.618513107299805, | |
| "learning_rate": 1.3086757990867582e-06, | |
| "loss": 0.377, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 3.4885844748858448, | |
| "grad_norm": 23.14171600341797, | |
| "learning_rate": 1.2858447488584475e-06, | |
| "loss": 0.4071, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 3.497716894977169, | |
| "grad_norm": 25.513710021972656, | |
| "learning_rate": 1.263013698630137e-06, | |
| "loss": 0.385, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 3.506849315068493, | |
| "grad_norm": 15.305421829223633, | |
| "learning_rate": 1.2401826484018267e-06, | |
| "loss": 0.4366, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 3.5159817351598175, | |
| "grad_norm": 15.072031021118164, | |
| "learning_rate": 1.2173515981735161e-06, | |
| "loss": 0.4637, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 3.5251141552511416, | |
| "grad_norm": 16.01192283630371, | |
| "learning_rate": 1.1945205479452056e-06, | |
| "loss": 0.4313, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 3.5342465753424657, | |
| "grad_norm": 18.003978729248047, | |
| "learning_rate": 1.1721461187214613e-06, | |
| "loss": 0.4291, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 3.54337899543379, | |
| "grad_norm": 33.20173263549805, | |
| "learning_rate": 1.1493150684931508e-06, | |
| "loss": 0.4792, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 3.5525114155251143, | |
| "grad_norm": 11.555415153503418, | |
| "learning_rate": 1.1264840182648403e-06, | |
| "loss": 0.4561, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 3.5616438356164384, | |
| "grad_norm": 12.765213966369629, | |
| "learning_rate": 1.1036529680365298e-06, | |
| "loss": 0.4258, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.5707762557077625, | |
| "grad_norm": 14.976128578186035, | |
| "learning_rate": 1.0808219178082192e-06, | |
| "loss": 0.4376, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 3.5799086757990866, | |
| "grad_norm": 21.439231872558594, | |
| "learning_rate": 1.0579908675799087e-06, | |
| "loss": 0.451, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 3.589041095890411, | |
| "grad_norm": 13.795149803161621, | |
| "learning_rate": 1.0351598173515982e-06, | |
| "loss": 0.389, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 3.598173515981735, | |
| "grad_norm": 10.679841995239258, | |
| "learning_rate": 1.0123287671232876e-06, | |
| "loss": 0.4535, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 3.6073059360730593, | |
| "grad_norm": 18.03670310974121, | |
| "learning_rate": 9.894977168949773e-07, | |
| "loss": 0.4252, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 3.616438356164384, | |
| "grad_norm": 68.05775451660156, | |
| "learning_rate": 9.666666666666668e-07, | |
| "loss": 0.4625, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 3.625570776255708, | |
| "grad_norm": 32.91834259033203, | |
| "learning_rate": 9.438356164383563e-07, | |
| "loss": 0.4295, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 3.634703196347032, | |
| "grad_norm": 34.28315734863281, | |
| "learning_rate": 9.210045662100457e-07, | |
| "loss": 0.3863, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 3.643835616438356, | |
| "grad_norm": 13.401795387268066, | |
| "learning_rate": 8.981735159817352e-07, | |
| "loss": 0.4354, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 3.65296803652968, | |
| "grad_norm": 9.942870140075684, | |
| "learning_rate": 8.753424657534247e-07, | |
| "loss": 0.3813, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.65296803652968, | |
| "eval_exact_match": 82.50709555345317, | |
| "eval_f1": 89.67676155395135, | |
| "eval_runtime": 810.5169, | |
| "eval_samples_per_second": 13.041, | |
| "eval_steps_per_second": 0.816, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.6621004566210047, | |
| "grad_norm": 33.140907287597656, | |
| "learning_rate": 8.525114155251143e-07, | |
| "loss": 0.4385, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 3.671232876712329, | |
| "grad_norm": 86.72850799560547, | |
| "learning_rate": 8.296803652968037e-07, | |
| "loss": 0.4785, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 3.680365296803653, | |
| "grad_norm": 11.17636775970459, | |
| "learning_rate": 8.068493150684932e-07, | |
| "loss": 0.4134, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 3.6894977168949774, | |
| "grad_norm": 14.498661994934082, | |
| "learning_rate": 7.840182648401827e-07, | |
| "loss": 0.3903, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 3.6986301369863015, | |
| "grad_norm": 18.49641227722168, | |
| "learning_rate": 7.611872146118722e-07, | |
| "loss": 0.4131, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 3.7077625570776256, | |
| "grad_norm": 23.517004013061523, | |
| "learning_rate": 7.383561643835617e-07, | |
| "loss": 0.4074, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 3.7168949771689497, | |
| "grad_norm": 14.428693771362305, | |
| "learning_rate": 7.155251141552512e-07, | |
| "loss": 0.3898, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 3.7260273972602738, | |
| "grad_norm": 13.074502944946289, | |
| "learning_rate": 6.926940639269407e-07, | |
| "loss": 0.4251, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 3.7351598173515983, | |
| "grad_norm": 10.06019401550293, | |
| "learning_rate": 6.698630136986301e-07, | |
| "loss": 0.409, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 3.7442922374429224, | |
| "grad_norm": 17.26006507873535, | |
| "learning_rate": 6.470319634703197e-07, | |
| "loss": 0.374, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.7534246575342465, | |
| "grad_norm": 19.47539710998535, | |
| "learning_rate": 6.242009132420092e-07, | |
| "loss": 0.3683, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 3.762557077625571, | |
| "grad_norm": 20.99232292175293, | |
| "learning_rate": 6.013698630136987e-07, | |
| "loss": 0.4675, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 3.771689497716895, | |
| "grad_norm": 48.530582427978516, | |
| "learning_rate": 5.785388127853881e-07, | |
| "loss": 0.4275, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 3.780821917808219, | |
| "grad_norm": 15.112560272216797, | |
| "learning_rate": 5.557077625570777e-07, | |
| "loss": 0.38, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 3.7899543378995433, | |
| "grad_norm": 13.57143497467041, | |
| "learning_rate": 5.328767123287672e-07, | |
| "loss": 0.3885, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 3.7990867579908674, | |
| "grad_norm": 24.962566375732422, | |
| "learning_rate": 5.100456621004567e-07, | |
| "loss": 0.4148, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 3.808219178082192, | |
| "grad_norm": 35.816993713378906, | |
| "learning_rate": 4.872146118721461e-07, | |
| "loss": 0.4211, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 3.817351598173516, | |
| "grad_norm": 29.423664093017578, | |
| "learning_rate": 4.6438356164383565e-07, | |
| "loss": 0.5014, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 3.82648401826484, | |
| "grad_norm": 39.580039978027344, | |
| "learning_rate": 4.415525114155252e-07, | |
| "loss": 0.4521, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 3.8356164383561646, | |
| "grad_norm": 6.263591766357422, | |
| "learning_rate": 4.1872146118721465e-07, | |
| "loss": 0.452, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.8447488584474887, | |
| "grad_norm": 68.41936492919922, | |
| "learning_rate": 3.958904109589042e-07, | |
| "loss": 0.4276, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 3.853881278538813, | |
| "grad_norm": 22.434158325195312, | |
| "learning_rate": 3.7305936073059365e-07, | |
| "loss": 0.4951, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 3.863013698630137, | |
| "grad_norm": 16.98661231994629, | |
| "learning_rate": 3.5022831050228317e-07, | |
| "loss": 0.4212, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 3.872146118721461, | |
| "grad_norm": 12.409255981445312, | |
| "learning_rate": 3.2739726027397264e-07, | |
| "loss": 0.4156, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 3.8812785388127855, | |
| "grad_norm": 26.019704818725586, | |
| "learning_rate": 3.0456621004566217e-07, | |
| "loss": 0.4105, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 3.8904109589041096, | |
| "grad_norm": 28.032352447509766, | |
| "learning_rate": 2.817351598173516e-07, | |
| "loss": 0.4564, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 3.8995433789954337, | |
| "grad_norm": 18.951528549194336, | |
| "learning_rate": 2.589041095890411e-07, | |
| "loss": 0.4687, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 3.908675799086758, | |
| "grad_norm": 36.14161682128906, | |
| "learning_rate": 2.360730593607306e-07, | |
| "loss": 0.4194, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 3.9178082191780823, | |
| "grad_norm": 22.935070037841797, | |
| "learning_rate": 2.132420091324201e-07, | |
| "loss": 0.4582, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 3.9269406392694064, | |
| "grad_norm": 15.434907913208008, | |
| "learning_rate": 1.904109589041096e-07, | |
| "loss": 0.3789, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.9360730593607305, | |
| "grad_norm": 5.678657054901123, | |
| "learning_rate": 1.675799086757991e-07, | |
| "loss": 0.4196, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 3.9452054794520546, | |
| "grad_norm": 14.900694847106934, | |
| "learning_rate": 1.447488584474886e-07, | |
| "loss": 0.4129, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 3.954337899543379, | |
| "grad_norm": 13.7803316116333, | |
| "learning_rate": 1.2191780821917807e-07, | |
| "loss": 0.4436, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 3.963470319634703, | |
| "grad_norm": 15.580832481384277, | |
| "learning_rate": 9.908675799086758e-08, | |
| "loss": 0.3984, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 3.9726027397260273, | |
| "grad_norm": 12.353429794311523, | |
| "learning_rate": 7.625570776255708e-08, | |
| "loss": 0.4279, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 3.981735159817352, | |
| "grad_norm": 10.456271171569824, | |
| "learning_rate": 5.3424657534246586e-08, | |
| "loss": 0.4438, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 3.990867579908676, | |
| "grad_norm": 17.142772674560547, | |
| "learning_rate": 3.059360730593608e-08, | |
| "loss": 0.4569, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.582050323486328, | |
| "learning_rate": 7.762557077625571e-09, | |
| "loss": 0.4129, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 21900, | |
| "total_flos": 2.6750162778225377e+18, | |
| "train_loss": 0.7768389247214957, | |
| "train_runtime": 96436.1031, | |
| "train_samples_per_second": 3.633, | |
| "train_steps_per_second": 0.227 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 21900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 2.6750162778225377e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |