| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9976711690731253, | |
| "eval_steps": 500, | |
| "global_step": 3219, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004657661853749418, | |
| "grad_norm": 82.12092468163539, | |
| "learning_rate": 7.763975155279503e-07, | |
| "loss": 10.4849, | |
| "num_tokens": 5242880.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.009315323707498836, | |
| "grad_norm": 98.320792520794, | |
| "learning_rate": 1.5527950310559006e-06, | |
| "loss": 10.1636, | |
| "num_tokens": 10485760.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013972985561248253, | |
| "grad_norm": 158.09089905651268, | |
| "learning_rate": 2.329192546583851e-06, | |
| "loss": 8.0718, | |
| "num_tokens": 15691438.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.018630647414997672, | |
| "grad_norm": 25.135264827961844, | |
| "learning_rate": 3.1055900621118013e-06, | |
| "loss": 2.4027, | |
| "num_tokens": 20915252.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02328830926874709, | |
| "grad_norm": 4.229120769856315, | |
| "learning_rate": 3.881987577639752e-06, | |
| "loss": 1.4239, | |
| "num_tokens": 26134828.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.027945971122496506, | |
| "grad_norm": 1.4417480815272306, | |
| "learning_rate": 4.658385093167702e-06, | |
| "loss": 1.116, | |
| "num_tokens": 31377708.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.032603632976245925, | |
| "grad_norm": 0.8895502352161005, | |
| "learning_rate": 5.4347826086956525e-06, | |
| "loss": 0.9274, | |
| "num_tokens": 36557688.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.037261294829995344, | |
| "grad_norm": 0.7606006196045029, | |
| "learning_rate": 6.2111801242236025e-06, | |
| "loss": 0.8235, | |
| "num_tokens": 41800322.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04191895668374476, | |
| "grad_norm": 0.4707900398683412, | |
| "learning_rate": 6.9875776397515525e-06, | |
| "loss": 0.7182, | |
| "num_tokens": 47043202.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04657661853749418, | |
| "grad_norm": 0.4681723308735776, | |
| "learning_rate": 7.763975155279503e-06, | |
| "loss": 0.6927, | |
| "num_tokens": 52286082.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05123428039124359, | |
| "grad_norm": 0.35978087935261616, | |
| "learning_rate": 8.540372670807453e-06, | |
| "loss": 0.6581, | |
| "num_tokens": 57528962.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05589194224499301, | |
| "grad_norm": 0.3485647823977247, | |
| "learning_rate": 9.316770186335403e-06, | |
| "loss": 0.6389, | |
| "num_tokens": 62771842.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06054960409874243, | |
| "grad_norm": 0.29715248155962004, | |
| "learning_rate": 1.0093167701863353e-05, | |
| "loss": 0.6195, | |
| "num_tokens": 67980354.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06520726595249185, | |
| "grad_norm": 0.27686969511133136, | |
| "learning_rate": 1.0869565217391305e-05, | |
| "loss": 0.5873, | |
| "num_tokens": 73223234.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06986492780624126, | |
| "grad_norm": 0.29819893243181794, | |
| "learning_rate": 1.1645962732919255e-05, | |
| "loss": 0.5836, | |
| "num_tokens": 78437814.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07452258965999069, | |
| "grad_norm": 0.30051884104149545, | |
| "learning_rate": 1.2422360248447205e-05, | |
| "loss": 0.5911, | |
| "num_tokens": 83680694.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0791802515137401, | |
| "grad_norm": 0.3074376375864693, | |
| "learning_rate": 1.3198757763975155e-05, | |
| "loss": 0.5594, | |
| "num_tokens": 88923574.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08383791336748952, | |
| "grad_norm": 0.3014931535858037, | |
| "learning_rate": 1.3975155279503105e-05, | |
| "loss": 0.5618, | |
| "num_tokens": 94166454.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08849557522123894, | |
| "grad_norm": 0.2549176193333227, | |
| "learning_rate": 1.4751552795031057e-05, | |
| "loss": 0.5672, | |
| "num_tokens": 99409334.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09315323707498836, | |
| "grad_norm": 0.2780121707349434, | |
| "learning_rate": 1.5527950310559007e-05, | |
| "loss": 0.5529, | |
| "num_tokens": 104652214.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09781089892873777, | |
| "grad_norm": 0.2743186816350753, | |
| "learning_rate": 1.630434782608696e-05, | |
| "loss": 0.55, | |
| "num_tokens": 109895094.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10246856078248719, | |
| "grad_norm": 0.2715732620875137, | |
| "learning_rate": 1.7080745341614907e-05, | |
| "loss": 0.535, | |
| "num_tokens": 115137974.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10712622263623661, | |
| "grad_norm": 0.2718170301749333, | |
| "learning_rate": 1.785714285714286e-05, | |
| "loss": 0.533, | |
| "num_tokens": 120380854.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11178388448998602, | |
| "grad_norm": 0.2858665286125675, | |
| "learning_rate": 1.8633540372670807e-05, | |
| "loss": 0.5443, | |
| "num_tokens": 125623734.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11644154634373545, | |
| "grad_norm": 0.28972170081341825, | |
| "learning_rate": 1.940993788819876e-05, | |
| "loss": 0.5182, | |
| "num_tokens": 130866614.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12109920819748486, | |
| "grad_norm": 0.30798443431324013, | |
| "learning_rate": 2.0186335403726707e-05, | |
| "loss": 0.5192, | |
| "num_tokens": 136068456.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1257568700512343, | |
| "grad_norm": 0.32977866738670863, | |
| "learning_rate": 2.096273291925466e-05, | |
| "loss": 0.5268, | |
| "num_tokens": 141311336.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1304145319049837, | |
| "grad_norm": 0.27904302302889866, | |
| "learning_rate": 2.173913043478261e-05, | |
| "loss": 0.5162, | |
| "num_tokens": 146554216.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1350721937587331, | |
| "grad_norm": 0.3080103359309234, | |
| "learning_rate": 2.2515527950310562e-05, | |
| "loss": 0.527, | |
| "num_tokens": 151797096.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.13972985561248252, | |
| "grad_norm": 0.3277378616458447, | |
| "learning_rate": 2.329192546583851e-05, | |
| "loss": 0.5177, | |
| "num_tokens": 157039976.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.14438751746623196, | |
| "grad_norm": 0.30741986520380804, | |
| "learning_rate": 2.4068322981366462e-05, | |
| "loss": 0.4976, | |
| "num_tokens": 162282856.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.14904517931998137, | |
| "grad_norm": 0.36421433688652194, | |
| "learning_rate": 2.484472049689441e-05, | |
| "loss": 0.5062, | |
| "num_tokens": 167466294.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1537028411737308, | |
| "grad_norm": 0.3923856818891026, | |
| "learning_rate": 2.5621118012422362e-05, | |
| "loss": 0.5029, | |
| "num_tokens": 172703130.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1583605030274802, | |
| "grad_norm": 0.38300198750223347, | |
| "learning_rate": 2.639751552795031e-05, | |
| "loss": 0.5007, | |
| "num_tokens": 177946010.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1630181648812296, | |
| "grad_norm": 0.40320303249503864, | |
| "learning_rate": 2.7173913043478262e-05, | |
| "loss": 0.5124, | |
| "num_tokens": 183145930.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.16767582673497905, | |
| "grad_norm": 0.3423947139334269, | |
| "learning_rate": 2.795031055900621e-05, | |
| "loss": 0.5037, | |
| "num_tokens": 188388810.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.17233348858872846, | |
| "grad_norm": 0.3556666362657707, | |
| "learning_rate": 2.8726708074534165e-05, | |
| "loss": 0.504, | |
| "num_tokens": 193628148.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.17699115044247787, | |
| "grad_norm": 0.40080744776481453, | |
| "learning_rate": 2.9503105590062114e-05, | |
| "loss": 0.4952, | |
| "num_tokens": 198871028.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.18164881229622729, | |
| "grad_norm": 0.36733063743089683, | |
| "learning_rate": 3.0279503105590062e-05, | |
| "loss": 0.5003, | |
| "num_tokens": 204113908.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.18630647414997673, | |
| "grad_norm": 0.3861047834029622, | |
| "learning_rate": 3.1055900621118014e-05, | |
| "loss": 0.4704, | |
| "num_tokens": 209266418.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19096413600372614, | |
| "grad_norm": 0.6189173524324959, | |
| "learning_rate": 3.183229813664597e-05, | |
| "loss": 0.4927, | |
| "num_tokens": 214355738.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.19562179785747555, | |
| "grad_norm": 0.47772587508912245, | |
| "learning_rate": 3.260869565217392e-05, | |
| "loss": 0.4807, | |
| "num_tokens": 219598618.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.20027945971122496, | |
| "grad_norm": 0.5283740796756944, | |
| "learning_rate": 3.3385093167701865e-05, | |
| "loss": 0.4925, | |
| "num_tokens": 224779924.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.20493712156497437, | |
| "grad_norm": 0.4202170263910499, | |
| "learning_rate": 3.4161490683229814e-05, | |
| "loss": 0.4882, | |
| "num_tokens": 230001220.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2095947834187238, | |
| "grad_norm": 0.34791401171097375, | |
| "learning_rate": 3.493788819875777e-05, | |
| "loss": 0.4882, | |
| "num_tokens": 235244100.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.21425244527247322, | |
| "grad_norm": 0.30832335986698484, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.487, | |
| "num_tokens": 240459436.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21891010712622264, | |
| "grad_norm": 0.29886980753100034, | |
| "learning_rate": 3.6490683229813665e-05, | |
| "loss": 0.4804, | |
| "num_tokens": 245702316.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.22356776897997205, | |
| "grad_norm": 0.35599557732154424, | |
| "learning_rate": 3.7267080745341614e-05, | |
| "loss": 0.48, | |
| "num_tokens": 250945196.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.22822543083372146, | |
| "grad_norm": 0.29094898667165214, | |
| "learning_rate": 3.804347826086957e-05, | |
| "loss": 0.4671, | |
| "num_tokens": 256130440.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2328830926874709, | |
| "grad_norm": 0.34049539691430897, | |
| "learning_rate": 3.881987577639752e-05, | |
| "loss": 0.4634, | |
| "num_tokens": 261373320.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2375407545412203, | |
| "grad_norm": 0.2947606929142029, | |
| "learning_rate": 3.9596273291925465e-05, | |
| "loss": 0.4769, | |
| "num_tokens": 266595056.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.24219841639496972, | |
| "grad_norm": 0.30044046548629566, | |
| "learning_rate": 4.0372670807453414e-05, | |
| "loss": 0.4806, | |
| "num_tokens": 271837936.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.24685607824871914, | |
| "grad_norm": 0.33115256549687727, | |
| "learning_rate": 4.114906832298137e-05, | |
| "loss": 0.4737, | |
| "num_tokens": 277080816.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2515137401024686, | |
| "grad_norm": 0.3886276423561874, | |
| "learning_rate": 4.192546583850932e-05, | |
| "loss": 0.4619, | |
| "num_tokens": 282323696.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.25617140195621796, | |
| "grad_norm": 0.3444718966849633, | |
| "learning_rate": 4.270186335403727e-05, | |
| "loss": 0.4784, | |
| "num_tokens": 287548638.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2608290638099674, | |
| "grad_norm": 0.4646518224546721, | |
| "learning_rate": 4.347826086956522e-05, | |
| "loss": 0.4667, | |
| "num_tokens": 292791518.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.26548672566371684, | |
| "grad_norm": 0.45335655630825383, | |
| "learning_rate": 4.425465838509317e-05, | |
| "loss": 0.4643, | |
| "num_tokens": 298034398.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2701443875174662, | |
| "grad_norm": 0.4013684893199347, | |
| "learning_rate": 4.5031055900621124e-05, | |
| "loss": 0.4623, | |
| "num_tokens": 303277278.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.27480204937121566, | |
| "grad_norm": 0.5041343435622528, | |
| "learning_rate": 4.580745341614907e-05, | |
| "loss": 0.4558, | |
| "num_tokens": 308520158.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.27945971122496505, | |
| "grad_norm": 0.4011936210144379, | |
| "learning_rate": 4.658385093167702e-05, | |
| "loss": 0.4697, | |
| "num_tokens": 313763038.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2841173730787145, | |
| "grad_norm": 0.4440241081390936, | |
| "learning_rate": 4.736024844720497e-05, | |
| "loss": 0.4752, | |
| "num_tokens": 318941236.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.2887750349324639, | |
| "grad_norm": 0.3812174168901137, | |
| "learning_rate": 4.8136645962732924e-05, | |
| "loss": 0.4694, | |
| "num_tokens": 324184116.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2934326967862133, | |
| "grad_norm": 0.364267368765205, | |
| "learning_rate": 4.891304347826087e-05, | |
| "loss": 0.4584, | |
| "num_tokens": 329426996.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.29809035863996275, | |
| "grad_norm": 0.37844475840024255, | |
| "learning_rate": 4.968944099378882e-05, | |
| "loss": 0.466, | |
| "num_tokens": 334669876.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.30274802049371213, | |
| "grad_norm": 0.4407288220956239, | |
| "learning_rate": 4.994822229892993e-05, | |
| "loss": 0.4678, | |
| "num_tokens": 339912756.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3074056823474616, | |
| "grad_norm": 0.4591495159705848, | |
| "learning_rate": 4.986192613047981e-05, | |
| "loss": 0.4673, | |
| "num_tokens": 345155636.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.312063344201211, | |
| "grad_norm": 0.329700517411992, | |
| "learning_rate": 4.977562996202969e-05, | |
| "loss": 0.4508, | |
| "num_tokens": 350398516.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3167210060549604, | |
| "grad_norm": 0.3012066875460979, | |
| "learning_rate": 4.968933379357957e-05, | |
| "loss": 0.4642, | |
| "num_tokens": 355641396.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.32137866790870984, | |
| "grad_norm": 0.34142093884298497, | |
| "learning_rate": 4.9603037625129445e-05, | |
| "loss": 0.4517, | |
| "num_tokens": 360884276.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3260363297624592, | |
| "grad_norm": 0.3845229834868296, | |
| "learning_rate": 4.951674145667933e-05, | |
| "loss": 0.4541, | |
| "num_tokens": 366127156.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.33069399161620866, | |
| "grad_norm": 0.44258567239380464, | |
| "learning_rate": 4.94304452882292e-05, | |
| "loss": 0.4509, | |
| "num_tokens": 371370036.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3353516534699581, | |
| "grad_norm": 0.3336062988980636, | |
| "learning_rate": 4.934414911977908e-05, | |
| "loss": 0.4603, | |
| "num_tokens": 376600048.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3400093153237075, | |
| "grad_norm": 0.30758421959340787, | |
| "learning_rate": 4.9257852951328965e-05, | |
| "loss": 0.458, | |
| "num_tokens": 381842928.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3446669771774569, | |
| "grad_norm": 0.2856509710996205, | |
| "learning_rate": 4.917155678287884e-05, | |
| "loss": 0.4716, | |
| "num_tokens": 387073976.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3493246390312063, | |
| "grad_norm": 0.3295399028410657, | |
| "learning_rate": 4.908526061442872e-05, | |
| "loss": 0.4634, | |
| "num_tokens": 392316856.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.28289110029052567, | |
| "learning_rate": 4.89989644459786e-05, | |
| "loss": 0.45, | |
| "num_tokens": 397559736.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3586399627387052, | |
| "grad_norm": 0.34988241905890866, | |
| "learning_rate": 4.891266827752848e-05, | |
| "loss": 0.4443, | |
| "num_tokens": 402802616.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.36329762459245457, | |
| "grad_norm": 0.3296915763265243, | |
| "learning_rate": 4.882637210907836e-05, | |
| "loss": 0.4476, | |
| "num_tokens": 408014414.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.367955286446204, | |
| "grad_norm": 0.3349688176292715, | |
| "learning_rate": 4.874007594062824e-05, | |
| "loss": 0.4435, | |
| "num_tokens": 413257294.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.37261294829995345, | |
| "grad_norm": 0.2628711146772846, | |
| "learning_rate": 4.865377977217811e-05, | |
| "loss": 0.4421, | |
| "num_tokens": 418500174.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.37727061015370283, | |
| "grad_norm": 0.33777569370054417, | |
| "learning_rate": 4.8567483603728e-05, | |
| "loss": 0.442, | |
| "num_tokens": 423692782.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3819282720074523, | |
| "grad_norm": 0.29683573320560597, | |
| "learning_rate": 4.8481187435277875e-05, | |
| "loss": 0.4492, | |
| "num_tokens": 428845062.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.38658593386120166, | |
| "grad_norm": 0.35272599374617136, | |
| "learning_rate": 4.839489126682776e-05, | |
| "loss": 0.4477, | |
| "num_tokens": 434087942.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3912435957149511, | |
| "grad_norm": 0.34467996977933857, | |
| "learning_rate": 4.830859509837763e-05, | |
| "loss": 0.4384, | |
| "num_tokens": 439278078.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.39590125756870054, | |
| "grad_norm": 0.2819923194176978, | |
| "learning_rate": 4.822229892992751e-05, | |
| "loss": 0.4558, | |
| "num_tokens": 444520958.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4005589194224499, | |
| "grad_norm": 0.3799050976439397, | |
| "learning_rate": 4.8136002761477395e-05, | |
| "loss": 0.4478, | |
| "num_tokens": 449763838.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.40521658127619936, | |
| "grad_norm": 0.3110751722247092, | |
| "learning_rate": 4.804970659302727e-05, | |
| "loss": 0.4518, | |
| "num_tokens": 454963270.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.40987424312994875, | |
| "grad_norm": 0.35941404373912506, | |
| "learning_rate": 4.796341042457715e-05, | |
| "loss": 0.4612, | |
| "num_tokens": 460206150.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4145319049836982, | |
| "grad_norm": 0.32752899348089387, | |
| "learning_rate": 4.787711425612703e-05, | |
| "loss": 0.4367, | |
| "num_tokens": 465378662.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4191895668374476, | |
| "grad_norm": 0.3139139204020394, | |
| "learning_rate": 4.779081808767691e-05, | |
| "loss": 0.4401, | |
| "num_tokens": 470602566.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.423847228691197, | |
| "grad_norm": 0.2825862135492498, | |
| "learning_rate": 4.770452191922679e-05, | |
| "loss": 0.462, | |
| "num_tokens": 475800020.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.42850489054494645, | |
| "grad_norm": 0.2723199749905416, | |
| "learning_rate": 4.761822575077667e-05, | |
| "loss": 0.4396, | |
| "num_tokens": 481042900.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.43316255239869583, | |
| "grad_norm": 0.3182623966019266, | |
| "learning_rate": 4.753192958232654e-05, | |
| "loss": 0.45, | |
| "num_tokens": 486220500.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.43782021425244527, | |
| "grad_norm": 0.3045532555366469, | |
| "learning_rate": 4.744563341387643e-05, | |
| "loss": 0.4411, | |
| "num_tokens": 491463380.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 0.2850890965575483, | |
| "learning_rate": 4.7359337245426306e-05, | |
| "loss": 0.4497, | |
| "num_tokens": 496706260.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4471355379599441, | |
| "grad_norm": 0.2320309714284213, | |
| "learning_rate": 4.7273041076976184e-05, | |
| "loss": 0.434, | |
| "num_tokens": 501949140.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.45179319981369354, | |
| "grad_norm": 0.2879495707302432, | |
| "learning_rate": 4.718674490852606e-05, | |
| "loss": 0.4373, | |
| "num_tokens": 507192020.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4564508616674429, | |
| "grad_norm": 0.2741330386708762, | |
| "learning_rate": 4.710044874007594e-05, | |
| "loss": 0.4414, | |
| "num_tokens": 512433054.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.46110852352119236, | |
| "grad_norm": 0.262251172224242, | |
| "learning_rate": 4.7014152571625826e-05, | |
| "loss": 0.4325, | |
| "num_tokens": 517658160.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4657661853749418, | |
| "grad_norm": 0.26053513015916885, | |
| "learning_rate": 4.6927856403175704e-05, | |
| "loss": 0.4403, | |
| "num_tokens": 522901040.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4704238472286912, | |
| "grad_norm": 0.24108379170888813, | |
| "learning_rate": 4.684156023472558e-05, | |
| "loss": 0.4335, | |
| "num_tokens": 528127586.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.4750815090824406, | |
| "grad_norm": 0.29587698347052144, | |
| "learning_rate": 4.675526406627546e-05, | |
| "loss": 0.4297, | |
| "num_tokens": 533355284.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.47973917093619, | |
| "grad_norm": 0.33652344097172093, | |
| "learning_rate": 4.666896789782534e-05, | |
| "loss": 0.4425, | |
| "num_tokens": 538598164.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.48439683278993945, | |
| "grad_norm": 0.27976999375954476, | |
| "learning_rate": 4.658267172937522e-05, | |
| "loss": 0.4388, | |
| "num_tokens": 543841044.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4890544946436889, | |
| "grad_norm": 0.2837801350103133, | |
| "learning_rate": 4.64963755609251e-05, | |
| "loss": 0.4465, | |
| "num_tokens": 549083924.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.49371215649743827, | |
| "grad_norm": 0.3099491399331756, | |
| "learning_rate": 4.641007939247497e-05, | |
| "loss": 0.4322, | |
| "num_tokens": 554323932.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4983698183511877, | |
| "grad_norm": 0.4452299129981611, | |
| "learning_rate": 4.632378322402486e-05, | |
| "loss": 0.4412, | |
| "num_tokens": 559555680.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5030274802049371, | |
| "grad_norm": 0.26935333830330205, | |
| "learning_rate": 4.6237487055574736e-05, | |
| "loss": 0.4439, | |
| "num_tokens": 564798560.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5076851420586865, | |
| "grad_norm": 0.34739553354007896, | |
| "learning_rate": 4.6151190887124615e-05, | |
| "loss": 0.4317, | |
| "num_tokens": 569918486.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5123428039124359, | |
| "grad_norm": 0.3472592651761058, | |
| "learning_rate": 4.606489471867449e-05, | |
| "loss": 0.4475, | |
| "num_tokens": 575160720.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5170004657661854, | |
| "grad_norm": 0.3035605824849283, | |
| "learning_rate": 4.597859855022437e-05, | |
| "loss": 0.433, | |
| "num_tokens": 580403600.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5216581276199348, | |
| "grad_norm": 0.2574408507810935, | |
| "learning_rate": 4.589230238177425e-05, | |
| "loss": 0.4379, | |
| "num_tokens": 585635440.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.2639972065360183, | |
| "learning_rate": 4.5806006213324134e-05, | |
| "loss": 0.4359, | |
| "num_tokens": 590878320.0, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 0.267849189370849, | |
| "learning_rate": 4.5719710044874006e-05, | |
| "loss": 0.4351, | |
| "num_tokens": 596063606.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5356311131811831, | |
| "grad_norm": 0.28287517961378306, | |
| "learning_rate": 4.563341387642389e-05, | |
| "loss": 0.4335, | |
| "num_tokens": 601306486.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5402887750349324, | |
| "grad_norm": 0.2649886191030391, | |
| "learning_rate": 4.554711770797377e-05, | |
| "loss": 0.4362, | |
| "num_tokens": 606451006.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5449464368886818, | |
| "grad_norm": 0.2631971923656154, | |
| "learning_rate": 4.546082153952365e-05, | |
| "loss": 0.4296, | |
| "num_tokens": 611693886.0, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5496040987424313, | |
| "grad_norm": 0.29977311865819234, | |
| "learning_rate": 4.5374525371073526e-05, | |
| "loss": 0.4462, | |
| "num_tokens": 616936766.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5542617605961807, | |
| "grad_norm": 0.28494644236198535, | |
| "learning_rate": 4.5288229202623404e-05, | |
| "loss": 0.4376, | |
| "num_tokens": 622179646.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5589194224499301, | |
| "grad_norm": 0.2641694997511625, | |
| "learning_rate": 4.520193303417328e-05, | |
| "loss": 0.441, | |
| "num_tokens": 627422526.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5635770843036796, | |
| "grad_norm": 0.27897918189506365, | |
| "learning_rate": 4.511563686572317e-05, | |
| "loss": 0.4158, | |
| "num_tokens": 632619928.0, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.568234746157429, | |
| "grad_norm": 0.25761220975231697, | |
| "learning_rate": 4.5029340697273045e-05, | |
| "loss": 0.4345, | |
| "num_tokens": 637862808.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5728924080111784, | |
| "grad_norm": 0.25001128422952085, | |
| "learning_rate": 4.4943044528822923e-05, | |
| "loss": 0.4273, | |
| "num_tokens": 643105688.0, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5775500698649279, | |
| "grad_norm": 0.2914170543484923, | |
| "learning_rate": 4.48567483603728e-05, | |
| "loss": 0.4315, | |
| "num_tokens": 648348568.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5822077317186772, | |
| "grad_norm": 0.31058817794036714, | |
| "learning_rate": 4.477045219192268e-05, | |
| "loss": 0.4411, | |
| "num_tokens": 653591448.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5868653935724266, | |
| "grad_norm": 0.28937125577842665, | |
| "learning_rate": 4.4684156023472565e-05, | |
| "loss": 0.4264, | |
| "num_tokens": 658834328.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5915230554261761, | |
| "grad_norm": 0.22786079347617752, | |
| "learning_rate": 4.4597859855022436e-05, | |
| "loss": 0.4262, | |
| "num_tokens": 664028780.0, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5961807172799255, | |
| "grad_norm": 0.24552824986807895, | |
| "learning_rate": 4.4511563686572315e-05, | |
| "loss": 0.4324, | |
| "num_tokens": 669256250.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6008383791336749, | |
| "grad_norm": 0.2890518447174677, | |
| "learning_rate": 4.44252675181222e-05, | |
| "loss": 0.4303, | |
| "num_tokens": 674460550.0, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6054960409874243, | |
| "grad_norm": 0.30310921639067445, | |
| "learning_rate": 4.433897134967208e-05, | |
| "loss": 0.4278, | |
| "num_tokens": 679701528.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6101537028411738, | |
| "grad_norm": 0.27140239795246707, | |
| "learning_rate": 4.4252675181221956e-05, | |
| "loss": 0.4358, | |
| "num_tokens": 684944408.0, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6148113646949231, | |
| "grad_norm": 0.3083441049780654, | |
| "learning_rate": 4.4166379012771834e-05, | |
| "loss": 0.4369, | |
| "num_tokens": 690187288.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6194690265486725, | |
| "grad_norm": 0.23955876261742862, | |
| "learning_rate": 4.408008284432171e-05, | |
| "loss": 0.4298, | |
| "num_tokens": 695372640.0, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.624126688402422, | |
| "grad_norm": 0.23442625308166126, | |
| "learning_rate": 4.39937866758716e-05, | |
| "loss": 0.4329, | |
| "num_tokens": 700615520.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6287843502561714, | |
| "grad_norm": 0.2871627229865842, | |
| "learning_rate": 4.3907490507421476e-05, | |
| "loss": 0.4256, | |
| "num_tokens": 705831004.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6334420121099208, | |
| "grad_norm": 0.2687212114297664, | |
| "learning_rate": 4.382119433897135e-05, | |
| "loss": 0.4292, | |
| "num_tokens": 711015208.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6380996739636703, | |
| "grad_norm": 0.33541709740891185, | |
| "learning_rate": 4.373489817052123e-05, | |
| "loss": 0.4393, | |
| "num_tokens": 716205240.0, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6427573358174197, | |
| "grad_norm": 0.31495578260456913, | |
| "learning_rate": 4.364860200207111e-05, | |
| "loss": 0.4285, | |
| "num_tokens": 721448120.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6474149976711691, | |
| "grad_norm": 0.2898791627197594, | |
| "learning_rate": 4.356230583362099e-05, | |
| "loss": 0.4285, | |
| "num_tokens": 726691000.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6520726595249184, | |
| "grad_norm": 0.3390871520671219, | |
| "learning_rate": 4.347600966517087e-05, | |
| "loss": 0.4248, | |
| "num_tokens": 731930908.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6567303213786679, | |
| "grad_norm": 0.26322702461053304, | |
| "learning_rate": 4.3389713496720745e-05, | |
| "loss": 0.422, | |
| "num_tokens": 737133156.0, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6613879832324173, | |
| "grad_norm": 0.2411711720508501, | |
| "learning_rate": 4.330341732827063e-05, | |
| "loss": 0.4358, | |
| "num_tokens": 742336330.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6660456450861667, | |
| "grad_norm": 0.2666350329538405, | |
| "learning_rate": 4.321712115982051e-05, | |
| "loss": 0.4288, | |
| "num_tokens": 747579210.0, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6707033069399162, | |
| "grad_norm": 0.2834934990333827, | |
| "learning_rate": 4.3130824991370387e-05, | |
| "loss": 0.4202, | |
| "num_tokens": 752752000.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6753609687936656, | |
| "grad_norm": 0.23475621737499716, | |
| "learning_rate": 4.3044528822920265e-05, | |
| "loss": 0.4242, | |
| "num_tokens": 757994880.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.680018630647415, | |
| "grad_norm": 0.28206945452066673, | |
| "learning_rate": 4.295823265447014e-05, | |
| "loss": 0.4324, | |
| "num_tokens": 763237760.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6846762925011645, | |
| "grad_norm": 0.261061824062115, | |
| "learning_rate": 4.287193648602002e-05, | |
| "loss": 0.4198, | |
| "num_tokens": 768480640.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.6893339543549138, | |
| "grad_norm": 0.24020616053274785, | |
| "learning_rate": 4.27856403175699e-05, | |
| "loss": 0.4237, | |
| "num_tokens": 773723520.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6939916162086632, | |
| "grad_norm": 0.28219654388949683, | |
| "learning_rate": 4.269934414911978e-05, | |
| "loss": 0.4223, | |
| "num_tokens": 778966400.0, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.6986492780624126, | |
| "grad_norm": 0.27272817696730994, | |
| "learning_rate": 4.261304798066966e-05, | |
| "loss": 0.4303, | |
| "num_tokens": 784209280.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7033069399161621, | |
| "grad_norm": 0.2932280320598997, | |
| "learning_rate": 4.252675181221954e-05, | |
| "loss": 0.4179, | |
| "num_tokens": 789452160.0, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.25543865186709225, | |
| "learning_rate": 4.244045564376942e-05, | |
| "loss": 0.4226, | |
| "num_tokens": 794695040.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7126222636236609, | |
| "grad_norm": 0.2432347853756722, | |
| "learning_rate": 4.23541594753193e-05, | |
| "loss": 0.4295, | |
| "num_tokens": 799937676.0, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7172799254774104, | |
| "grad_norm": 0.2685563256620536, | |
| "learning_rate": 4.2267863306869176e-05, | |
| "loss": 0.4291, | |
| "num_tokens": 805180556.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7219375873311598, | |
| "grad_norm": 0.3401605685530348, | |
| "learning_rate": 4.2181567138419054e-05, | |
| "loss": 0.4177, | |
| "num_tokens": 810397834.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7265952491849091, | |
| "grad_norm": 0.2814222563747529, | |
| "learning_rate": 4.209527096996894e-05, | |
| "loss": 0.4347, | |
| "num_tokens": 815565546.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7312529110386586, | |
| "grad_norm": 0.2969737663505764, | |
| "learning_rate": 4.200897480151881e-05, | |
| "loss": 0.4265, | |
| "num_tokens": 820802646.0, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.735910572892408, | |
| "grad_norm": 0.22711045496661186, | |
| "learning_rate": 4.1922678633068695e-05, | |
| "loss": 0.4176, | |
| "num_tokens": 826045526.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7405682347461574, | |
| "grad_norm": 0.24316002701758074, | |
| "learning_rate": 4.1836382464618573e-05, | |
| "loss": 0.4276, | |
| "num_tokens": 831245076.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7452258965999069, | |
| "grad_norm": 0.23443573121049466, | |
| "learning_rate": 4.175008629616845e-05, | |
| "loss": 0.4212, | |
| "num_tokens": 836487956.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7498835584536563, | |
| "grad_norm": 0.2177184558266077, | |
| "learning_rate": 4.166379012771833e-05, | |
| "loss": 0.4386, | |
| "num_tokens": 841670226.0, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.7545412203074057, | |
| "grad_norm": 0.2822890601620429, | |
| "learning_rate": 4.157749395926821e-05, | |
| "loss": 0.427, | |
| "num_tokens": 846891280.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.759198882161155, | |
| "grad_norm": 0.2680333802224899, | |
| "learning_rate": 4.1491197790818086e-05, | |
| "loss": 0.4262, | |
| "num_tokens": 852076096.0, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7638565440149045, | |
| "grad_norm": 0.2563374963785661, | |
| "learning_rate": 4.140490162236797e-05, | |
| "loss": 0.4252, | |
| "num_tokens": 857318976.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7685142058686539, | |
| "grad_norm": 0.2741256247096192, | |
| "learning_rate": 4.131860545391785e-05, | |
| "loss": 0.4141, | |
| "num_tokens": 862561856.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7731718677224033, | |
| "grad_norm": 0.2217312700121506, | |
| "learning_rate": 4.123230928546773e-05, | |
| "loss": 0.4238, | |
| "num_tokens": 867804736.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7778295295761528, | |
| "grad_norm": 0.24929317063270387, | |
| "learning_rate": 4.1146013117017606e-05, | |
| "loss": 0.4288, | |
| "num_tokens": 873045328.0, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7824871914299022, | |
| "grad_norm": 0.239155265203606, | |
| "learning_rate": 4.1059716948567484e-05, | |
| "loss": 0.4137, | |
| "num_tokens": 878249566.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7871448532836516, | |
| "grad_norm": 0.23463365870880432, | |
| "learning_rate": 4.097342078011737e-05, | |
| "loss": 0.4214, | |
| "num_tokens": 883492446.0, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7918025151374011, | |
| "grad_norm": 0.24834022806840694, | |
| "learning_rate": 4.088712461166724e-05, | |
| "loss": 0.4217, | |
| "num_tokens": 888735326.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7964601769911505, | |
| "grad_norm": 0.263701388819094, | |
| "learning_rate": 4.080082844321712e-05, | |
| "loss": 0.4133, | |
| "num_tokens": 893978206.0, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8011178388448998, | |
| "grad_norm": 0.24483255965235723, | |
| "learning_rate": 4.0714532274767004e-05, | |
| "loss": 0.416, | |
| "num_tokens": 899180370.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8057755006986492, | |
| "grad_norm": 0.26789454184774325, | |
| "learning_rate": 4.062823610631688e-05, | |
| "loss": 0.4194, | |
| "num_tokens": 904405480.0, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.8104331625523987, | |
| "grad_norm": 0.2219920377956709, | |
| "learning_rate": 4.054193993786676e-05, | |
| "loss": 0.4123, | |
| "num_tokens": 909648360.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8150908244061481, | |
| "grad_norm": 0.2478368054109279, | |
| "learning_rate": 4.045564376941664e-05, | |
| "loss": 0.4231, | |
| "num_tokens": 914891240.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.8197484862598975, | |
| "grad_norm": 0.24420526285043473, | |
| "learning_rate": 4.036934760096652e-05, | |
| "loss": 0.4072, | |
| "num_tokens": 920134120.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.824406148113647, | |
| "grad_norm": 0.23673717599556984, | |
| "learning_rate": 4.02830514325164e-05, | |
| "loss": 0.4234, | |
| "num_tokens": 925348344.0, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8290638099673964, | |
| "grad_norm": 0.20687140115596026, | |
| "learning_rate": 4.019675526406628e-05, | |
| "loss": 0.4221, | |
| "num_tokens": 930591224.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8337214718211458, | |
| "grad_norm": 0.23577724933371347, | |
| "learning_rate": 4.011045909561615e-05, | |
| "loss": 0.4208, | |
| "num_tokens": 935834104.0, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8383791336748952, | |
| "grad_norm": 0.22950933284274888, | |
| "learning_rate": 4.0024162927166037e-05, | |
| "loss": 0.4167, | |
| "num_tokens": 941076984.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8430367955286446, | |
| "grad_norm": 0.23397135664041768, | |
| "learning_rate": 3.9937866758715915e-05, | |
| "loss": 0.4203, | |
| "num_tokens": 946319864.0, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.847694457382394, | |
| "grad_norm": 0.25132849541590513, | |
| "learning_rate": 3.98515705902658e-05, | |
| "loss": 0.4267, | |
| "num_tokens": 951515422.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8523521192361434, | |
| "grad_norm": 0.23255546519793427, | |
| "learning_rate": 3.976527442181567e-05, | |
| "loss": 0.4177, | |
| "num_tokens": 956704248.0, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.8570097810898929, | |
| "grad_norm": 0.2572102709817985, | |
| "learning_rate": 3.967897825336555e-05, | |
| "loss": 0.4247, | |
| "num_tokens": 961947128.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8616674429436423, | |
| "grad_norm": 0.2794655020066489, | |
| "learning_rate": 3.9592682084915434e-05, | |
| "loss": 0.4187, | |
| "num_tokens": 967190008.0, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8663251047973917, | |
| "grad_norm": 0.2616474979319117, | |
| "learning_rate": 3.950638591646531e-05, | |
| "loss": 0.4218, | |
| "num_tokens": 972432888.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8709827666511412, | |
| "grad_norm": 0.25818286592762985, | |
| "learning_rate": 3.942008974801519e-05, | |
| "loss": 0.4079, | |
| "num_tokens": 977675768.0, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.8756404285048905, | |
| "grad_norm": 0.2620708843817634, | |
| "learning_rate": 3.933379357956507e-05, | |
| "loss": 0.4098, | |
| "num_tokens": 982908826.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8802980903586399, | |
| "grad_norm": 0.2347432430996397, | |
| "learning_rate": 3.924749741111495e-05, | |
| "loss": 0.4122, | |
| "num_tokens": 988129568.0, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 0.24950154908542746, | |
| "learning_rate": 3.916120124266483e-05, | |
| "loss": 0.4021, | |
| "num_tokens": 993372448.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8896134140661388, | |
| "grad_norm": 0.23273641557799368, | |
| "learning_rate": 3.9074905074214704e-05, | |
| "loss": 0.4165, | |
| "num_tokens": 998615328.0, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8942710759198882, | |
| "grad_norm": 0.8612696564377977, | |
| "learning_rate": 3.898860890576458e-05, | |
| "loss": 0.4232, | |
| "num_tokens": 1003840426.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8989287377736377, | |
| "grad_norm": 0.285677341095466, | |
| "learning_rate": 3.890231273731447e-05, | |
| "loss": 0.4169, | |
| "num_tokens": 1009083306.0, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.9035863996273871, | |
| "grad_norm": 0.24828614450017533, | |
| "learning_rate": 3.8816016568864345e-05, | |
| "loss": 0.4197, | |
| "num_tokens": 1014326186.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9082440614811365, | |
| "grad_norm": 0.2510856869417985, | |
| "learning_rate": 3.8729720400414224e-05, | |
| "loss": 0.4141, | |
| "num_tokens": 1019569066.0, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.9129017233348858, | |
| "grad_norm": 0.2675729105260259, | |
| "learning_rate": 3.86434242319641e-05, | |
| "loss": 0.4171, | |
| "num_tokens": 1024811946.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9175593851886353, | |
| "grad_norm": 0.24069901373503072, | |
| "learning_rate": 3.855712806351398e-05, | |
| "loss": 0.4214, | |
| "num_tokens": 1030054826.0, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.9222170470423847, | |
| "grad_norm": 0.22691653413707366, | |
| "learning_rate": 3.8470831895063865e-05, | |
| "loss": 0.3934, | |
| "num_tokens": 1035279144.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9268747088961341, | |
| "grad_norm": 0.2446560685959566, | |
| "learning_rate": 3.838453572661374e-05, | |
| "loss": 0.41, | |
| "num_tokens": 1040522024.0, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.9315323707498836, | |
| "grad_norm": 0.23589892031832072, | |
| "learning_rate": 3.8298239558163615e-05, | |
| "loss": 0.4123, | |
| "num_tokens": 1045764904.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.936190032603633, | |
| "grad_norm": 0.25777225090756506, | |
| "learning_rate": 3.82119433897135e-05, | |
| "loss": 0.4216, | |
| "num_tokens": 1051007784.0, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.9408476944573824, | |
| "grad_norm": 0.21844053190545842, | |
| "learning_rate": 3.812564722126338e-05, | |
| "loss": 0.4123, | |
| "num_tokens": 1056250664.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9455053563111319, | |
| "grad_norm": 0.21445376667996244, | |
| "learning_rate": 3.8039351052813256e-05, | |
| "loss": 0.4134, | |
| "num_tokens": 1061493544.0, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.9501630181648812, | |
| "grad_norm": 0.24320130003758428, | |
| "learning_rate": 3.7953054884363134e-05, | |
| "loss": 0.4221, | |
| "num_tokens": 1066736424.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9548206800186306, | |
| "grad_norm": 0.22671648010549939, | |
| "learning_rate": 3.786675871591301e-05, | |
| "loss": 0.4142, | |
| "num_tokens": 1071979304.0, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.95947834187238, | |
| "grad_norm": 0.22117437951254582, | |
| "learning_rate": 3.77804625474629e-05, | |
| "loss": 0.4061, | |
| "num_tokens": 1077222184.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9641360037261295, | |
| "grad_norm": 0.23704403632072713, | |
| "learning_rate": 3.7694166379012776e-05, | |
| "loss": 0.4091, | |
| "num_tokens": 1082465064.0, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.9687936655798789, | |
| "grad_norm": 0.2181135424522087, | |
| "learning_rate": 3.7607870210562654e-05, | |
| "loss": 0.4079, | |
| "num_tokens": 1087707944.0, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9734513274336283, | |
| "grad_norm": 0.2201471206294805, | |
| "learning_rate": 3.752157404211253e-05, | |
| "loss": 0.409, | |
| "num_tokens": 1092923046.0, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.9781089892873778, | |
| "grad_norm": 0.2481687896895825, | |
| "learning_rate": 3.743527787366241e-05, | |
| "loss": 0.409, | |
| "num_tokens": 1098165926.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9827666511411272, | |
| "grad_norm": 0.22426850689631653, | |
| "learning_rate": 3.734898170521229e-05, | |
| "loss": 0.4132, | |
| "num_tokens": 1103324696.0, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.9874243129948765, | |
| "grad_norm": 0.23293678058861622, | |
| "learning_rate": 3.7262685536762174e-05, | |
| "loss": 0.4124, | |
| "num_tokens": 1108567576.0, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.992081974848626, | |
| "grad_norm": 0.22971773619907454, | |
| "learning_rate": 3.7176389368312045e-05, | |
| "loss": 0.4113, | |
| "num_tokens": 1113810456.0, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9967396367023754, | |
| "grad_norm": 0.2623764973513331, | |
| "learning_rate": 3.709009319986193e-05, | |
| "loss": 0.4019, | |
| "num_tokens": 1119053336.0, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.00093153237075, | |
| "grad_norm": 0.3672694994896235, | |
| "learning_rate": 3.700379703141181e-05, | |
| "loss": 0.4142, | |
| "num_tokens": 1123444248.0, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.0055891942244992, | |
| "grad_norm": 0.2824258365122951, | |
| "learning_rate": 3.6917500862961687e-05, | |
| "loss": 0.3608, | |
| "num_tokens": 1128625632.0, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.0102468560782487, | |
| "grad_norm": 0.2220444817259926, | |
| "learning_rate": 3.6831204694511565e-05, | |
| "loss": 0.3465, | |
| "num_tokens": 1133868512.0, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.0149045179319982, | |
| "grad_norm": 0.2287154958640183, | |
| "learning_rate": 3.674490852606144e-05, | |
| "loss": 0.36, | |
| "num_tokens": 1139111392.0, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.0195621797857475, | |
| "grad_norm": 0.24698825266576346, | |
| "learning_rate": 3.665861235761132e-05, | |
| "loss": 0.3556, | |
| "num_tokens": 1144313368.0, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.024219841639497, | |
| "grad_norm": 0.2604025594854447, | |
| "learning_rate": 3.6572316189161206e-05, | |
| "loss": 0.3597, | |
| "num_tokens": 1149498184.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.0288775034932465, | |
| "grad_norm": 0.22666270690738718, | |
| "learning_rate": 3.6486020020711085e-05, | |
| "loss": 0.3583, | |
| "num_tokens": 1154741064.0, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.0335351653469957, | |
| "grad_norm": 0.21462990749032038, | |
| "learning_rate": 3.639972385226096e-05, | |
| "loss": 0.3584, | |
| "num_tokens": 1159983944.0, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.0381928272007452, | |
| "grad_norm": 0.23911611952529468, | |
| "learning_rate": 3.631342768381084e-05, | |
| "loss": 0.3509, | |
| "num_tokens": 1165183376.0, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.0428504890544947, | |
| "grad_norm": 0.24512784651248454, | |
| "learning_rate": 3.622713151536072e-05, | |
| "loss": 0.3575, | |
| "num_tokens": 1170426256.0, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.047508150908244, | |
| "grad_norm": 0.2333802503793358, | |
| "learning_rate": 3.6140835346910604e-05, | |
| "loss": 0.3616, | |
| "num_tokens": 1175669136.0, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.0521658127619935, | |
| "grad_norm": 0.2511459845467595, | |
| "learning_rate": 3.6054539178460476e-05, | |
| "loss": 0.35, | |
| "num_tokens": 1180830872.0, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.056823474615743, | |
| "grad_norm": 0.24201094782681035, | |
| "learning_rate": 3.5968243010010354e-05, | |
| "loss": 0.3577, | |
| "num_tokens": 1186073752.0, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.0614811364694923, | |
| "grad_norm": 0.23368138863917715, | |
| "learning_rate": 3.588194684156024e-05, | |
| "loss": 0.358, | |
| "num_tokens": 1191316632.0, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.0661387983232418, | |
| "grad_norm": 0.23275627460132062, | |
| "learning_rate": 3.579565067311012e-05, | |
| "loss": 0.3571, | |
| "num_tokens": 1196496612.0, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.0707964601769913, | |
| "grad_norm": 0.20892181367377105, | |
| "learning_rate": 3.5709354504659995e-05, | |
| "loss": 0.3541, | |
| "num_tokens": 1201739492.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.0754541220307405, | |
| "grad_norm": 0.2947512380750944, | |
| "learning_rate": 3.5623058336209874e-05, | |
| "loss": 0.3577, | |
| "num_tokens": 1206939480.0, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.08011178388449, | |
| "grad_norm": 0.24415115511422772, | |
| "learning_rate": 3.553676216775975e-05, | |
| "loss": 0.3658, | |
| "num_tokens": 1212064592.0, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.0847694457382393, | |
| "grad_norm": 0.22693331138134526, | |
| "learning_rate": 3.545046599930964e-05, | |
| "loss": 0.3627, | |
| "num_tokens": 1217281960.0, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.0894271075919888, | |
| "grad_norm": 0.1963016262673903, | |
| "learning_rate": 3.536416983085951e-05, | |
| "loss": 0.3606, | |
| "num_tokens": 1222524840.0, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.0940847694457383, | |
| "grad_norm": 0.22725888539783665, | |
| "learning_rate": 3.5277873662409386e-05, | |
| "loss": 0.3604, | |
| "num_tokens": 1227767720.0, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.0987424312994876, | |
| "grad_norm": 0.24639322397682187, | |
| "learning_rate": 3.519157749395927e-05, | |
| "loss": 0.3531, | |
| "num_tokens": 1232972020.0, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.103400093153237, | |
| "grad_norm": 0.2165339172344169, | |
| "learning_rate": 3.510528132550915e-05, | |
| "loss": 0.3542, | |
| "num_tokens": 1238214900.0, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.1080577550069866, | |
| "grad_norm": 0.23564183999743302, | |
| "learning_rate": 3.501898515705903e-05, | |
| "loss": 0.3362, | |
| "num_tokens": 1243457780.0, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.1127154168607358, | |
| "grad_norm": 0.2291128967075881, | |
| "learning_rate": 3.4932688988608906e-05, | |
| "loss": 0.3467, | |
| "num_tokens": 1248618688.0, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.1173730787144853, | |
| "grad_norm": 0.22383859224734165, | |
| "learning_rate": 3.4846392820158784e-05, | |
| "loss": 0.3555, | |
| "num_tokens": 1253838334.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.1220307405682348, | |
| "grad_norm": 0.2264578088311139, | |
| "learning_rate": 3.476009665170867e-05, | |
| "loss": 0.364, | |
| "num_tokens": 1259062652.0, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.126688402421984, | |
| "grad_norm": 0.2505640583911167, | |
| "learning_rate": 3.467380048325855e-05, | |
| "loss": 0.3566, | |
| "num_tokens": 1264270772.0, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.1313460642757336, | |
| "grad_norm": 0.22987819447343966, | |
| "learning_rate": 3.458750431480842e-05, | |
| "loss": 0.3522, | |
| "num_tokens": 1269495882.0, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.136003726129483, | |
| "grad_norm": 0.2518194235864917, | |
| "learning_rate": 3.4501208146358304e-05, | |
| "loss": 0.3632, | |
| "num_tokens": 1274688984.0, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.1406613879832324, | |
| "grad_norm": 0.2548213621153145, | |
| "learning_rate": 3.441491197790818e-05, | |
| "loss": 0.3516, | |
| "num_tokens": 1279931864.0, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.1453190498369819, | |
| "grad_norm": 0.23439086402593545, | |
| "learning_rate": 3.432861580945806e-05, | |
| "loss": 0.3553, | |
| "num_tokens": 1285174744.0, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.1499767116907313, | |
| "grad_norm": 0.24996036439883423, | |
| "learning_rate": 3.424231964100794e-05, | |
| "loss": 0.355, | |
| "num_tokens": 1290357536.0, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.1546343735444806, | |
| "grad_norm": 0.22700170648009696, | |
| "learning_rate": 3.415602347255782e-05, | |
| "loss": 0.3596, | |
| "num_tokens": 1295587548.0, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.1592920353982301, | |
| "grad_norm": 0.20305574931314446, | |
| "learning_rate": 3.40697273041077e-05, | |
| "loss": 0.3624, | |
| "num_tokens": 1300830428.0, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.1639496972519794, | |
| "grad_norm": 0.21770072942642166, | |
| "learning_rate": 3.398343113565758e-05, | |
| "loss": 0.3646, | |
| "num_tokens": 1306073308.0, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.1686073591057289, | |
| "grad_norm": 0.21103382747916685, | |
| "learning_rate": 3.389713496720746e-05, | |
| "loss": 0.355, | |
| "num_tokens": 1311249846.0, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.1732650209594784, | |
| "grad_norm": 0.22753156643631783, | |
| "learning_rate": 3.381083879875734e-05, | |
| "loss": 0.3637, | |
| "num_tokens": 1316492726.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.1779226828132279, | |
| "grad_norm": 0.22364876562779948, | |
| "learning_rate": 3.3724542630307215e-05, | |
| "loss": 0.3515, | |
| "num_tokens": 1321729528.0, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.1825803446669771, | |
| "grad_norm": 0.22017951349349468, | |
| "learning_rate": 3.363824646185709e-05, | |
| "loss": 0.3453, | |
| "num_tokens": 1326922446.0, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.1872380065207266, | |
| "grad_norm": 0.22202056706931508, | |
| "learning_rate": 3.355195029340698e-05, | |
| "loss": 0.3566, | |
| "num_tokens": 1332165326.0, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.191895668374476, | |
| "grad_norm": 0.22338433758494683, | |
| "learning_rate": 3.346565412495685e-05, | |
| "loss": 0.3589, | |
| "num_tokens": 1337408206.0, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.1965533302282254, | |
| "grad_norm": 0.22483337024003108, | |
| "learning_rate": 3.3379357956506735e-05, | |
| "loss": 0.3635, | |
| "num_tokens": 1342651086.0, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.201210992081975, | |
| "grad_norm": 0.24346359617349606, | |
| "learning_rate": 3.329306178805661e-05, | |
| "loss": 0.353, | |
| "num_tokens": 1347876028.0, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.2058686539357242, | |
| "grad_norm": 0.22661755141039527, | |
| "learning_rate": 3.320676561960649e-05, | |
| "loss": 0.3569, | |
| "num_tokens": 1353118908.0, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.20263082211858716, | |
| "learning_rate": 3.312046945115637e-05, | |
| "loss": 0.3526, | |
| "num_tokens": 1358361788.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.2151839776432232, | |
| "grad_norm": 0.2439981650114557, | |
| "learning_rate": 3.303417328270625e-05, | |
| "loss": 0.363, | |
| "num_tokens": 1363556240.0, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 1.2198416394969724, | |
| "grad_norm": 0.24652335258420222, | |
| "learning_rate": 3.2947877114256126e-05, | |
| "loss": 0.3658, | |
| "num_tokens": 1368799120.0, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.224499301350722, | |
| "grad_norm": 0.22679812500420424, | |
| "learning_rate": 3.286158094580601e-05, | |
| "loss": 0.355, | |
| "num_tokens": 1374042000.0, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 1.2291569632044714, | |
| "grad_norm": 0.24955574806002354, | |
| "learning_rate": 3.277528477735589e-05, | |
| "loss": 0.3653, | |
| "num_tokens": 1379284880.0, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.2338146250582207, | |
| "grad_norm": 0.22202888067052107, | |
| "learning_rate": 3.268898860890577e-05, | |
| "loss": 0.3566, | |
| "num_tokens": 1384484350.0, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.2384722869119702, | |
| "grad_norm": 0.23194015579620933, | |
| "learning_rate": 3.2602692440455645e-05, | |
| "loss": 0.3628, | |
| "num_tokens": 1389727230.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.2431299487657197, | |
| "grad_norm": 0.2615386609253307, | |
| "learning_rate": 3.2516396272005524e-05, | |
| "loss": 0.3558, | |
| "num_tokens": 1394919838.0, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 1.247787610619469, | |
| "grad_norm": 0.257100183800459, | |
| "learning_rate": 3.243010010355541e-05, | |
| "loss": 0.3556, | |
| "num_tokens": 1400162718.0, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.2524452724732185, | |
| "grad_norm": 0.21464705390229605, | |
| "learning_rate": 3.234380393510528e-05, | |
| "loss": 0.346, | |
| "num_tokens": 1405378054.0, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 1.257102934326968, | |
| "grad_norm": 0.2180807929044995, | |
| "learning_rate": 3.225750776665516e-05, | |
| "loss": 0.3568, | |
| "num_tokens": 1410620934.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.2617605961807172, | |
| "grad_norm": 0.1934722938763174, | |
| "learning_rate": 3.217121159820504e-05, | |
| "loss": 0.3505, | |
| "num_tokens": 1415863814.0, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 1.2664182580344667, | |
| "grad_norm": 0.2286860861249808, | |
| "learning_rate": 3.208491542975492e-05, | |
| "loss": 0.3459, | |
| "num_tokens": 1421106694.0, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.271075919888216, | |
| "grad_norm": 0.1949118607209114, | |
| "learning_rate": 3.19986192613048e-05, | |
| "loss": 0.3594, | |
| "num_tokens": 1426349574.0, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 1.2757335817419655, | |
| "grad_norm": 0.2518695507342612, | |
| "learning_rate": 3.191232309285468e-05, | |
| "loss": 0.3573, | |
| "num_tokens": 1431592454.0, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.280391243595715, | |
| "grad_norm": 0.18853372243176753, | |
| "learning_rate": 3.1826026924404556e-05, | |
| "loss": 0.3561, | |
| "num_tokens": 1436835334.0, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.2850489054494645, | |
| "grad_norm": 0.2198966088361658, | |
| "learning_rate": 3.173973075595444e-05, | |
| "loss": 0.3553, | |
| "num_tokens": 1442078214.0, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.2897065673032138, | |
| "grad_norm": 0.19670516928430468, | |
| "learning_rate": 3.165343458750431e-05, | |
| "loss": 0.3553, | |
| "num_tokens": 1447321094.0, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 1.2943642291569633, | |
| "grad_norm": 0.22828433520374913, | |
| "learning_rate": 3.156713841905419e-05, | |
| "loss": 0.3627, | |
| "num_tokens": 1452563974.0, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.2990218910107125, | |
| "grad_norm": 0.19683120837834497, | |
| "learning_rate": 3.1480842250604076e-05, | |
| "loss": 0.3555, | |
| "num_tokens": 1457790520.0, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 1.303679552864462, | |
| "grad_norm": 0.25515592338843396, | |
| "learning_rate": 3.1394546082153954e-05, | |
| "loss": 0.3609, | |
| "num_tokens": 1463015630.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.3083372147182115, | |
| "grad_norm": 0.22154667275288062, | |
| "learning_rate": 3.130824991370383e-05, | |
| "loss": 0.351, | |
| "num_tokens": 1468217794.0, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 1.312994876571961, | |
| "grad_norm": 0.23215300521021542, | |
| "learning_rate": 3.122195374525371e-05, | |
| "loss": 0.3501, | |
| "num_tokens": 1473372988.0, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.3176525384257103, | |
| "grad_norm": 0.20074856829426602, | |
| "learning_rate": 3.113565757680359e-05, | |
| "loss": 0.3693, | |
| "num_tokens": 1478610484.0, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 1.3223102002794598, | |
| "grad_norm": 0.23913181777037554, | |
| "learning_rate": 3.1049361408353474e-05, | |
| "loss": 0.36, | |
| "num_tokens": 1483853364.0, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.326967862133209, | |
| "grad_norm": 0.20003731962610308, | |
| "learning_rate": 3.096306523990335e-05, | |
| "loss": 0.351, | |
| "num_tokens": 1489042190.0, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.3316255239869585, | |
| "grad_norm": 0.20644674819443287, | |
| "learning_rate": 3.0876769071453223e-05, | |
| "loss": 0.3542, | |
| "num_tokens": 1494232222.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.336283185840708, | |
| "grad_norm": 0.1974035173138762, | |
| "learning_rate": 3.079047290300311e-05, | |
| "loss": 0.357, | |
| "num_tokens": 1499475102.0, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 1.3409408476944573, | |
| "grad_norm": 0.19268558436708005, | |
| "learning_rate": 3.070417673455299e-05, | |
| "loss": 0.3563, | |
| "num_tokens": 1504716080.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.3455985095482068, | |
| "grad_norm": 0.2067534023532924, | |
| "learning_rate": 3.061788056610287e-05, | |
| "loss": 0.3627, | |
| "num_tokens": 1509958960.0, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 1.350256171401956, | |
| "grad_norm": 0.2056434127332076, | |
| "learning_rate": 3.053158439765274e-05, | |
| "loss": 0.3519, | |
| "num_tokens": 1515201840.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.3549138332557056, | |
| "grad_norm": 0.19894158768579334, | |
| "learning_rate": 3.0445288229202625e-05, | |
| "loss": 0.3608, | |
| "num_tokens": 1520444720.0, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 1.359571495109455, | |
| "grad_norm": 0.22175281587634732, | |
| "learning_rate": 3.0358992060752506e-05, | |
| "loss": 0.3446, | |
| "num_tokens": 1525687356.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.3642291569632046, | |
| "grad_norm": 0.2042329128619712, | |
| "learning_rate": 3.027269589230238e-05, | |
| "loss": 0.3518, | |
| "num_tokens": 1530885262.0, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 1.3688868188169538, | |
| "grad_norm": 0.22057793167514664, | |
| "learning_rate": 3.018639972385226e-05, | |
| "loss": 0.357, | |
| "num_tokens": 1536099842.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.3735444806707033, | |
| "grad_norm": 0.21367814598801835, | |
| "learning_rate": 3.010010355540214e-05, | |
| "loss": 0.3508, | |
| "num_tokens": 1541293860.0, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.3782021425244526, | |
| "grad_norm": 0.21788842442142192, | |
| "learning_rate": 3.001380738695202e-05, | |
| "loss": 0.3537, | |
| "num_tokens": 1546481532.0, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.382859804378202, | |
| "grad_norm": 0.2129006536280718, | |
| "learning_rate": 2.99275112185019e-05, | |
| "loss": 0.3605, | |
| "num_tokens": 1551724412.0, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 1.3875174662319516, | |
| "grad_norm": 0.20311647150112996, | |
| "learning_rate": 2.984121505005178e-05, | |
| "loss": 0.3597, | |
| "num_tokens": 1556967292.0, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.392175128085701, | |
| "grad_norm": 0.1995467221942951, | |
| "learning_rate": 2.9754918881601657e-05, | |
| "loss": 0.3592, | |
| "num_tokens": 1562210172.0, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 1.3968327899394504, | |
| "grad_norm": 0.19328803971523173, | |
| "learning_rate": 2.966862271315154e-05, | |
| "loss": 0.3479, | |
| "num_tokens": 1567453052.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4014904517931999, | |
| "grad_norm": 0.22159643242508917, | |
| "learning_rate": 2.9582326544701417e-05, | |
| "loss": 0.3533, | |
| "num_tokens": 1572651232.0, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 1.4061481136469491, | |
| "grad_norm": 0.201453046647755, | |
| "learning_rate": 2.9496030376251292e-05, | |
| "loss": 0.352, | |
| "num_tokens": 1577894112.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.4108057755006986, | |
| "grad_norm": 0.1923816720246006, | |
| "learning_rate": 2.9409734207801177e-05, | |
| "loss": 0.3621, | |
| "num_tokens": 1583076382.0, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 1.4154634373544481, | |
| "grad_norm": 0.214579585116131, | |
| "learning_rate": 2.9323438039351052e-05, | |
| "loss": 0.3534, | |
| "num_tokens": 1588319262.0, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.4201210992081974, | |
| "grad_norm": 0.22056328737620753, | |
| "learning_rate": 2.9237141870900937e-05, | |
| "loss": 0.3539, | |
| "num_tokens": 1593562142.0, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.424778761061947, | |
| "grad_norm": 0.19736190427070113, | |
| "learning_rate": 2.9150845702450812e-05, | |
| "loss": 0.3448, | |
| "num_tokens": 1598805022.0, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.4294364229156964, | |
| "grad_norm": 0.20349068024202002, | |
| "learning_rate": 2.906454953400069e-05, | |
| "loss": 0.3623, | |
| "num_tokens": 1604047902.0, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 1.4340940847694457, | |
| "grad_norm": 0.17533490600591858, | |
| "learning_rate": 2.897825336555057e-05, | |
| "loss": 0.3525, | |
| "num_tokens": 1609290782.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.4387517466231952, | |
| "grad_norm": 0.17875571100522028, | |
| "learning_rate": 2.889195719710045e-05, | |
| "loss": 0.3531, | |
| "num_tokens": 1614533662.0, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.4434094084769447, | |
| "grad_norm": 0.18501668823747172, | |
| "learning_rate": 2.8805661028650328e-05, | |
| "loss": 0.3477, | |
| "num_tokens": 1619776542.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.448067070330694, | |
| "grad_norm": 0.19692511936469895, | |
| "learning_rate": 2.871936486020021e-05, | |
| "loss": 0.3444, | |
| "num_tokens": 1625019422.0, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 1.4527247321844434, | |
| "grad_norm": 0.22264199491481562, | |
| "learning_rate": 2.8633068691750088e-05, | |
| "loss": 0.3525, | |
| "num_tokens": 1630262302.0, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.4573823940381927, | |
| "grad_norm": 0.20218356001094864, | |
| "learning_rate": 2.854677252329997e-05, | |
| "loss": 0.3608, | |
| "num_tokens": 1635461518.0, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 1.4620400558919422, | |
| "grad_norm": 0.23582448523809849, | |
| "learning_rate": 2.8460476354849848e-05, | |
| "loss": 0.3722, | |
| "num_tokens": 1640704398.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.4666977177456917, | |
| "grad_norm": 0.24526018198436725, | |
| "learning_rate": 2.8374180186399723e-05, | |
| "loss": 0.354, | |
| "num_tokens": 1645947278.0, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.4713553795994412, | |
| "grad_norm": 0.23262166899112952, | |
| "learning_rate": 2.8287884017949608e-05, | |
| "loss": 0.3458, | |
| "num_tokens": 1651190158.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.4760130414531905, | |
| "grad_norm": 0.21147202390792988, | |
| "learning_rate": 2.8201587849499482e-05, | |
| "loss": 0.3541, | |
| "num_tokens": 1656433038.0, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 1.48067070330694, | |
| "grad_norm": 0.21140335302132418, | |
| "learning_rate": 2.811529168104936e-05, | |
| "loss": 0.3528, | |
| "num_tokens": 1661675918.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.4853283651606892, | |
| "grad_norm": 0.21138321951558706, | |
| "learning_rate": 2.8028995512599242e-05, | |
| "loss": 0.3516, | |
| "num_tokens": 1666918798.0, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.4899860270144387, | |
| "grad_norm": 0.2217336802074178, | |
| "learning_rate": 2.794269934414912e-05, | |
| "loss": 0.3576, | |
| "num_tokens": 1672161678.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.4946436888681882, | |
| "grad_norm": 0.1947287994954116, | |
| "learning_rate": 2.7856403175699002e-05, | |
| "loss": 0.349, | |
| "num_tokens": 1677379408.0, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.4993013507219377, | |
| "grad_norm": 0.210795644222948, | |
| "learning_rate": 2.777010700724888e-05, | |
| "loss": 0.3551, | |
| "num_tokens": 1682574966.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.503959012575687, | |
| "grad_norm": 0.21678781186790855, | |
| "learning_rate": 2.768381083879876e-05, | |
| "loss": 0.3533, | |
| "num_tokens": 1687810252.0, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 1.5086166744294365, | |
| "grad_norm": 0.1958072936276665, | |
| "learning_rate": 2.759751467034864e-05, | |
| "loss": 0.3548, | |
| "num_tokens": 1693051286.0, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.5132743362831858, | |
| "grad_norm": 0.195888740286851, | |
| "learning_rate": 2.751121850189852e-05, | |
| "loss": 0.3578, | |
| "num_tokens": 1698294166.0, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.5179319981369352, | |
| "grad_norm": 0.2218425516701336, | |
| "learning_rate": 2.7424922333448393e-05, | |
| "loss": 0.355, | |
| "num_tokens": 1703537046.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.5225896599906847, | |
| "grad_norm": 0.20731923133305644, | |
| "learning_rate": 2.7338626164998278e-05, | |
| "loss": 0.3653, | |
| "num_tokens": 1708779926.0, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.5272473218444342, | |
| "grad_norm": 0.19045004460891435, | |
| "learning_rate": 2.7252329996548153e-05, | |
| "loss": 0.3584, | |
| "num_tokens": 1714022806.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.5319049836981835, | |
| "grad_norm": 0.23615985512522628, | |
| "learning_rate": 2.7166033828098038e-05, | |
| "loss": 0.3614, | |
| "num_tokens": 1719265686.0, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 1.5365626455519328, | |
| "grad_norm": 0.19367844618271696, | |
| "learning_rate": 2.7079737659647913e-05, | |
| "loss": 0.3469, | |
| "num_tokens": 1724508566.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.5412203074056823, | |
| "grad_norm": 0.22293150745044873, | |
| "learning_rate": 2.699344149119779e-05, | |
| "loss": 0.3457, | |
| "num_tokens": 1729751446.0, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 1.5458779692594318, | |
| "grad_norm": 0.21756885950072824, | |
| "learning_rate": 2.6907145322747673e-05, | |
| "loss": 0.3457, | |
| "num_tokens": 1734965670.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.5505356311131813, | |
| "grad_norm": 0.19339837979852165, | |
| "learning_rate": 2.682084915429755e-05, | |
| "loss": 0.3586, | |
| "num_tokens": 1740208550.0, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.5551932929669308, | |
| "grad_norm": 0.19943103944516885, | |
| "learning_rate": 2.673455298584743e-05, | |
| "loss": 0.355, | |
| "num_tokens": 1745389856.0, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.55985095482068, | |
| "grad_norm": 0.2250859278951836, | |
| "learning_rate": 2.664825681739731e-05, | |
| "loss": 0.3514, | |
| "num_tokens": 1750632736.0, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.5645086166744293, | |
| "grad_norm": 0.21015513353230048, | |
| "learning_rate": 2.6561960648947186e-05, | |
| "loss": 0.3585, | |
| "num_tokens": 1755875616.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.5691662785281788, | |
| "grad_norm": 0.22625589287371234, | |
| "learning_rate": 2.647566448049707e-05, | |
| "loss": 0.3482, | |
| "num_tokens": 1761107456.0, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 1.5738239403819283, | |
| "grad_norm": 0.189125334788574, | |
| "learning_rate": 2.6389368312046945e-05, | |
| "loss": 0.3559, | |
| "num_tokens": 1766350336.0, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.5784816022356778, | |
| "grad_norm": 0.18538752004557982, | |
| "learning_rate": 2.6303072143596824e-05, | |
| "loss": 0.352, | |
| "num_tokens": 1771593216.0, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.583139264089427, | |
| "grad_norm": 0.21673192590300572, | |
| "learning_rate": 2.6216775975146705e-05, | |
| "loss": 0.3555, | |
| "num_tokens": 1776820058.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.5877969259431766, | |
| "grad_norm": 0.19588832881605528, | |
| "learning_rate": 2.6130479806696584e-05, | |
| "loss": 0.3504, | |
| "num_tokens": 1782062938.0, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 1.5924545877969258, | |
| "grad_norm": 0.1901323469487215, | |
| "learning_rate": 2.6044183638246462e-05, | |
| "loss": 0.3575, | |
| "num_tokens": 1787305818.0, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.5971122496506753, | |
| "grad_norm": 0.23307185612370301, | |
| "learning_rate": 2.5957887469796343e-05, | |
| "loss": 0.3591, | |
| "num_tokens": 1792484830.0, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 1.6017699115044248, | |
| "grad_norm": 0.18955997048607617, | |
| "learning_rate": 2.587159130134622e-05, | |
| "loss": 0.3456, | |
| "num_tokens": 1797727710.0, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.6064275733581743, | |
| "grad_norm": 0.19919423944314782, | |
| "learning_rate": 2.5785295132896096e-05, | |
| "loss": 0.3526, | |
| "num_tokens": 1802970590.0, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.6110852352119236, | |
| "grad_norm": 0.2008337468455049, | |
| "learning_rate": 2.569899896444598e-05, | |
| "loss": 0.3646, | |
| "num_tokens": 1808213470.0, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.6157428970656729, | |
| "grad_norm": 0.2148858425734766, | |
| "learning_rate": 2.5612702795995856e-05, | |
| "loss": 0.3633, | |
| "num_tokens": 1813456350.0, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 1.6204005589194224, | |
| "grad_norm": 0.18402307523917638, | |
| "learning_rate": 2.552640662754574e-05, | |
| "loss": 0.3443, | |
| "num_tokens": 1818699230.0, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.6250582207731719, | |
| "grad_norm": 0.19347894305711252, | |
| "learning_rate": 2.5440110459095616e-05, | |
| "loss": 0.3727, | |
| "num_tokens": 1823942110.0, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 1.6297158826269214, | |
| "grad_norm": 0.18572609103365975, | |
| "learning_rate": 2.5353814290645494e-05, | |
| "loss": 0.3541, | |
| "num_tokens": 1829161686.0, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.6343735444806708, | |
| "grad_norm": 0.19627015288400848, | |
| "learning_rate": 2.5267518122195376e-05, | |
| "loss": 0.3609, | |
| "num_tokens": 1834404566.0, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.6390312063344201, | |
| "grad_norm": 0.20673127977064684, | |
| "learning_rate": 2.5181221953745254e-05, | |
| "loss": 0.3477, | |
| "num_tokens": 1839647446.0, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.6436888681881694, | |
| "grad_norm": 0.1986433721149521, | |
| "learning_rate": 2.5094925785295132e-05, | |
| "loss": 0.3485, | |
| "num_tokens": 1844890326.0, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 1.648346530041919, | |
| "grad_norm": 0.21357902897129788, | |
| "learning_rate": 2.5008629616845014e-05, | |
| "loss": 0.3468, | |
| "num_tokens": 1850121374.0, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.6530041918956684, | |
| "grad_norm": 0.16971779325075495, | |
| "learning_rate": 2.4922333448394892e-05, | |
| "loss": 0.3618, | |
| "num_tokens": 1855364254.0, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.6576618537494179, | |
| "grad_norm": 0.1936636533838936, | |
| "learning_rate": 2.483603727994477e-05, | |
| "loss": 0.3466, | |
| "num_tokens": 1860607134.0, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.6623195156031674, | |
| "grad_norm": 0.20039378245146786, | |
| "learning_rate": 2.4749741111494652e-05, | |
| "loss": 0.3571, | |
| "num_tokens": 1865850014.0, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.6669771774569166, | |
| "grad_norm": 0.19382602283515765, | |
| "learning_rate": 2.466344494304453e-05, | |
| "loss": 0.3497, | |
| "num_tokens": 1871092894.0, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.671634839310666, | |
| "grad_norm": 0.211364742709401, | |
| "learning_rate": 2.457714877459441e-05, | |
| "loss": 0.3472, | |
| "num_tokens": 1876335774.0, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 1.6762925011644154, | |
| "grad_norm": 0.19180430593648623, | |
| "learning_rate": 2.4490852606144287e-05, | |
| "loss": 0.3464, | |
| "num_tokens": 1881578654.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.680950163018165, | |
| "grad_norm": 0.2130514306266424, | |
| "learning_rate": 2.440455643769417e-05, | |
| "loss": 0.3539, | |
| "num_tokens": 1886793756.0, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 1.6856078248719144, | |
| "grad_norm": 0.2030821234662106, | |
| "learning_rate": 2.4318260269244047e-05, | |
| "loss": 0.3521, | |
| "num_tokens": 1892022954.0, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.6902654867256637, | |
| "grad_norm": 0.20444410391580545, | |
| "learning_rate": 2.4231964100793925e-05, | |
| "loss": 0.352, | |
| "num_tokens": 1897265834.0, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.6949231485794132, | |
| "grad_norm": 0.19509942761755505, | |
| "learning_rate": 2.4145667932343803e-05, | |
| "loss": 0.3398, | |
| "num_tokens": 1902508714.0, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.6995808104331624, | |
| "grad_norm": 0.19453315780096914, | |
| "learning_rate": 2.4059371763893685e-05, | |
| "loss": 0.3425, | |
| "num_tokens": 1907751594.0, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.704238472286912, | |
| "grad_norm": 0.23611653131971386, | |
| "learning_rate": 2.3973075595443563e-05, | |
| "loss": 0.3479, | |
| "num_tokens": 1912994474.0, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.7088961341406614, | |
| "grad_norm": 0.20356224901969366, | |
| "learning_rate": 2.388677942699344e-05, | |
| "loss": 0.3554, | |
| "num_tokens": 1918171986.0, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 1.713553795994411, | |
| "grad_norm": 0.23223162232171488, | |
| "learning_rate": 2.3800483258543323e-05, | |
| "loss": 0.3487, | |
| "num_tokens": 1923414866.0, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.7182114578481602, | |
| "grad_norm": 0.1944291211611693, | |
| "learning_rate": 2.37141870900932e-05, | |
| "loss": 0.353, | |
| "num_tokens": 1928641926.0, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.7228691197019095, | |
| "grad_norm": 0.2202839172320452, | |
| "learning_rate": 2.3627890921643083e-05, | |
| "loss": 0.3634, | |
| "num_tokens": 1933884806.0, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.727526781555659, | |
| "grad_norm": 0.19477900066568313, | |
| "learning_rate": 2.3541594753192957e-05, | |
| "loss": 0.344, | |
| "num_tokens": 1939127686.0, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.7321844434094085, | |
| "grad_norm": 0.19762015317203627, | |
| "learning_rate": 2.345529858474284e-05, | |
| "loss": 0.3672, | |
| "num_tokens": 1944370566.0, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.21142213900532833, | |
| "learning_rate": 2.3369002416292717e-05, | |
| "loss": 0.3471, | |
| "num_tokens": 1949613446.0, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 1.7414997671169075, | |
| "grad_norm": 0.1713123247092682, | |
| "learning_rate": 2.32827062478426e-05, | |
| "loss": 0.3522, | |
| "num_tokens": 1954856326.0, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.7461574289706567, | |
| "grad_norm": 0.20041446699241133, | |
| "learning_rate": 2.3196410079392474e-05, | |
| "loss": 0.3667, | |
| "num_tokens": 1960099206.0, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.750815090824406, | |
| "grad_norm": 0.19275848539341905, | |
| "learning_rate": 2.3110113910942355e-05, | |
| "loss": 0.3576, | |
| "num_tokens": 1965342086.0, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.7554727526781555, | |
| "grad_norm": 0.1866884152248216, | |
| "learning_rate": 2.3023817742492234e-05, | |
| "loss": 0.3577, | |
| "num_tokens": 1970582094.0, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 1.760130414531905, | |
| "grad_norm": 0.198425732588647, | |
| "learning_rate": 2.2937521574042115e-05, | |
| "loss": 0.3548, | |
| "num_tokens": 1975767338.0, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.7647880763856545, | |
| "grad_norm": 0.20133312155090208, | |
| "learning_rate": 2.285122540559199e-05, | |
| "loss": 0.3649, | |
| "num_tokens": 1980963902.0, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 1.7694457382394038, | |
| "grad_norm": 0.18940101588201483, | |
| "learning_rate": 2.276492923714187e-05, | |
| "loss": 0.3539, | |
| "num_tokens": 1986206782.0, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.7741034000931533, | |
| "grad_norm": 0.20406694515597523, | |
| "learning_rate": 2.267863306869175e-05, | |
| "loss": 0.3546, | |
| "num_tokens": 1991388204.0, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.7787610619469025, | |
| "grad_norm": 0.205464936537111, | |
| "learning_rate": 2.259233690024163e-05, | |
| "loss": 0.3456, | |
| "num_tokens": 1996631084.0, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.783418723800652, | |
| "grad_norm": 0.19832171301293253, | |
| "learning_rate": 2.250604073179151e-05, | |
| "loss": 0.342, | |
| "num_tokens": 2001873964.0, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 1.7880763856544015, | |
| "grad_norm": 0.20002311530633637, | |
| "learning_rate": 2.2419744563341388e-05, | |
| "loss": 0.3479, | |
| "num_tokens": 2007116844.0, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.792734047508151, | |
| "grad_norm": 0.2121285205201717, | |
| "learning_rate": 2.233344839489127e-05, | |
| "loss": 0.3581, | |
| "num_tokens": 2012298654.0, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.7973917093619003, | |
| "grad_norm": 0.18676189776784669, | |
| "learning_rate": 2.2247152226441148e-05, | |
| "loss": 0.3464, | |
| "num_tokens": 2017523752.0, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.8020493712156498, | |
| "grad_norm": 0.19096766874610935, | |
| "learning_rate": 2.2160856057991026e-05, | |
| "loss": 0.3418, | |
| "num_tokens": 2022766632.0, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.806707033069399, | |
| "grad_norm": 0.18363334824646743, | |
| "learning_rate": 2.2074559889540904e-05, | |
| "loss": 0.3452, | |
| "num_tokens": 2028009512.0, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.8113646949231486, | |
| "grad_norm": 0.19910259062104715, | |
| "learning_rate": 2.1988263721090786e-05, | |
| "loss": 0.3561, | |
| "num_tokens": 2033209432.0, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 1.816022356776898, | |
| "grad_norm": 0.1945409738344291, | |
| "learning_rate": 2.1901967552640664e-05, | |
| "loss": 0.3462, | |
| "num_tokens": 2038449340.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.8206800186306475, | |
| "grad_norm": 0.1918697300584693, | |
| "learning_rate": 2.1815671384190542e-05, | |
| "loss": 0.3508, | |
| "num_tokens": 2043692220.0, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 1.8253376804843968, | |
| "grad_norm": 0.20167268318651987, | |
| "learning_rate": 2.172937521574042e-05, | |
| "loss": 0.3517, | |
| "num_tokens": 2048935100.0, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.829995342338146, | |
| "grad_norm": 0.19999157120954503, | |
| "learning_rate": 2.1643079047290302e-05, | |
| "loss": 0.3531, | |
| "num_tokens": 2054123726.0, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.8346530041918956, | |
| "grad_norm": 0.2171556550330138, | |
| "learning_rate": 2.155678287884018e-05, | |
| "loss": 0.3485, | |
| "num_tokens": 2059366606.0, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.839310666045645, | |
| "grad_norm": 0.1936458133167658, | |
| "learning_rate": 2.147048671039006e-05, | |
| "loss": 0.3537, | |
| "num_tokens": 2064609486.0, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.8439683278993946, | |
| "grad_norm": 0.18210053995006473, | |
| "learning_rate": 2.1384190541939937e-05, | |
| "loss": 0.3468, | |
| "num_tokens": 2069852366.0, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.848625989753144, | |
| "grad_norm": 0.1992735548717075, | |
| "learning_rate": 2.129789437348982e-05, | |
| "loss": 0.3598, | |
| "num_tokens": 2075076504.0, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 1.8532836516068933, | |
| "grad_norm": 0.19201990731423438, | |
| "learning_rate": 2.1211598205039697e-05, | |
| "loss": 0.3554, | |
| "num_tokens": 2080319384.0, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.8579413134606426, | |
| "grad_norm": 0.18876762304804304, | |
| "learning_rate": 2.1125302036589575e-05, | |
| "loss": 0.3421, | |
| "num_tokens": 2085562264.0, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.8625989753143921, | |
| "grad_norm": 0.19023101667147826, | |
| "learning_rate": 2.1039005868139457e-05, | |
| "loss": 0.3515, | |
| "num_tokens": 2090805144.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.8672566371681416, | |
| "grad_norm": 0.22663085873082117, | |
| "learning_rate": 2.0952709699689335e-05, | |
| "loss": 0.3589, | |
| "num_tokens": 2096036892.0, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 1.871914299021891, | |
| "grad_norm": 0.18853462557649636, | |
| "learning_rate": 2.0866413531239216e-05, | |
| "loss": 0.3468, | |
| "num_tokens": 2101279772.0, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.8765719608756404, | |
| "grad_norm": 0.19493596788320008, | |
| "learning_rate": 2.078011736278909e-05, | |
| "loss": 0.3585, | |
| "num_tokens": 2106522652.0, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 1.8812296227293899, | |
| "grad_norm": 0.19894437999469852, | |
| "learning_rate": 2.0693821194338973e-05, | |
| "loss": 0.3442, | |
| "num_tokens": 2111743394.0, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.8858872845831391, | |
| "grad_norm": 0.19853665562057782, | |
| "learning_rate": 2.060752502588885e-05, | |
| "loss": 0.3509, | |
| "num_tokens": 2116986274.0, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.8905449464368886, | |
| "grad_norm": 0.20723274813649697, | |
| "learning_rate": 2.0521228857438733e-05, | |
| "loss": 0.348, | |
| "num_tokens": 2122229154.0, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.8952026082906381, | |
| "grad_norm": 0.18957779798248753, | |
| "learning_rate": 2.0434932688988608e-05, | |
| "loss": 0.346, | |
| "num_tokens": 2127472034.0, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 1.8998602701443876, | |
| "grad_norm": 1.2026559860404993, | |
| "learning_rate": 2.034863652053849e-05, | |
| "loss": 0.3635, | |
| "num_tokens": 2132714914.0, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.904517931998137, | |
| "grad_norm": 0.1945098070162303, | |
| "learning_rate": 2.0262340352088367e-05, | |
| "loss": 0.3575, | |
| "num_tokens": 2137957794.0, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 1.9091755938518864, | |
| "grad_norm": 0.21030924977357682, | |
| "learning_rate": 2.017604418363825e-05, | |
| "loss": 0.352, | |
| "num_tokens": 2143200674.0, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.9138332557056357, | |
| "grad_norm": 0.19484740573940798, | |
| "learning_rate": 2.0089748015188127e-05, | |
| "loss": 0.3523, | |
| "num_tokens": 2148417246.0, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.9184909175593852, | |
| "grad_norm": 0.2103812727795531, | |
| "learning_rate": 2.0003451846738005e-05, | |
| "loss": 0.3469, | |
| "num_tokens": 2153571210.0, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.9231485794131347, | |
| "grad_norm": 0.19536923184828212, | |
| "learning_rate": 1.9917155678287887e-05, | |
| "loss": 0.348, | |
| "num_tokens": 2158814090.0, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 1.9278062412668842, | |
| "grad_norm": 0.19377601993038385, | |
| "learning_rate": 1.9830859509837765e-05, | |
| "loss": 0.3517, | |
| "num_tokens": 2164056970.0, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.9324639031206334, | |
| "grad_norm": 0.1801265909451719, | |
| "learning_rate": 1.9744563341387643e-05, | |
| "loss": 0.35, | |
| "num_tokens": 2169299850.0, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.9371215649743827, | |
| "grad_norm": 0.19859650333088924, | |
| "learning_rate": 1.9658267172937522e-05, | |
| "loss": 0.3444, | |
| "num_tokens": 2174542730.0, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.9417792268281322, | |
| "grad_norm": 0.17667258267968372, | |
| "learning_rate": 1.9571971004487403e-05, | |
| "loss": 0.3552, | |
| "num_tokens": 2179720330.0, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 1.9464368886818817, | |
| "grad_norm": 0.16845965059921447, | |
| "learning_rate": 1.948567483603728e-05, | |
| "loss": 0.3396, | |
| "num_tokens": 2184963210.0, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.9510945505356312, | |
| "grad_norm": 0.1954643591448432, | |
| "learning_rate": 1.939937866758716e-05, | |
| "loss": 0.3476, | |
| "num_tokens": 2190165052.0, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 1.9557522123893807, | |
| "grad_norm": 0.1838999036973933, | |
| "learning_rate": 1.9313082499137038e-05, | |
| "loss": 0.3475, | |
| "num_tokens": 2195407932.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.96040987424313, | |
| "grad_norm": 0.18847540418823933, | |
| "learning_rate": 1.922678633068692e-05, | |
| "loss": 0.356, | |
| "num_tokens": 2200650812.0, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.9650675360968792, | |
| "grad_norm": 0.17911602185865064, | |
| "learning_rate": 1.9140490162236798e-05, | |
| "loss": 0.3417, | |
| "num_tokens": 2205893692.0, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.9697251979506287, | |
| "grad_norm": 0.2022059241272546, | |
| "learning_rate": 1.9054193993786676e-05, | |
| "loss": 0.3491, | |
| "num_tokens": 2211109840.0, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 1.9743828598043782, | |
| "grad_norm": 0.18090976759298982, | |
| "learning_rate": 1.8967897825336554e-05, | |
| "loss": 0.3522, | |
| "num_tokens": 2216352720.0, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.9790405216581277, | |
| "grad_norm": 0.19473887379297544, | |
| "learning_rate": 1.8881601656886436e-05, | |
| "loss": 0.3464, | |
| "num_tokens": 2221537760.0, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.983698183511877, | |
| "grad_norm": 0.17531749663446256, | |
| "learning_rate": 1.8795305488436314e-05, | |
| "loss": 0.3487, | |
| "num_tokens": 2226780640.0, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.9883558453656265, | |
| "grad_norm": 0.20812529277503486, | |
| "learning_rate": 1.8709009319986192e-05, | |
| "loss": 0.3553, | |
| "num_tokens": 2232023520.0, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 1.9930135072193758, | |
| "grad_norm": 0.19732858913976015, | |
| "learning_rate": 1.8622713151536074e-05, | |
| "loss": 0.3528, | |
| "num_tokens": 2237266400.0, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.9976711690731253, | |
| "grad_norm": 0.1844073713398463, | |
| "learning_rate": 1.8536416983085952e-05, | |
| "loss": 0.3547, | |
| "num_tokens": 2242509280.0, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 2.0018630647415, | |
| "grad_norm": 0.20466166303184266, | |
| "learning_rate": 1.8450120814635834e-05, | |
| "loss": 0.3314, | |
| "num_tokens": 2246888496.0, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.0065207265952494, | |
| "grad_norm": 0.222458418353215, | |
| "learning_rate": 1.836382464618571e-05, | |
| "loss": 0.2929, | |
| "num_tokens": 2252112814.0, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 2.0111783884489984, | |
| "grad_norm": 0.20316966252913563, | |
| "learning_rate": 1.827752847773559e-05, | |
| "loss": 0.2902, | |
| "num_tokens": 2257355694.0, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.015836050302748, | |
| "grad_norm": 0.20613499251609135, | |
| "learning_rate": 1.819123230928547e-05, | |
| "loss": 0.2876, | |
| "num_tokens": 2262579456.0, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 2.0204937121564974, | |
| "grad_norm": 0.19969713294398614, | |
| "learning_rate": 1.810493614083535e-05, | |
| "loss": 0.2776, | |
| "num_tokens": 2267822336.0, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 2.025151374010247, | |
| "grad_norm": 0.21082403122347504, | |
| "learning_rate": 1.8018639972385225e-05, | |
| "loss": 0.2942, | |
| "num_tokens": 2273022256.0, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.0298090358639964, | |
| "grad_norm": 0.19200539656516838, | |
| "learning_rate": 1.7932343803935107e-05, | |
| "loss": 0.2819, | |
| "num_tokens": 2278265136.0, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.034466697717746, | |
| "grad_norm": 0.21869392385233344, | |
| "learning_rate": 1.7846047635484985e-05, | |
| "loss": 0.2872, | |
| "num_tokens": 2283508016.0, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 2.039124359571495, | |
| "grad_norm": 0.2053569151384136, | |
| "learning_rate": 1.7759751467034866e-05, | |
| "loss": 0.2884, | |
| "num_tokens": 2288728034.0, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 2.0437820214252445, | |
| "grad_norm": 0.1950796592024243, | |
| "learning_rate": 1.767345529858474e-05, | |
| "loss": 0.2864, | |
| "num_tokens": 2293970914.0, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 2.048439683278994, | |
| "grad_norm": 0.1895400050270748, | |
| "learning_rate": 1.7587159130134623e-05, | |
| "loss": 0.2833, | |
| "num_tokens": 2299152298.0, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.0530973451327434, | |
| "grad_norm": 0.1909161506568314, | |
| "learning_rate": 1.75008629616845e-05, | |
| "loss": 0.2891, | |
| "num_tokens": 2304395178.0, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 2.057755006986493, | |
| "grad_norm": 0.1882935659455252, | |
| "learning_rate": 1.7414566793234383e-05, | |
| "loss": 0.2769, | |
| "num_tokens": 2309599416.0, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 2.062412668840242, | |
| "grad_norm": 0.19827918295315924, | |
| "learning_rate": 1.732827062478426e-05, | |
| "loss": 0.2867, | |
| "num_tokens": 2314842296.0, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 2.0670703306939915, | |
| "grad_norm": 0.18476311669265127, | |
| "learning_rate": 1.724197445633414e-05, | |
| "loss": 0.284, | |
| "num_tokens": 2320085176.0, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.071727992547741, | |
| "grad_norm": 0.2119727488460815, | |
| "learning_rate": 1.715567828788402e-05, | |
| "loss": 0.2833, | |
| "num_tokens": 2325328056.0, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.0763856544014905, | |
| "grad_norm": 0.18860318145230429, | |
| "learning_rate": 1.70693821194339e-05, | |
| "loss": 0.286, | |
| "num_tokens": 2330570936.0, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 2.08104331625524, | |
| "grad_norm": 0.18124836438668304, | |
| "learning_rate": 1.6983085950983777e-05, | |
| "loss": 0.2797, | |
| "num_tokens": 2335813816.0, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 2.0857009781089895, | |
| "grad_norm": 0.19139159457936905, | |
| "learning_rate": 1.6896789782533655e-05, | |
| "loss": 0.2897, | |
| "num_tokens": 2341056696.0, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.0903586399627385, | |
| "grad_norm": 0.19218676570039245, | |
| "learning_rate": 1.6810493614083537e-05, | |
| "loss": 0.2851, | |
| "num_tokens": 2346238006.0, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 2.095016301816488, | |
| "grad_norm": 0.1885592960230182, | |
| "learning_rate": 1.6724197445633415e-05, | |
| "loss": 0.281, | |
| "num_tokens": 2351480886.0, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.0996739636702375, | |
| "grad_norm": 0.19518158152530876, | |
| "learning_rate": 1.6637901277183294e-05, | |
| "loss": 0.2925, | |
| "num_tokens": 2356679066.0, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 2.104331625523987, | |
| "grad_norm": 0.18758939915457917, | |
| "learning_rate": 1.6551605108733172e-05, | |
| "loss": 0.2812, | |
| "num_tokens": 2361921946.0, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.1089892873777365, | |
| "grad_norm": 0.1921925862833243, | |
| "learning_rate": 1.6465308940283053e-05, | |
| "loss": 0.2914, | |
| "num_tokens": 2367121162.0, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 2.113646949231486, | |
| "grad_norm": 0.17967843798725472, | |
| "learning_rate": 1.637901277183293e-05, | |
| "loss": 0.2837, | |
| "num_tokens": 2372364042.0, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 2.118304611085235, | |
| "grad_norm": 0.18381257503357742, | |
| "learning_rate": 1.629271660338281e-05, | |
| "loss": 0.2844, | |
| "num_tokens": 2377606922.0, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.1229622729389845, | |
| "grad_norm": 0.18303104892808997, | |
| "learning_rate": 1.620642043493269e-05, | |
| "loss": 0.2836, | |
| "num_tokens": 2382827664.0, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.127619934792734, | |
| "grad_norm": 0.17915349866213887, | |
| "learning_rate": 1.612012426648257e-05, | |
| "loss": 0.2886, | |
| "num_tokens": 2388027652.0, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 2.1322775966464835, | |
| "grad_norm": 0.18782066980876427, | |
| "learning_rate": 1.6033828098032448e-05, | |
| "loss": 0.2875, | |
| "num_tokens": 2393270532.0, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 2.136935258500233, | |
| "grad_norm": 0.18789300808432408, | |
| "learning_rate": 1.5947531929582326e-05, | |
| "loss": 0.2902, | |
| "num_tokens": 2398467742.0, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 2.1415929203539825, | |
| "grad_norm": 0.19312735251094643, | |
| "learning_rate": 1.5861235761132208e-05, | |
| "loss": 0.2926, | |
| "num_tokens": 2403710622.0, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.1462505822077316, | |
| "grad_norm": 0.19749618542525005, | |
| "learning_rate": 1.5774939592682086e-05, | |
| "loss": 0.2873, | |
| "num_tokens": 2408951214.0, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 2.150908244061481, | |
| "grad_norm": 0.17506466471685322, | |
| "learning_rate": 1.5688643424231964e-05, | |
| "loss": 0.2759, | |
| "num_tokens": 2414194094.0, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 2.1555659059152306, | |
| "grad_norm": 0.18345051978977678, | |
| "learning_rate": 1.5602347255781842e-05, | |
| "loss": 0.2816, | |
| "num_tokens": 2419371694.0, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 2.16022356776898, | |
| "grad_norm": 0.18881546398847446, | |
| "learning_rate": 1.5516051087331724e-05, | |
| "loss": 0.2802, | |
| "num_tokens": 2424614574.0, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.1648812296227296, | |
| "grad_norm": 0.20510019062629636, | |
| "learning_rate": 1.5429754918881602e-05, | |
| "loss": 0.2862, | |
| "num_tokens": 2429841120.0, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.1695388914764786, | |
| "grad_norm": 0.17780457830497634, | |
| "learning_rate": 1.534345875043148e-05, | |
| "loss": 0.2868, | |
| "num_tokens": 2435040552.0, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 2.174196553330228, | |
| "grad_norm": 0.18888800975493006, | |
| "learning_rate": 1.525716258198136e-05, | |
| "loss": 0.2814, | |
| "num_tokens": 2440283432.0, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 2.1788542151839776, | |
| "grad_norm": 0.19785987444812264, | |
| "learning_rate": 1.517086641353124e-05, | |
| "loss": 0.2846, | |
| "num_tokens": 2445485274.0, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.183511877037727, | |
| "grad_norm": 0.17883301445147745, | |
| "learning_rate": 1.508457024508112e-05, | |
| "loss": 0.29, | |
| "num_tokens": 2450728154.0, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 2.1881695388914766, | |
| "grad_norm": 0.1947043353091113, | |
| "learning_rate": 1.4998274076630997e-05, | |
| "loss": 0.2839, | |
| "num_tokens": 2455971034.0, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.192827200745226, | |
| "grad_norm": 0.1770640125215528, | |
| "learning_rate": 1.4911977908180877e-05, | |
| "loss": 0.2804, | |
| "num_tokens": 2461186370.0, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 2.197484862598975, | |
| "grad_norm": 0.2122232291610023, | |
| "learning_rate": 1.4825681739730757e-05, | |
| "loss": 0.2908, | |
| "num_tokens": 2466429250.0, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.2021425244527246, | |
| "grad_norm": 0.18924595153885262, | |
| "learning_rate": 1.4739385571280637e-05, | |
| "loss": 0.2888, | |
| "num_tokens": 2471561792.0, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 2.206800186306474, | |
| "grad_norm": 0.20138907434962766, | |
| "learning_rate": 1.4653089402830513e-05, | |
| "loss": 0.2882, | |
| "num_tokens": 2476744062.0, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 2.2114578481602236, | |
| "grad_norm": 0.2146833252038963, | |
| "learning_rate": 1.4566793234380393e-05, | |
| "loss": 0.2885, | |
| "num_tokens": 2481986942.0, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.216115510013973, | |
| "grad_norm": 0.18561110122941038, | |
| "learning_rate": 1.4480497065930273e-05, | |
| "loss": 0.2849, | |
| "num_tokens": 2487212052.0, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.2207731718677226, | |
| "grad_norm": 0.18442145502966026, | |
| "learning_rate": 1.4394200897480153e-05, | |
| "loss": 0.2967, | |
| "num_tokens": 2492454932.0, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 2.2254308337214717, | |
| "grad_norm": 0.16935577654858627, | |
| "learning_rate": 1.4307904729030031e-05, | |
| "loss": 0.2896, | |
| "num_tokens": 2497656374.0, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.230088495575221, | |
| "grad_norm": 0.19183522003950687, | |
| "learning_rate": 1.4221608560579911e-05, | |
| "loss": 0.2859, | |
| "num_tokens": 2502899254.0, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 2.2347461574289706, | |
| "grad_norm": 0.18220975585984353, | |
| "learning_rate": 1.4135312392129791e-05, | |
| "loss": 0.2756, | |
| "num_tokens": 2508113316.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.23940381928272, | |
| "grad_norm": 0.180730928951134, | |
| "learning_rate": 1.404901622367967e-05, | |
| "loss": 0.287, | |
| "num_tokens": 2513356196.0, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 2.2440614811364696, | |
| "grad_norm": 0.18965210318025177, | |
| "learning_rate": 1.3962720055229547e-05, | |
| "loss": 0.2863, | |
| "num_tokens": 2518599076.0, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.248719142990219, | |
| "grad_norm": 0.17257945172194086, | |
| "learning_rate": 1.3876423886779427e-05, | |
| "loss": 0.2847, | |
| "num_tokens": 2523815648.0, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 2.253376804843968, | |
| "grad_norm": 0.191482066683056, | |
| "learning_rate": 1.3790127718329307e-05, | |
| "loss": 0.2757, | |
| "num_tokens": 2529044846.0, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.2580344666977177, | |
| "grad_norm": 0.1830413316351366, | |
| "learning_rate": 1.3703831549879187e-05, | |
| "loss": 0.2889, | |
| "num_tokens": 2534287726.0, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.262692128551467, | |
| "grad_norm": 0.1840154400557875, | |
| "learning_rate": 1.3617535381429064e-05, | |
| "loss": 0.2793, | |
| "num_tokens": 2539530606.0, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.2673497904052167, | |
| "grad_norm": 0.19256723556560748, | |
| "learning_rate": 1.3531239212978944e-05, | |
| "loss": 0.288, | |
| "num_tokens": 2544773486.0, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 2.272007452258966, | |
| "grad_norm": 0.17658811837538962, | |
| "learning_rate": 1.3444943044528824e-05, | |
| "loss": 0.2836, | |
| "num_tokens": 2549964918.0, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.276665114112715, | |
| "grad_norm": 0.19497865199271125, | |
| "learning_rate": 1.3358646876078703e-05, | |
| "loss": 0.2843, | |
| "num_tokens": 2555170596.0, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 2.2813227759664647, | |
| "grad_norm": 0.17847664685164866, | |
| "learning_rate": 1.327235070762858e-05, | |
| "loss": 0.2816, | |
| "num_tokens": 2560411630.0, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.285980437820214, | |
| "grad_norm": 0.18056534918434158, | |
| "learning_rate": 1.318605453917846e-05, | |
| "loss": 0.2825, | |
| "num_tokens": 2565604238.0, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 2.2906380996739637, | |
| "grad_norm": 0.18186829013218692, | |
| "learning_rate": 1.309975837072834e-05, | |
| "loss": 0.2785, | |
| "num_tokens": 2570828142.0, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.295295761527713, | |
| "grad_norm": 0.19629580934115015, | |
| "learning_rate": 1.301346220227822e-05, | |
| "loss": 0.2953, | |
| "num_tokens": 2576047788.0, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 2.2999534233814627, | |
| "grad_norm": 0.20244354201672302, | |
| "learning_rate": 1.2927166033828098e-05, | |
| "loss": 0.2834, | |
| "num_tokens": 2581290668.0, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.3046110852352117, | |
| "grad_norm": 0.2079773331282868, | |
| "learning_rate": 1.2840869865377978e-05, | |
| "loss": 0.291, | |
| "num_tokens": 2586517510.0, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.3092687470889612, | |
| "grad_norm": 0.18386221467315658, | |
| "learning_rate": 1.2754573696927858e-05, | |
| "loss": 0.2824, | |
| "num_tokens": 2591760390.0, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.3139264089427107, | |
| "grad_norm": 0.18122335457008204, | |
| "learning_rate": 1.2668277528477738e-05, | |
| "loss": 0.2912, | |
| "num_tokens": 2597003270.0, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 2.3185840707964602, | |
| "grad_norm": 0.1896463646920755, | |
| "learning_rate": 1.2581981360027614e-05, | |
| "loss": 0.2928, | |
| "num_tokens": 2602243278.0, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.3232417326502097, | |
| "grad_norm": 0.17387509820568628, | |
| "learning_rate": 1.2495685191577494e-05, | |
| "loss": 0.2838, | |
| "num_tokens": 2607486158.0, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 2.3278993945039588, | |
| "grad_norm": 0.16584638362980814, | |
| "learning_rate": 1.2409389023127374e-05, | |
| "loss": 0.2756, | |
| "num_tokens": 2612655868.0, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.3325570563577083, | |
| "grad_norm": 0.18235002198953293, | |
| "learning_rate": 1.2323092854677252e-05, | |
| "loss": 0.2844, | |
| "num_tokens": 2617898748.0, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 2.3372147182114578, | |
| "grad_norm": 0.1776373008486078, | |
| "learning_rate": 1.2236796686227132e-05, | |
| "loss": 0.2835, | |
| "num_tokens": 2623141628.0, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 2.3418723800652073, | |
| "grad_norm": 0.18857596897653406, | |
| "learning_rate": 1.215050051777701e-05, | |
| "loss": 0.2851, | |
| "num_tokens": 2628384508.0, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 2.3465300419189568, | |
| "grad_norm": 0.19454812018849615, | |
| "learning_rate": 1.206420434932689e-05, | |
| "loss": 0.2874, | |
| "num_tokens": 2633569752.0, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.3511877037727063, | |
| "grad_norm": 0.18437277158481002, | |
| "learning_rate": 1.1977908180876769e-05, | |
| "loss": 0.2878, | |
| "num_tokens": 2638812632.0, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.3558453656264557, | |
| "grad_norm": 0.1783358484367809, | |
| "learning_rate": 1.1891612012426649e-05, | |
| "loss": 0.2856, | |
| "num_tokens": 2644055512.0, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 2.360503027480205, | |
| "grad_norm": 0.19407167700482955, | |
| "learning_rate": 1.1805315843976528e-05, | |
| "loss": 0.2782, | |
| "num_tokens": 2649298392.0, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 2.3651606893339543, | |
| "grad_norm": 0.18070351694345052, | |
| "learning_rate": 1.1719019675526408e-05, | |
| "loss": 0.2758, | |
| "num_tokens": 2654541272.0, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.369818351187704, | |
| "grad_norm": 0.17087538084804843, | |
| "learning_rate": 1.1632723507076287e-05, | |
| "loss": 0.2884, | |
| "num_tokens": 2659784152.0, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 2.3744760130414533, | |
| "grad_norm": 0.18681698228539706, | |
| "learning_rate": 1.1546427338626167e-05, | |
| "loss": 0.2874, | |
| "num_tokens": 2665027032.0, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.3791336748952028, | |
| "grad_norm": 0.18279345120527926, | |
| "learning_rate": 1.1460131170176045e-05, | |
| "loss": 0.2848, | |
| "num_tokens": 2670269912.0, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 2.383791336748952, | |
| "grad_norm": 0.2325755383458184, | |
| "learning_rate": 1.1373835001725925e-05, | |
| "loss": 0.285, | |
| "num_tokens": 2675512792.0, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.3884489986027013, | |
| "grad_norm": 0.19504765012097386, | |
| "learning_rate": 1.1287538833275803e-05, | |
| "loss": 0.2858, | |
| "num_tokens": 2680749892.0, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 2.393106660456451, | |
| "grad_norm": 0.18358175055002107, | |
| "learning_rate": 1.1201242664825683e-05, | |
| "loss": 0.2886, | |
| "num_tokens": 2685992772.0, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 2.3977643223102003, | |
| "grad_norm": 0.1848932611484623, | |
| "learning_rate": 1.1114946496375561e-05, | |
| "loss": 0.2873, | |
| "num_tokens": 2691235652.0, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.40242198416395, | |
| "grad_norm": 0.17873198879932573, | |
| "learning_rate": 1.1028650327925441e-05, | |
| "loss": 0.2811, | |
| "num_tokens": 2696478532.0, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.4070796460176993, | |
| "grad_norm": 0.18442209133634158, | |
| "learning_rate": 1.094235415947532e-05, | |
| "loss": 0.2829, | |
| "num_tokens": 2701721412.0, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 2.4117373078714484, | |
| "grad_norm": 0.16595581037127025, | |
| "learning_rate": 1.0856057991025199e-05, | |
| "loss": 0.2762, | |
| "num_tokens": 2706964292.0, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 2.416394969725198, | |
| "grad_norm": 0.18418899241104994, | |
| "learning_rate": 1.0769761822575077e-05, | |
| "loss": 0.2918, | |
| "num_tokens": 2712088070.0, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.18590411037862897, | |
| "learning_rate": 1.0683465654124957e-05, | |
| "loss": 0.2873, | |
| "num_tokens": 2717309806.0, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.425710293432697, | |
| "grad_norm": 0.17830003825808874, | |
| "learning_rate": 1.0597169485674835e-05, | |
| "loss": 0.2812, | |
| "num_tokens": 2722552686.0, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 2.4303679552864463, | |
| "grad_norm": 0.1801718337960902, | |
| "learning_rate": 1.0510873317224715e-05, | |
| "loss": 0.2779, | |
| "num_tokens": 2727795566.0, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 2.4350256171401954, | |
| "grad_norm": 0.18076032191543717, | |
| "learning_rate": 1.0424577148774595e-05, | |
| "loss": 0.2827, | |
| "num_tokens": 2733038446.0, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 2.439683278993945, | |
| "grad_norm": 0.17912382625564433, | |
| "learning_rate": 1.0338280980324475e-05, | |
| "loss": 0.2851, | |
| "num_tokens": 2738249702.0, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.4443409408476944, | |
| "grad_norm": 0.17067019304586425, | |
| "learning_rate": 1.0251984811874353e-05, | |
| "loss": 0.2793, | |
| "num_tokens": 2743492582.0, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.448998602701444, | |
| "grad_norm": 0.18248217715274923, | |
| "learning_rate": 1.0165688643424233e-05, | |
| "loss": 0.28, | |
| "num_tokens": 2748735462.0, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 2.4536562645551934, | |
| "grad_norm": 0.17238278253867284, | |
| "learning_rate": 1.0079392474974112e-05, | |
| "loss": 0.2859, | |
| "num_tokens": 2753978342.0, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 2.458313926408943, | |
| "grad_norm": 0.18512124552156498, | |
| "learning_rate": 9.993096306523992e-06, | |
| "loss": 0.2864, | |
| "num_tokens": 2759221222.0, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.4629715882626924, | |
| "grad_norm": 0.1890621280556425, | |
| "learning_rate": 9.90680013807387e-06, | |
| "loss": 0.2867, | |
| "num_tokens": 2764464102.0, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 2.4676292501164414, | |
| "grad_norm": 0.19251424249866586, | |
| "learning_rate": 9.82050396962375e-06, | |
| "loss": 0.2821, | |
| "num_tokens": 2769706982.0, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.472286911970191, | |
| "grad_norm": 0.18670639810765133, | |
| "learning_rate": 9.734207801173628e-06, | |
| "loss": 0.2841, | |
| "num_tokens": 2774879410.0, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 2.4769445738239404, | |
| "grad_norm": 0.17983653823647472, | |
| "learning_rate": 9.647911632723508e-06, | |
| "loss": 0.2777, | |
| "num_tokens": 2780111158.0, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.48160223567769, | |
| "grad_norm": 0.18915245656956345, | |
| "learning_rate": 9.561615464273386e-06, | |
| "loss": 0.2823, | |
| "num_tokens": 2785354038.0, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 2.4862598975314394, | |
| "grad_norm": 0.18851778549805698, | |
| "learning_rate": 9.475319295823266e-06, | |
| "loss": 0.2824, | |
| "num_tokens": 2790596918.0, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 2.4909175593851884, | |
| "grad_norm": 0.16948307025132278, | |
| "learning_rate": 9.389023127373144e-06, | |
| "loss": 0.2795, | |
| "num_tokens": 2795839798.0, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.495575221238938, | |
| "grad_norm": 0.17031573633766003, | |
| "learning_rate": 9.302726958923024e-06, | |
| "loss": 0.2797, | |
| "num_tokens": 2801082678.0, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.5002328830926874, | |
| "grad_norm": 0.185187415872164, | |
| "learning_rate": 9.216430790472904e-06, | |
| "loss": 0.2852, | |
| "num_tokens": 2806325558.0, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 2.504890544946437, | |
| "grad_norm": 0.18203758967949804, | |
| "learning_rate": 9.130134622022784e-06, | |
| "loss": 0.2756, | |
| "num_tokens": 2811568438.0, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.5095482068001864, | |
| "grad_norm": 0.18684276396851224, | |
| "learning_rate": 9.043838453572662e-06, | |
| "loss": 0.2809, | |
| "num_tokens": 2816811318.0, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 2.514205868653936, | |
| "grad_norm": 0.17872665434221408, | |
| "learning_rate": 8.957542285122542e-06, | |
| "loss": 0.2779, | |
| "num_tokens": 2822054198.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.5188635305076854, | |
| "grad_norm": 0.19613743907526726, | |
| "learning_rate": 8.87124611667242e-06, | |
| "loss": 0.2852, | |
| "num_tokens": 2827297078.0, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 2.5235211923614345, | |
| "grad_norm": 0.1785699414658985, | |
| "learning_rate": 8.7849499482223e-06, | |
| "loss": 0.2822, | |
| "num_tokens": 2832539958.0, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.528178854215184, | |
| "grad_norm": 0.18186283447461787, | |
| "learning_rate": 8.698653779772179e-06, | |
| "loss": 0.2845, | |
| "num_tokens": 2837782838.0, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 2.5328365160689335, | |
| "grad_norm": 0.17329701477582082, | |
| "learning_rate": 8.612357611322058e-06, | |
| "loss": 0.2925, | |
| "num_tokens": 2843025718.0, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.537494177922683, | |
| "grad_norm": 0.18514608064447788, | |
| "learning_rate": 8.526061442871937e-06, | |
| "loss": 0.2833, | |
| "num_tokens": 2848268598.0, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.542151839776432, | |
| "grad_norm": 0.17671204755555353, | |
| "learning_rate": 8.439765274421817e-06, | |
| "loss": 0.292, | |
| "num_tokens": 2853458630.0, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.5468095016301815, | |
| "grad_norm": 0.18719289844083756, | |
| "learning_rate": 8.353469105971695e-06, | |
| "loss": 0.2781, | |
| "num_tokens": 2858701510.0, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 2.551467163483931, | |
| "grad_norm": 0.18554228730156927, | |
| "learning_rate": 8.267172937521575e-06, | |
| "loss": 0.2778, | |
| "num_tokens": 2863944390.0, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.5561248253376805, | |
| "grad_norm": 0.1888358161746658, | |
| "learning_rate": 8.180876769071453e-06, | |
| "loss": 0.2821, | |
| "num_tokens": 2869187270.0, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 2.56078248719143, | |
| "grad_norm": 0.17167191763957684, | |
| "learning_rate": 8.094580600621333e-06, | |
| "loss": 0.2812, | |
| "num_tokens": 2874430150.0, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.5654401490451795, | |
| "grad_norm": 0.16853579106861402, | |
| "learning_rate": 8.008284432171211e-06, | |
| "loss": 0.2782, | |
| "num_tokens": 2879673030.0, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 2.570097810898929, | |
| "grad_norm": 0.1841774036066754, | |
| "learning_rate": 7.921988263721091e-06, | |
| "loss": 0.2829, | |
| "num_tokens": 2884915910.0, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.574755472752678, | |
| "grad_norm": 0.19813239208470615, | |
| "learning_rate": 7.835692095270971e-06, | |
| "loss": 0.2797, | |
| "num_tokens": 2890124422.0, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 2.5794131346064275, | |
| "grad_norm": 0.17903339643216915, | |
| "learning_rate": 7.749395926820851e-06, | |
| "loss": 0.2811, | |
| "num_tokens": 2895367302.0, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.584070796460177, | |
| "grad_norm": 0.18029365642367381, | |
| "learning_rate": 7.663099758370729e-06, | |
| "loss": 0.2767, | |
| "num_tokens": 2900606020.0, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.5887284583139265, | |
| "grad_norm": 0.179932562552873, | |
| "learning_rate": 7.576803589920608e-06, | |
| "loss": 0.2858, | |
| "num_tokens": 2905848900.0, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.5933861201676756, | |
| "grad_norm": 0.18170600624608446, | |
| "learning_rate": 7.490507421470487e-06, | |
| "loss": 0.2866, | |
| "num_tokens": 2911048450.0, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 2.598043782021425, | |
| "grad_norm": 0.1700919572735666, | |
| "learning_rate": 7.4042112530203655e-06, | |
| "loss": 0.2891, | |
| "num_tokens": 2916231888.0, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.6027014438751745, | |
| "grad_norm": 0.17818321488145758, | |
| "learning_rate": 7.317915084570245e-06, | |
| "loss": 0.2828, | |
| "num_tokens": 2921474768.0, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 2.607359105728924, | |
| "grad_norm": 0.17704262070315935, | |
| "learning_rate": 7.231618916120124e-06, | |
| "loss": 0.2838, | |
| "num_tokens": 2926717648.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.6120167675826735, | |
| "grad_norm": 0.175240680166841, | |
| "learning_rate": 7.1453227476700035e-06, | |
| "loss": 0.2821, | |
| "num_tokens": 2931960528.0, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 2.616674429436423, | |
| "grad_norm": 0.19619717844445844, | |
| "learning_rate": 7.059026579219883e-06, | |
| "loss": 0.2832, | |
| "num_tokens": 2937181824.0, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.6213320912901725, | |
| "grad_norm": 0.16580789294034584, | |
| "learning_rate": 6.9727304107697625e-06, | |
| "loss": 0.2765, | |
| "num_tokens": 2942424704.0, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 2.625989753143922, | |
| "grad_norm": 0.18240869212063554, | |
| "learning_rate": 6.886434242319641e-06, | |
| "loss": 0.2857, | |
| "num_tokens": 2947667584.0, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.630647414997671, | |
| "grad_norm": 0.17675152445137393, | |
| "learning_rate": 6.800138073869521e-06, | |
| "loss": 0.2878, | |
| "num_tokens": 2952910464.0, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 2.6353050768514206, | |
| "grad_norm": 0.1936430042230673, | |
| "learning_rate": 6.713841905419399e-06, | |
| "loss": 0.2797, | |
| "num_tokens": 2958153344.0, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.63996273870517, | |
| "grad_norm": 0.18979558867919527, | |
| "learning_rate": 6.627545736969279e-06, | |
| "loss": 0.2888, | |
| "num_tokens": 2963396224.0, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 2.6446204005589196, | |
| "grad_norm": 0.16863770196483935, | |
| "learning_rate": 6.541249568519157e-06, | |
| "loss": 0.2919, | |
| "num_tokens": 2968590242.0, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.6492780624126686, | |
| "grad_norm": 0.18997469828362676, | |
| "learning_rate": 6.454953400069037e-06, | |
| "loss": 0.2832, | |
| "num_tokens": 2973833122.0, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 2.653935724266418, | |
| "grad_norm": 0.17321367794198167, | |
| "learning_rate": 6.368657231618916e-06, | |
| "loss": 0.2885, | |
| "num_tokens": 2979031028.0, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.6585933861201676, | |
| "grad_norm": 0.17990951872316197, | |
| "learning_rate": 6.282361063168796e-06, | |
| "loss": 0.2785, | |
| "num_tokens": 2984273908.0, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 2.663251047973917, | |
| "grad_norm": 0.18123182801507975, | |
| "learning_rate": 6.196064894718675e-06, | |
| "loss": 0.2855, | |
| "num_tokens": 2989434158.0, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.6679087098276666, | |
| "grad_norm": 0.17259686444859876, | |
| "learning_rate": 6.109768726268554e-06, | |
| "loss": 0.2846, | |
| "num_tokens": 2994677038.0, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 2.672566371681416, | |
| "grad_norm": 0.1907813281424595, | |
| "learning_rate": 6.023472557818433e-06, | |
| "loss": 0.2818, | |
| "num_tokens": 2999919918.0, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.6772240335351656, | |
| "grad_norm": 0.17349507625528626, | |
| "learning_rate": 5.937176389368312e-06, | |
| "loss": 0.2821, | |
| "num_tokens": 3005147388.0, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 2.6818816953889146, | |
| "grad_norm": 0.17968937343762775, | |
| "learning_rate": 5.850880220918191e-06, | |
| "loss": 0.2834, | |
| "num_tokens": 3010305506.0, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.686539357242664, | |
| "grad_norm": 0.17340253364289687, | |
| "learning_rate": 5.76458405246807e-06, | |
| "loss": 0.2904, | |
| "num_tokens": 3015490322.0, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 2.6911970190964136, | |
| "grad_norm": 0.1723305299091664, | |
| "learning_rate": 5.67828788401795e-06, | |
| "loss": 0.2814, | |
| "num_tokens": 3020733202.0, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.695854680950163, | |
| "grad_norm": 0.18451566048268914, | |
| "learning_rate": 5.591991715567829e-06, | |
| "loss": 0.2792, | |
| "num_tokens": 3025976082.0, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 2.700512342803912, | |
| "grad_norm": 0.165663086067545, | |
| "learning_rate": 5.5056955471177085e-06, | |
| "loss": 0.2825, | |
| "num_tokens": 3031218962.0, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.7051700046576617, | |
| "grad_norm": 0.16217642076383101, | |
| "learning_rate": 5.4193993786675876e-06, | |
| "loss": 0.2753, | |
| "num_tokens": 3036461842.0, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 2.709827666511411, | |
| "grad_norm": 0.16976229308170754, | |
| "learning_rate": 5.333103210217467e-06, | |
| "loss": 0.2811, | |
| "num_tokens": 3041676422.0, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.7144853283651607, | |
| "grad_norm": 0.18889824902243735, | |
| "learning_rate": 5.246807041767346e-06, | |
| "loss": 0.288, | |
| "num_tokens": 3046901580.0, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 2.71914299021891, | |
| "grad_norm": 0.17544467084852128, | |
| "learning_rate": 5.160510873317225e-06, | |
| "loss": 0.2797, | |
| "num_tokens": 3052144460.0, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.7238006520726596, | |
| "grad_norm": 0.17821279060257453, | |
| "learning_rate": 5.074214704867105e-06, | |
| "loss": 0.2885, | |
| "num_tokens": 3057357558.0, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 2.728458313926409, | |
| "grad_norm": 0.16612575807045404, | |
| "learning_rate": 4.987918536416984e-06, | |
| "loss": 0.287, | |
| "num_tokens": 3062600438.0, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 2.7331159757801586, | |
| "grad_norm": 0.1699430030564144, | |
| "learning_rate": 4.901622367966863e-06, | |
| "loss": 0.2816, | |
| "num_tokens": 3067843318.0, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 2.7377736376339077, | |
| "grad_norm": 0.17975135175844872, | |
| "learning_rate": 4.815326199516742e-06, | |
| "loss": 0.2824, | |
| "num_tokens": 3073086198.0, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.742431299487657, | |
| "grad_norm": 0.17865434016729267, | |
| "learning_rate": 4.729030031066621e-06, | |
| "loss": 0.2904, | |
| "num_tokens": 3078319256.0, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 2.7470889613414067, | |
| "grad_norm": 0.1748671313142682, | |
| "learning_rate": 4.6427338626165e-06, | |
| "loss": 0.2792, | |
| "num_tokens": 3083562136.0, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.751746623195156, | |
| "grad_norm": 0.18607290999773793, | |
| "learning_rate": 4.556437694166379e-06, | |
| "loss": 0.2829, | |
| "num_tokens": 3088805016.0, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 2.7564042850489052, | |
| "grad_norm": 0.176671447418757, | |
| "learning_rate": 4.470141525716258e-06, | |
| "loss": 0.2854, | |
| "num_tokens": 3094047896.0, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.7610619469026547, | |
| "grad_norm": 0.16222888490227458, | |
| "learning_rate": 4.383845357266138e-06, | |
| "loss": 0.2842, | |
| "num_tokens": 3099290776.0, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 2.765719608756404, | |
| "grad_norm": 0.16870751856088081, | |
| "learning_rate": 4.297549188816017e-06, | |
| "loss": 0.2857, | |
| "num_tokens": 3104533656.0, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.7703772706101537, | |
| "grad_norm": 0.17948012087188597, | |
| "learning_rate": 4.211253020365896e-06, | |
| "loss": 0.281, | |
| "num_tokens": 3109751024.0, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 2.775034932463903, | |
| "grad_norm": 0.18108487558480812, | |
| "learning_rate": 4.124956851915775e-06, | |
| "loss": 0.2803, | |
| "num_tokens": 3114993904.0, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.7796925943176527, | |
| "grad_norm": 0.1781786879296704, | |
| "learning_rate": 4.0386606834656544e-06, | |
| "loss": 0.2877, | |
| "num_tokens": 3120236784.0, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 2.784350256171402, | |
| "grad_norm": 0.1688708102823595, | |
| "learning_rate": 3.9523645150155335e-06, | |
| "loss": 0.2836, | |
| "num_tokens": 3125461726.0, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 2.7890079180251512, | |
| "grad_norm": 0.17358002603794873, | |
| "learning_rate": 3.8660683465654126e-06, | |
| "loss": 0.2805, | |
| "num_tokens": 3130604236.0, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 2.7936655798789007, | |
| "grad_norm": 0.16993181340352262, | |
| "learning_rate": 3.779772178115292e-06, | |
| "loss": 0.2863, | |
| "num_tokens": 3135847116.0, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.7983232417326502, | |
| "grad_norm": 0.16400870434081694, | |
| "learning_rate": 3.693476009665171e-06, | |
| "loss": 0.2795, | |
| "num_tokens": 3141089996.0, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 2.8029809035863997, | |
| "grad_norm": 0.16860079783639653, | |
| "learning_rate": 3.6071798412150506e-06, | |
| "loss": 0.2879, | |
| "num_tokens": 3146332876.0, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 2.807638565440149, | |
| "grad_norm": 0.16869830224404767, | |
| "learning_rate": 3.5208836727649297e-06, | |
| "loss": 0.272, | |
| "num_tokens": 3151575756.0, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 2.8122962272938983, | |
| "grad_norm": 0.17013037406938528, | |
| "learning_rate": 3.434587504314809e-06, | |
| "loss": 0.2848, | |
| "num_tokens": 3156818636.0, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.8169538891476478, | |
| "grad_norm": 0.16142434942955675, | |
| "learning_rate": 3.348291335864688e-06, | |
| "loss": 0.2811, | |
| "num_tokens": 3162007886.0, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 2.8216115510013973, | |
| "grad_norm": 0.16142479636656198, | |
| "learning_rate": 3.2619951674145674e-06, | |
| "loss": 0.2803, | |
| "num_tokens": 3167202338.0, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.8262692128551468, | |
| "grad_norm": 0.1711508514298411, | |
| "learning_rate": 3.1756989989644464e-06, | |
| "loss": 0.2827, | |
| "num_tokens": 3172445218.0, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 2.8309268747088963, | |
| "grad_norm": 0.18220156053713799, | |
| "learning_rate": 3.089402830514325e-06, | |
| "loss": 0.2886, | |
| "num_tokens": 3177682714.0, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.8355845365626458, | |
| "grad_norm": 0.1578762490127166, | |
| "learning_rate": 3.0031066620642046e-06, | |
| "loss": 0.2804, | |
| "num_tokens": 3182925594.0, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 2.840242198416395, | |
| "grad_norm": 0.18072000631431973, | |
| "learning_rate": 2.9168104936140837e-06, | |
| "loss": 0.2808, | |
| "num_tokens": 3188162430.0, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.8448998602701443, | |
| "grad_norm": 0.1847757384746603, | |
| "learning_rate": 2.8305143251639627e-06, | |
| "loss": 0.2871, | |
| "num_tokens": 3193342410.0, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 2.849557522123894, | |
| "grad_norm": 0.16740240324280342, | |
| "learning_rate": 2.7442181567138422e-06, | |
| "loss": 0.2898, | |
| "num_tokens": 3198585290.0, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.8542151839776433, | |
| "grad_norm": 0.1864233088534882, | |
| "learning_rate": 2.6579219882637213e-06, | |
| "loss": 0.286, | |
| "num_tokens": 3203814572.0, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 2.858872845831393, | |
| "grad_norm": 0.16528790852909972, | |
| "learning_rate": 2.5716258198136004e-06, | |
| "loss": 0.2753, | |
| "num_tokens": 3209051374.0, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.863530507685142, | |
| "grad_norm": 0.1736020259915108, | |
| "learning_rate": 2.4853296513634795e-06, | |
| "loss": 0.2825, | |
| "num_tokens": 3214294254.0, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 2.8681881695388913, | |
| "grad_norm": 0.17239407301851922, | |
| "learning_rate": 2.399033482913359e-06, | |
| "loss": 0.2758, | |
| "num_tokens": 3219537134.0, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.872845831392641, | |
| "grad_norm": 0.160921363209712, | |
| "learning_rate": 2.312737314463238e-06, | |
| "loss": 0.2825, | |
| "num_tokens": 3224780014.0, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 2.8775034932463903, | |
| "grad_norm": 0.16796360529759863, | |
| "learning_rate": 2.226441146013117e-06, | |
| "loss": 0.2892, | |
| "num_tokens": 3230001068.0, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.88216115510014, | |
| "grad_norm": 0.1615021086178067, | |
| "learning_rate": 2.140144977562996e-06, | |
| "loss": 0.2773, | |
| "num_tokens": 3235243948.0, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 2.8868188169538893, | |
| "grad_norm": 0.16227746520671288, | |
| "learning_rate": 2.0538488091128757e-06, | |
| "loss": 0.2839, | |
| "num_tokens": 3240486828.0, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.891476478807639, | |
| "grad_norm": 0.1711645094393323, | |
| "learning_rate": 1.9675526406627547e-06, | |
| "loss": 0.2848, | |
| "num_tokens": 3245729708.0, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 2.896134140661388, | |
| "grad_norm": 0.16190359991778525, | |
| "learning_rate": 1.8812564722126338e-06, | |
| "loss": 0.2769, | |
| "num_tokens": 3250972588.0, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 2.9007918025151374, | |
| "grad_norm": 0.16767984196106225, | |
| "learning_rate": 1.794960303762513e-06, | |
| "loss": 0.2747, | |
| "num_tokens": 3256215468.0, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 2.905449464368887, | |
| "grad_norm": 0.17187093592071814, | |
| "learning_rate": 1.7086641353123924e-06, | |
| "loss": 0.2834, | |
| "num_tokens": 3261418642.0, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.9101071262226363, | |
| "grad_norm": 0.16593307567809112, | |
| "learning_rate": 1.6223679668622715e-06, | |
| "loss": 0.285, | |
| "num_tokens": 3266596154.0, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 2.9147647880763854, | |
| "grad_norm": 0.16936788209066822, | |
| "learning_rate": 1.5360717984121505e-06, | |
| "loss": 0.2729, | |
| "num_tokens": 3271821252.0, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 2.919422449930135, | |
| "grad_norm": 0.16199781431823276, | |
| "learning_rate": 1.4497756299620296e-06, | |
| "loss": 0.281, | |
| "num_tokens": 3277064132.0, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 2.9240801117838844, | |
| "grad_norm": 0.1599415617676873, | |
| "learning_rate": 1.363479461511909e-06, | |
| "loss": 0.2879, | |
| "num_tokens": 3282307012.0, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.928737773637634, | |
| "grad_norm": 0.17169051338273975, | |
| "learning_rate": 1.277183293061788e-06, | |
| "loss": 0.2926, | |
| "num_tokens": 3287546350.0, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 2.9333954354913834, | |
| "grad_norm": 0.16976958060720237, | |
| "learning_rate": 1.1908871246116673e-06, | |
| "loss": 0.2844, | |
| "num_tokens": 3292761750.0, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.938053097345133, | |
| "grad_norm": 0.21762149953130888, | |
| "learning_rate": 1.1045909561615463e-06, | |
| "loss": 0.281, | |
| "num_tokens": 3297949422.0, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 2.9427107591988824, | |
| "grad_norm": 0.16692672814770282, | |
| "learning_rate": 1.0182947877114256e-06, | |
| "loss": 0.2841, | |
| "num_tokens": 3303192302.0, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.1703069504653585, | |
| "learning_rate": 9.319986192613048e-07, | |
| "loss": 0.2861, | |
| "num_tokens": 3308435182.0, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 2.952026082906381, | |
| "grad_norm": 0.1585555018764834, | |
| "learning_rate": 8.45702450811184e-07, | |
| "loss": 0.2808, | |
| "num_tokens": 3313678062.0, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 2.9566837447601304, | |
| "grad_norm": 0.1698111811621913, | |
| "learning_rate": 7.594062823610632e-07, | |
| "loss": 0.2852, | |
| "num_tokens": 3318920942.0, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 2.96134140661388, | |
| "grad_norm": 0.1620121265740979, | |
| "learning_rate": 6.731101139109423e-07, | |
| "loss": 0.2803, | |
| "num_tokens": 3324163822.0, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.9659990684676294, | |
| "grad_norm": 0.16593396203533156, | |
| "learning_rate": 5.868139454608215e-07, | |
| "loss": 0.282, | |
| "num_tokens": 3329404888.0, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 2.9706567303213784, | |
| "grad_norm": 0.1661780559267366, | |
| "learning_rate": 5.005177770107007e-07, | |
| "loss": 0.2744, | |
| "num_tokens": 3334635936.0, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.975314392175128, | |
| "grad_norm": 0.17134864073976327, | |
| "learning_rate": 4.142216085605799e-07, | |
| "loss": 0.2866, | |
| "num_tokens": 3339878816.0, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 2.9799720540288774, | |
| "grad_norm": 0.16516199160895237, | |
| "learning_rate": 3.279254401104591e-07, | |
| "loss": 0.286, | |
| "num_tokens": 3345029172.0, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.984629715882627, | |
| "grad_norm": 0.1631031999805717, | |
| "learning_rate": 2.416292716603383e-07, | |
| "loss": 0.2862, | |
| "num_tokens": 3350272052.0, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 2.9892873777363764, | |
| "grad_norm": 0.16329840168944437, | |
| "learning_rate": 1.5533310321021747e-07, | |
| "loss": 0.2703, | |
| "num_tokens": 3355456072.0, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.993945039590126, | |
| "grad_norm": 0.16826145904096465, | |
| "learning_rate": 6.903693476009665e-08, | |
| "loss": 0.2725, | |
| "num_tokens": 3360698952.0, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 2.9976711690731253, | |
| "num_tokens": 3364893256.0, | |
| "step": 3219, | |
| "total_flos": 2.754977641940386e+18, | |
| "train_loss": 0.41339050008851425, | |
| "train_runtime": 31033.4987, | |
| "train_samples_per_second": 1.66, | |
| "train_steps_per_second": 0.104 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 3219, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.754977641940386e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |