| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9992217898832685, | |
| "eval_steps": 500, | |
| "global_step": 963, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010376134889753567, | |
| "grad_norm": 23.969658904297845, | |
| "learning_rate": 1.0309278350515465e-07, | |
| "loss": 1.3725, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005188067444876783, | |
| "grad_norm": 22.449071975220065, | |
| "learning_rate": 5.154639175257732e-07, | |
| "loss": 1.3719, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010376134889753566, | |
| "grad_norm": 8.548339530816044, | |
| "learning_rate": 1.0309278350515464e-06, | |
| "loss": 1.2558, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01556420233463035, | |
| "grad_norm": 8.215121326664015, | |
| "learning_rate": 1.5463917525773197e-06, | |
| "loss": 1.081, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.020752269779507133, | |
| "grad_norm": 3.0841636846785283, | |
| "learning_rate": 2.061855670103093e-06, | |
| "loss": 0.9504, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02594033722438392, | |
| "grad_norm": 2.3540131831672575, | |
| "learning_rate": 2.577319587628866e-06, | |
| "loss": 0.9092, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0311284046692607, | |
| "grad_norm": 2.217217553056043, | |
| "learning_rate": 3.0927835051546395e-06, | |
| "loss": 0.8692, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03631647211413749, | |
| "grad_norm": 2.2563290398206615, | |
| "learning_rate": 3.6082474226804126e-06, | |
| "loss": 0.8445, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.041504539559014265, | |
| "grad_norm": 2.2874485501473907, | |
| "learning_rate": 4.123711340206186e-06, | |
| "loss": 0.8336, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04669260700389105, | |
| "grad_norm": 2.3330889165134967, | |
| "learning_rate": 4.639175257731959e-06, | |
| "loss": 0.8218, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05188067444876784, | |
| "grad_norm": 2.16633956379982, | |
| "learning_rate": 5.154639175257732e-06, | |
| "loss": 0.8255, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.057068741893644616, | |
| "grad_norm": 2.355679025636223, | |
| "learning_rate": 5.670103092783505e-06, | |
| "loss": 0.7891, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0622568093385214, | |
| "grad_norm": 2.4839432117685036, | |
| "learning_rate": 6.185567010309279e-06, | |
| "loss": 0.7814, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06744487678339818, | |
| "grad_norm": 2.480013950899919, | |
| "learning_rate": 6.701030927835052e-06, | |
| "loss": 0.7674, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07263294422827497, | |
| "grad_norm": 2.3378408946103284, | |
| "learning_rate": 7.216494845360825e-06, | |
| "loss": 0.766, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07782101167315175, | |
| "grad_norm": 2.2751461973205482, | |
| "learning_rate": 7.731958762886599e-06, | |
| "loss": 0.7433, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08300907911802853, | |
| "grad_norm": 2.428511002623931, | |
| "learning_rate": 8.247422680412371e-06, | |
| "loss": 0.7414, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08819714656290532, | |
| "grad_norm": 2.4298836600324045, | |
| "learning_rate": 8.762886597938146e-06, | |
| "loss": 0.7358, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0933852140077821, | |
| "grad_norm": 2.41313947506571, | |
| "learning_rate": 9.278350515463918e-06, | |
| "loss": 0.7319, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09857328145265888, | |
| "grad_norm": 2.4451429150679274, | |
| "learning_rate": 9.793814432989691e-06, | |
| "loss": 0.7323, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10376134889753567, | |
| "grad_norm": 2.450193589248992, | |
| "learning_rate": 9.999703897419048e-06, | |
| "loss": 0.7231, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10894941634241245, | |
| "grad_norm": 2.271786014084883, | |
| "learning_rate": 9.997894508649995e-06, | |
| "loss": 0.7149, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11413748378728923, | |
| "grad_norm": 2.354564055245926, | |
| "learning_rate": 9.99444082710777e-06, | |
| "loss": 0.708, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11932555123216602, | |
| "grad_norm": 2.220428698962425, | |
| "learning_rate": 9.989343989043563e-06, | |
| "loss": 0.7216, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1245136186770428, | |
| "grad_norm": 2.3141712328751396, | |
| "learning_rate": 9.982605671302293e-06, | |
| "loss": 0.7091, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1297016861219196, | |
| "grad_norm": 2.100396054783955, | |
| "learning_rate": 9.97422809077092e-06, | |
| "loss": 0.7066, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13488975356679636, | |
| "grad_norm": 2.2484885982675413, | |
| "learning_rate": 9.9642140036491e-06, | |
| "loss": 0.7085, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14007782101167315, | |
| "grad_norm": 2.1795476193729413, | |
| "learning_rate": 9.9525667045424e-06, | |
| "loss": 0.6889, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.14526588845654995, | |
| "grad_norm": 2.1757051871338593, | |
| "learning_rate": 9.93929002537839e-06, | |
| "loss": 0.6921, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1504539559014267, | |
| "grad_norm": 2.143005235580036, | |
| "learning_rate": 9.924388334145943e-06, | |
| "loss": 0.6907, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1556420233463035, | |
| "grad_norm": 2.1989760690420157, | |
| "learning_rate": 9.90786653345818e-06, | |
| "loss": 0.6912, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1608300907911803, | |
| "grad_norm": 2.004571277860471, | |
| "learning_rate": 9.889730058939529e-06, | |
| "loss": 0.6859, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.16601815823605706, | |
| "grad_norm": 2.05691987455993, | |
| "learning_rate": 9.869984877437413e-06, | |
| "loss": 0.6894, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17120622568093385, | |
| "grad_norm": 2.230053895792029, | |
| "learning_rate": 9.848637485059183e-06, | |
| "loss": 0.6814, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.17639429312581065, | |
| "grad_norm": 1.9493958638517837, | |
| "learning_rate": 9.82569490503491e-06, | |
| "loss": 0.6731, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1815823605706874, | |
| "grad_norm": 2.133120594361784, | |
| "learning_rate": 9.80116468540677e-06, | |
| "loss": 0.6594, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1867704280155642, | |
| "grad_norm": 2.01624934264464, | |
| "learning_rate": 9.775054896545755e-06, | |
| "loss": 0.6751, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.191958495460441, | |
| "grad_norm": 2.1502691215852527, | |
| "learning_rate": 9.747374128496541e-06, | |
| "loss": 0.6789, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.19714656290531776, | |
| "grad_norm": 2.0484102083185194, | |
| "learning_rate": 9.718131488151399e-06, | |
| "loss": 0.6676, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20233463035019456, | |
| "grad_norm": 2.0715841424222337, | |
| "learning_rate": 9.687336596254045e-06, | |
| "loss": 0.6616, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.20752269779507135, | |
| "grad_norm": 2.012157328183036, | |
| "learning_rate": 9.654999584234444e-06, | |
| "loss": 0.652, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2127107652399481, | |
| "grad_norm": 2.0669739212271923, | |
| "learning_rate": 9.621131090875603e-06, | |
| "loss": 0.6426, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2178988326848249, | |
| "grad_norm": 2.0105636015375143, | |
| "learning_rate": 9.585742258813447e-06, | |
| "loss": 0.6445, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2230869001297017, | |
| "grad_norm": 2.1108266544110688, | |
| "learning_rate": 9.548844730870903e-06, | |
| "loss": 0.6438, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.22827496757457846, | |
| "grad_norm": 2.072355913378756, | |
| "learning_rate": 9.51045064622747e-06, | |
| "loss": 0.6565, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23346303501945526, | |
| "grad_norm": 2.166007360772802, | |
| "learning_rate": 9.470572636425451e-06, | |
| "loss": 0.647, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.23865110246433205, | |
| "grad_norm": 2.022875957881762, | |
| "learning_rate": 9.429223821214213e-06, | |
| "loss": 0.6325, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2438391699092088, | |
| "grad_norm": 2.006861087987301, | |
| "learning_rate": 9.386417804233836e-06, | |
| "loss": 0.6477, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2490272373540856, | |
| "grad_norm": 2.0140489204477645, | |
| "learning_rate": 9.34216866853954e-06, | |
| "loss": 0.6391, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.25421530479896237, | |
| "grad_norm": 1.9489606047213677, | |
| "learning_rate": 9.296490971968416e-06, | |
| "loss": 0.6283, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2594033722438392, | |
| "grad_norm": 2.072486707132733, | |
| "learning_rate": 9.249399742349928e-06, | |
| "loss": 0.6377, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.26459143968871596, | |
| "grad_norm": 1.9650189580925839, | |
| "learning_rate": 9.20091047256181e-06, | |
| "loss": 0.6261, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2697795071335927, | |
| "grad_norm": 1.9241991797476943, | |
| "learning_rate": 9.151039115432946e-06, | |
| "loss": 0.6184, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27496757457846954, | |
| "grad_norm": 1.9743470888532664, | |
| "learning_rate": 9.099802078494947e-06, | |
| "loss": 0.6142, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2801556420233463, | |
| "grad_norm": 2.160988187935936, | |
| "learning_rate": 9.047216218584105e-06, | |
| "loss": 0.6094, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2853437094682231, | |
| "grad_norm": 1.9697508480614465, | |
| "learning_rate": 8.993298836295556e-06, | |
| "loss": 0.6196, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2905317769130999, | |
| "grad_norm": 1.8771524751425768, | |
| "learning_rate": 8.93806767029143e-06, | |
| "loss": 0.6163, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.29571984435797666, | |
| "grad_norm": 2.125863779805947, | |
| "learning_rate": 8.88154089146488e-06, | |
| "loss": 0.6167, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3009079118028534, | |
| "grad_norm": 2.1188493077731514, | |
| "learning_rate": 8.823737096961916e-06, | |
| "loss": 0.5992, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.30609597924773024, | |
| "grad_norm": 2.1335267497592807, | |
| "learning_rate": 8.764675304062992e-06, | |
| "loss": 0.6071, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.311284046692607, | |
| "grad_norm": 2.036189297244598, | |
| "learning_rate": 8.704374943926386e-06, | |
| "loss": 0.609, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3164721141374838, | |
| "grad_norm": 1.915927299304865, | |
| "learning_rate": 8.642855855195394e-06, | |
| "loss": 0.5945, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3216601815823606, | |
| "grad_norm": 2.005194485630929, | |
| "learning_rate": 8.580138277471476e-06, | |
| "loss": 0.5959, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.32684824902723736, | |
| "grad_norm": 2.1368034472887527, | |
| "learning_rate": 8.516242844655498e-06, | |
| "loss": 0.5941, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3320363164721141, | |
| "grad_norm": 1.9360804934529585, | |
| "learning_rate": 8.45119057815922e-06, | |
| "loss": 0.5915, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.33722438391699094, | |
| "grad_norm": 1.9356101875463727, | |
| "learning_rate": 8.385002879989328e-06, | |
| "loss": 0.5838, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3424124513618677, | |
| "grad_norm": 2.4311425501079023, | |
| "learning_rate": 8.317701525706226e-06, | |
| "loss": 0.5946, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3476005188067445, | |
| "grad_norm": 2.356263841306792, | |
| "learning_rate": 8.249308657259943e-06, | |
| "loss": 0.567, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3527885862516213, | |
| "grad_norm": 2.048334150791661, | |
| "learning_rate": 8.179846775705504e-06, | |
| "loss": 0.5795, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.35797665369649806, | |
| "grad_norm": 1.9977511587812506, | |
| "learning_rate": 8.109338733800132e-06, | |
| "loss": 0.5751, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3631647211413748, | |
| "grad_norm": 1.8688618314869894, | |
| "learning_rate": 8.03780772848477e-06, | |
| "loss": 0.568, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.36835278858625164, | |
| "grad_norm": 1.93022130905715, | |
| "learning_rate": 7.965277293252354e-06, | |
| "loss": 0.5682, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3735408560311284, | |
| "grad_norm": 2.0382225242835528, | |
| "learning_rate": 7.891771290405351e-06, | |
| "loss": 0.5617, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3787289234760052, | |
| "grad_norm": 1.9924209327442368, | |
| "learning_rate": 7.817313903205148e-06, | |
| "loss": 0.5577, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.383916990920882, | |
| "grad_norm": 1.9678458173326334, | |
| "learning_rate": 7.741929627915814e-06, | |
| "loss": 0.56, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "grad_norm": 2.2405618654805215, | |
| "learning_rate": 7.66564326574491e-06, | |
| "loss": 0.5513, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3942931258106355, | |
| "grad_norm": 1.9971872990885233, | |
| "learning_rate": 7.588479914683954e-06, | |
| "loss": 0.5445, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.39948119325551235, | |
| "grad_norm": 2.06807252227761, | |
| "learning_rate": 7.510464961251271e-06, | |
| "loss": 0.5674, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4046692607003891, | |
| "grad_norm": 1.9627368535332135, | |
| "learning_rate": 7.431624072139884e-06, | |
| "loss": 0.5435, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4098573281452659, | |
| "grad_norm": 1.9716804464407136, | |
| "learning_rate": 7.351983185773259e-06, | |
| "loss": 0.5552, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4150453955901427, | |
| "grad_norm": 1.9693396583392846, | |
| "learning_rate": 7.271568503771632e-06, | |
| "loss": 0.5343, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42023346303501946, | |
| "grad_norm": 1.9432949161104107, | |
| "learning_rate": 7.190406482331757e-06, | |
| "loss": 0.5475, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4254215304798962, | |
| "grad_norm": 2.0194917717314045, | |
| "learning_rate": 7.108523823522891e-06, | |
| "loss": 0.5477, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.43060959792477305, | |
| "grad_norm": 2.206404974952941, | |
| "learning_rate": 7.0259474665018915e-06, | |
| "loss": 0.5425, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4357976653696498, | |
| "grad_norm": 1.9526533277899327, | |
| "learning_rate": 6.942704578650312e-06, | |
| "loss": 0.5161, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4409857328145266, | |
| "grad_norm": 2.0097466124913117, | |
| "learning_rate": 6.858822546636417e-06, | |
| "loss": 0.5331, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4461738002594034, | |
| "grad_norm": 1.8348649689633039, | |
| "learning_rate": 6.774328967405035e-06, | |
| "loss": 0.523, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.45136186770428016, | |
| "grad_norm": 2.139084532722164, | |
| "learning_rate": 6.689251639098261e-06, | |
| "loss": 0.5251, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4565499351491569, | |
| "grad_norm": 1.9708479629081865, | |
| "learning_rate": 6.603618551909935e-06, | |
| "loss": 0.5232, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.46173800259403375, | |
| "grad_norm": 1.9331722289768318, | |
| "learning_rate": 6.517457878876958e-06, | |
| "loss": 0.5305, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4669260700389105, | |
| "grad_norm": 1.859009250403284, | |
| "learning_rate": 6.430797966610436e-06, | |
| "loss": 0.5159, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4721141374837873, | |
| "grad_norm": 1.986527066309499, | |
| "learning_rate": 6.343667325969736e-06, | |
| "loss": 0.5367, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4773022049286641, | |
| "grad_norm": 1.9771277544299588, | |
| "learning_rate": 6.256094622682493e-06, | |
| "loss": 0.5123, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.48249027237354086, | |
| "grad_norm": 2.0022259730400904, | |
| "learning_rate": 6.168108667913666e-06, | |
| "loss": 0.5166, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4876783398184176, | |
| "grad_norm": 1.9991961519932744, | |
| "learning_rate": 6.079738408786753e-06, | |
| "loss": 0.5161, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.49286640726329445, | |
| "grad_norm": 2.0805595238898307, | |
| "learning_rate": 5.9910129188602665e-06, | |
| "loss": 0.5179, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4980544747081712, | |
| "grad_norm": 1.929253006230254, | |
| "learning_rate": 5.9019613885626235e-06, | |
| "loss": 0.5097, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.503242542153048, | |
| "grad_norm": 2.25129632838715, | |
| "learning_rate": 5.812613115588575e-06, | |
| "loss": 0.4971, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5084306095979247, | |
| "grad_norm": 1.9119339241166262, | |
| "learning_rate": 5.722997495260348e-06, | |
| "loss": 0.4988, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5136186770428015, | |
| "grad_norm": 1.8300200112998326, | |
| "learning_rate": 5.6331440108566735e-06, | |
| "loss": 0.4941, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5188067444876784, | |
| "grad_norm": 1.9591247994452368, | |
| "learning_rate": 5.543082223912875e-06, | |
| "loss": 0.492, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5239948119325551, | |
| "grad_norm": 1.99136453982626, | |
| "learning_rate": 5.452841764495203e-06, | |
| "loss": 0.5002, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5291828793774319, | |
| "grad_norm": 1.9961024804052654, | |
| "learning_rate": 5.362452321452636e-06, | |
| "loss": 0.4772, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5343709468223087, | |
| "grad_norm": 1.9607124098040063, | |
| "learning_rate": 5.2719436326493255e-06, | |
| "loss": 0.4908, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5395590142671854, | |
| "grad_norm": 1.9303906010446525, | |
| "learning_rate": 5.181345475180941e-06, | |
| "loss": 0.4866, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5447470817120622, | |
| "grad_norm": 2.0420688559734503, | |
| "learning_rate": 5.090687655578078e-06, | |
| "loss": 0.4769, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5499351491569391, | |
| "grad_norm": 1.9908642175713687, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4742, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5551232166018158, | |
| "grad_norm": 1.9960779934532675, | |
| "learning_rate": 4.909312344421923e-06, | |
| "loss": 0.4666, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5603112840466926, | |
| "grad_norm": 1.9274839933909422, | |
| "learning_rate": 4.8186545248190604e-06, | |
| "loss": 0.4866, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5654993514915694, | |
| "grad_norm": 1.9162466337096817, | |
| "learning_rate": 4.7280563673506745e-06, | |
| "loss": 0.4692, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5706874189364461, | |
| "grad_norm": 2.07386431606307, | |
| "learning_rate": 4.637547678547366e-06, | |
| "loss": 0.4859, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5758754863813229, | |
| "grad_norm": 2.0201984812958385, | |
| "learning_rate": 4.547158235504797e-06, | |
| "loss": 0.4718, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5810635538261998, | |
| "grad_norm": 1.95015272613481, | |
| "learning_rate": 4.4569177760871255e-06, | |
| "loss": 0.475, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5862516212710766, | |
| "grad_norm": 1.944586565605588, | |
| "learning_rate": 4.366855989143326e-06, | |
| "loss": 0.4551, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5914396887159533, | |
| "grad_norm": 1.9208589567145171, | |
| "learning_rate": 4.277002504739653e-06, | |
| "loss": 0.4686, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5966277561608301, | |
| "grad_norm": 1.8639671285460482, | |
| "learning_rate": 4.187386884411426e-06, | |
| "loss": 0.4557, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6018158236057068, | |
| "grad_norm": 1.9975578797091653, | |
| "learning_rate": 4.098038611437377e-06, | |
| "loss": 0.4651, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6070038910505836, | |
| "grad_norm": 1.961651938542185, | |
| "learning_rate": 4.008987081139734e-06, | |
| "loss": 0.4643, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6121919584954605, | |
| "grad_norm": 1.9374158302120401, | |
| "learning_rate": 3.920261591213249e-06, | |
| "loss": 0.4556, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6173800259403373, | |
| "grad_norm": 1.9090835435895448, | |
| "learning_rate": 3.8318913320863355e-06, | |
| "loss": 0.4536, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.622568093385214, | |
| "grad_norm": 1.8975263865890188, | |
| "learning_rate": 3.7439053773175092e-06, | |
| "loss": 0.4615, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6277561608300908, | |
| "grad_norm": 1.9060390294655216, | |
| "learning_rate": 3.6563326740302664e-06, | |
| "loss": 0.4459, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6329442282749675, | |
| "grad_norm": 1.9725006931962796, | |
| "learning_rate": 3.569202033389565e-06, | |
| "loss": 0.4451, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6381322957198443, | |
| "grad_norm": 1.9621067476956515, | |
| "learning_rate": 3.4825421211230437e-06, | |
| "loss": 0.4419, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6433203631647212, | |
| "grad_norm": 2.098443239659209, | |
| "learning_rate": 3.3963814480900665e-06, | |
| "loss": 0.4415, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.648508430609598, | |
| "grad_norm": 1.8981208726840302, | |
| "learning_rate": 3.310748360901741e-06, | |
| "loss": 0.4456, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6536964980544747, | |
| "grad_norm": 1.8947168989269416, | |
| "learning_rate": 3.225671032594966e-06, | |
| "loss": 0.4229, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6588845654993515, | |
| "grad_norm": 2.0138652650509288, | |
| "learning_rate": 3.1411774533635854e-06, | |
| "loss": 0.437, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6640726329442282, | |
| "grad_norm": 1.8903378440015823, | |
| "learning_rate": 3.0572954213496897e-06, | |
| "loss": 0.4454, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.669260700389105, | |
| "grad_norm": 1.8448484960177367, | |
| "learning_rate": 2.9740525334981105e-06, | |
| "loss": 0.4398, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6744487678339819, | |
| "grad_norm": 1.9976530631786225, | |
| "learning_rate": 2.8914761764771093e-06, | |
| "loss": 0.429, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6796368352788587, | |
| "grad_norm": 1.9155018572353837, | |
| "learning_rate": 2.809593517668243e-06, | |
| "loss": 0.4309, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6848249027237354, | |
| "grad_norm": 1.942714148946629, | |
| "learning_rate": 2.728431496228369e-06, | |
| "loss": 0.4248, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6900129701686122, | |
| "grad_norm": 2.013023734418392, | |
| "learning_rate": 2.648016814226742e-06, | |
| "loss": 0.4304, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.695201037613489, | |
| "grad_norm": 1.9023117871214554, | |
| "learning_rate": 2.5683759278601174e-06, | |
| "loss": 0.4338, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7003891050583657, | |
| "grad_norm": 1.8911448184302957, | |
| "learning_rate": 2.4895350387487304e-06, | |
| "loss": 0.4245, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7055771725032426, | |
| "grad_norm": 2.0358392917626813, | |
| "learning_rate": 2.4115200853160475e-06, | |
| "loss": 0.4194, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7107652399481194, | |
| "grad_norm": 1.9510576677492195, | |
| "learning_rate": 2.3343567342550933e-06, | |
| "loss": 0.4267, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7159533073929961, | |
| "grad_norm": 1.8690267408594539, | |
| "learning_rate": 2.258070372084188e-06, | |
| "loss": 0.4312, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7211413748378729, | |
| "grad_norm": 1.8322122073891454, | |
| "learning_rate": 2.182686096794852e-06, | |
| "loss": 0.4207, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7263294422827496, | |
| "grad_norm": 2.0311002524177253, | |
| "learning_rate": 2.108228709594649e-06, | |
| "loss": 0.4227, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7315175097276264, | |
| "grad_norm": 1.8678394687630775, | |
| "learning_rate": 2.0347227067476478e-06, | |
| "loss": 0.4149, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7367055771725033, | |
| "grad_norm": 1.8521301731665931, | |
| "learning_rate": 1.962192271515232e-06, | |
| "loss": 0.4192, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.74189364461738, | |
| "grad_norm": 1.9291143236144128, | |
| "learning_rate": 1.8906612661998698e-06, | |
| "loss": 0.4128, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7470817120622568, | |
| "grad_norm": 1.991134829662921, | |
| "learning_rate": 1.820153224294498e-06, | |
| "loss": 0.4102, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7522697795071336, | |
| "grad_norm": 1.8597303553848081, | |
| "learning_rate": 1.750691342740058e-06, | |
| "loss": 0.4104, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7574578469520103, | |
| "grad_norm": 1.8334844899907363, | |
| "learning_rate": 1.6822984742937764e-06, | |
| "loss": 0.4049, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7626459143968871, | |
| "grad_norm": 1.839241095874111, | |
| "learning_rate": 1.6149971200106723e-06, | |
| "loss": 0.4009, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.767833981841764, | |
| "grad_norm": 1.8614288971061537, | |
| "learning_rate": 1.548809421840779e-06, | |
| "loss": 0.4029, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7730220492866408, | |
| "grad_norm": 1.9712640153496117, | |
| "learning_rate": 1.483757155344503e-06, | |
| "loss": 0.4056, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "grad_norm": 1.7973975593361922, | |
| "learning_rate": 1.4198617225285244e-06, | |
| "loss": 0.409, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7833981841763943, | |
| "grad_norm": 1.8870139707940816, | |
| "learning_rate": 1.3571441448046086e-06, | |
| "loss": 0.4117, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.788586251621271, | |
| "grad_norm": 1.9255920717839368, | |
| "learning_rate": 1.2956250560736143e-06, | |
| "loss": 0.4097, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7937743190661478, | |
| "grad_norm": 1.8605570734597534, | |
| "learning_rate": 1.2353246959370086e-06, | |
| "loss": 0.3885, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7989623865110247, | |
| "grad_norm": 1.8678208345700735, | |
| "learning_rate": 1.1762629030380867e-06, | |
| "loss": 0.4044, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8041504539559015, | |
| "grad_norm": 1.8406724398818959, | |
| "learning_rate": 1.118459108535122e-06, | |
| "loss": 0.3991, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8093385214007782, | |
| "grad_norm": 1.9447178497450672, | |
| "learning_rate": 1.061932329708572e-06, | |
| "loss": 0.3878, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.814526588845655, | |
| "grad_norm": 1.903347183666585, | |
| "learning_rate": 1.006701163704445e-06, | |
| "loss": 0.3994, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8197146562905318, | |
| "grad_norm": 1.7817648380438804, | |
| "learning_rate": 9.527837814158963e-07, | |
| "loss": 0.3943, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8249027237354085, | |
| "grad_norm": 1.89718875917406, | |
| "learning_rate": 9.001979215050544e-07, | |
| "loss": 0.3929, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8300907911802854, | |
| "grad_norm": 1.8825895138353903, | |
| "learning_rate": 8.489608845670527e-07, | |
| "loss": 0.3924, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8352788586251622, | |
| "grad_norm": 1.7908515887362904, | |
| "learning_rate": 7.99089527438191e-07, | |
| "loss": 0.3919, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8404669260700389, | |
| "grad_norm": 2.0314129121613034, | |
| "learning_rate": 7.506002576500732e-07, | |
| "loss": 0.3941, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8456549935149157, | |
| "grad_norm": 1.838737045068825, | |
| "learning_rate": 7.035090280315854e-07, | |
| "loss": 0.398, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8508430609597925, | |
| "grad_norm": 1.8701463175206698, | |
| "learning_rate": 6.578313314604612e-07, | |
| "loss": 0.395, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8560311284046692, | |
| "grad_norm": 1.8734176088672492, | |
| "learning_rate": 6.135821957661658e-07, | |
| "loss": 0.3945, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8612191958495461, | |
| "grad_norm": 1.8454123160341045, | |
| "learning_rate": 5.707761787857879e-07, | |
| "loss": 0.3855, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8664072632944229, | |
| "grad_norm": 1.7755466173110739, | |
| "learning_rate": 5.294273635745517e-07, | |
| "loss": 0.3971, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8715953307392996, | |
| "grad_norm": 1.8394864397787671, | |
| "learning_rate": 4.895493537725326e-07, | |
| "loss": 0.3966, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8767833981841764, | |
| "grad_norm": 1.7915928948304078, | |
| "learning_rate": 4.511552691290988e-07, | |
| "loss": 0.3979, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.8819714656290532, | |
| "grad_norm": 1.7863787006400424, | |
| "learning_rate": 4.1425774118655505e-07, | |
| "loss": 0.3826, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8871595330739299, | |
| "grad_norm": 1.8024141112662704, | |
| "learning_rate": 3.7886890912439633e-07, | |
| "loss": 0.3862, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8923476005188068, | |
| "grad_norm": 1.8180972720099156, | |
| "learning_rate": 3.4500041576555733e-07, | |
| "loss": 0.3859, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8975356679636836, | |
| "grad_norm": 1.7622051589037506, | |
| "learning_rate": 3.1266340374595693e-07, | |
| "loss": 0.3831, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9027237354085603, | |
| "grad_norm": 1.7926800043760007, | |
| "learning_rate": 2.818685118486025e-07, | |
| "loss": 0.3927, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9079118028534371, | |
| "grad_norm": 1.8515815235983688, | |
| "learning_rate": 2.526258715034602e-07, | |
| "loss": 0.3787, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9130998702983139, | |
| "grad_norm": 1.761786866550431, | |
| "learning_rate": 2.2494510345424657e-07, | |
| "loss": 0.3881, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9182879377431906, | |
| "grad_norm": 1.731506703869926, | |
| "learning_rate": 1.988353145932298e-07, | |
| "loss": 0.3762, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9234760051880675, | |
| "grad_norm": 1.8427166106595052, | |
| "learning_rate": 1.7430509496508985e-07, | |
| "loss": 0.3975, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9286640726329443, | |
| "grad_norm": 1.761769698023775, | |
| "learning_rate": 1.5136251494081822e-07, | |
| "loss": 0.3842, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.933852140077821, | |
| "grad_norm": 1.8297504100937483, | |
| "learning_rate": 1.3001512256258841e-07, | |
| "loss": 0.3916, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9390402075226978, | |
| "grad_norm": 1.8143369848190358, | |
| "learning_rate": 1.1026994106047296e-07, | |
| "loss": 0.3911, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9442282749675746, | |
| "grad_norm": 1.7462314691918333, | |
| "learning_rate": 9.213346654182054e-08, | |
| "loss": 0.3888, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9494163424124513, | |
| "grad_norm": 1.842285372864709, | |
| "learning_rate": 7.561166585405789e-08, | |
| "loss": 0.3823, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9546044098573282, | |
| "grad_norm": 1.798454935332072, | |
| "learning_rate": 6.070997462161055e-08, | |
| "loss": 0.4032, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.959792477302205, | |
| "grad_norm": 1.8579672164577692, | |
| "learning_rate": 4.743329545760122e-08, | |
| "loss": 0.3811, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9649805447470817, | |
| "grad_norm": 1.764976690651984, | |
| "learning_rate": 3.578599635090163e-08, | |
| "loss": 0.3806, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9701686121919585, | |
| "grad_norm": 1.7085373084916373, | |
| "learning_rate": 2.577190922908035e-08, | |
| "loss": 0.3888, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9753566796368353, | |
| "grad_norm": 1.7431684765639506, | |
| "learning_rate": 1.7394328697707407e-08, | |
| "loss": 0.3901, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.980544747081712, | |
| "grad_norm": 1.8495600056895127, | |
| "learning_rate": 1.0656010956437979e-08, | |
| "loss": 0.3918, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9857328145265889, | |
| "grad_norm": 1.8616847493274582, | |
| "learning_rate": 5.5591728922316235e-09, | |
| "loss": 0.3895, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9909208819714657, | |
| "grad_norm": 1.8274058784400706, | |
| "learning_rate": 2.1054913500051512e-09, | |
| "loss": 0.3831, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.9961089494163424, | |
| "grad_norm": 1.7888916632814764, | |
| "learning_rate": 2.9610258095169596e-10, | |
| "loss": 0.3863, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9992217898832685, | |
| "eval_loss": 0.35284245014190674, | |
| "eval_runtime": 0.9437, | |
| "eval_samples_per_second": 2.119, | |
| "eval_steps_per_second": 1.06, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.9992217898832685, | |
| "step": 963, | |
| "total_flos": 201580263505920.0, | |
| "train_loss": 0.5411187405403034, | |
| "train_runtime": 23935.6127, | |
| "train_samples_per_second": 1.288, | |
| "train_steps_per_second": 0.04 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 963, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 201580263505920.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |