| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999360204734485, | |
| "eval_steps": 500, | |
| "global_step": 781, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012795905310300703, | |
| "grad_norm": 6.290177490440239, | |
| "learning_rate": 2.5316455696202533e-07, | |
| "loss": 1.0784, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006397952655150352, | |
| "grad_norm": 6.702182800413163, | |
| "learning_rate": 1.2658227848101267e-06, | |
| "loss": 1.0415, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.012795905310300703, | |
| "grad_norm": 5.487859701594733, | |
| "learning_rate": 2.5316455696202535e-06, | |
| "loss": 1.0444, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.019193857965451054, | |
| "grad_norm": 2.2091656111344116, | |
| "learning_rate": 3.7974683544303802e-06, | |
| "loss": 0.9625, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.025591810620601407, | |
| "grad_norm": 1.3024745363944248, | |
| "learning_rate": 5.063291139240507e-06, | |
| "loss": 0.8519, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03198976327575176, | |
| "grad_norm": 1.0696576308505852, | |
| "learning_rate": 6.329113924050634e-06, | |
| "loss": 0.8425, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03838771593090211, | |
| "grad_norm": 0.9289972759140979, | |
| "learning_rate": 7.5949367088607605e-06, | |
| "loss": 0.8055, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.044785668586052464, | |
| "grad_norm": 0.8881490776255759, | |
| "learning_rate": 8.860759493670886e-06, | |
| "loss": 0.8101, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05118362124120281, | |
| "grad_norm": 1.003105745392554, | |
| "learning_rate": 1.0126582278481014e-05, | |
| "loss": 0.8034, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05758157389635317, | |
| "grad_norm": 0.9061421007764652, | |
| "learning_rate": 1.139240506329114e-05, | |
| "loss": 0.8149, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06397952655150352, | |
| "grad_norm": 0.947301258001698, | |
| "learning_rate": 1.2658227848101268e-05, | |
| "loss": 0.8066, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07037747920665387, | |
| "grad_norm": 0.9442508051067858, | |
| "learning_rate": 1.3924050632911395e-05, | |
| "loss": 0.7829, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07677543186180422, | |
| "grad_norm": 0.9505498602464607, | |
| "learning_rate": 1.5189873417721521e-05, | |
| "loss": 0.789, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08317338451695458, | |
| "grad_norm": 1.0294440611861784, | |
| "learning_rate": 1.6455696202531647e-05, | |
| "loss": 0.8047, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08957133717210493, | |
| "grad_norm": 0.9037139700901843, | |
| "learning_rate": 1.7721518987341772e-05, | |
| "loss": 0.7808, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09596928982725528, | |
| "grad_norm": 0.9155324723416859, | |
| "learning_rate": 1.89873417721519e-05, | |
| "loss": 0.7693, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.10236724248240563, | |
| "grad_norm": 1.01405799880107, | |
| "learning_rate": 1.999989986294826e-05, | |
| "loss": 0.7701, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10876519513755598, | |
| "grad_norm": 0.8936790580352, | |
| "learning_rate": 1.9996395276708856e-05, | |
| "loss": 0.7559, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11516314779270634, | |
| "grad_norm": 0.9533062095051035, | |
| "learning_rate": 1.9987885843187717e-05, | |
| "loss": 0.7755, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12156110044785669, | |
| "grad_norm": 0.8655719966380863, | |
| "learning_rate": 1.9974375822762117e-05, | |
| "loss": 0.7693, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.12795905310300704, | |
| "grad_norm": 0.9292508995916059, | |
| "learning_rate": 1.9955871979429188e-05, | |
| "loss": 0.7672, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1343570057581574, | |
| "grad_norm": 0.8848494940071496, | |
| "learning_rate": 1.9932383577419432e-05, | |
| "loss": 0.7669, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.14075495841330773, | |
| "grad_norm": 0.8492671125517353, | |
| "learning_rate": 1.9903922376558432e-05, | |
| "loss": 0.7807, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1471529110684581, | |
| "grad_norm": 0.8680755550279518, | |
| "learning_rate": 1.9870502626379127e-05, | |
| "loss": 0.7884, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15355086372360843, | |
| "grad_norm": 0.8715305157008323, | |
| "learning_rate": 1.983214105898757e-05, | |
| "loss": 0.7785, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1599488163787588, | |
| "grad_norm": 0.920271623258983, | |
| "learning_rate": 1.978885688068572e-05, | |
| "loss": 0.7723, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16634676903390916, | |
| "grad_norm": 0.8601122413411659, | |
| "learning_rate": 1.9740671762355548e-05, | |
| "loss": 0.7548, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1727447216890595, | |
| "grad_norm": 0.9508364162043275, | |
| "learning_rate": 1.9687609828609156e-05, | |
| "loss": 0.7782, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.17914267434420986, | |
| "grad_norm": 0.9570098046333952, | |
| "learning_rate": 1.9629697645710432e-05, | |
| "loss": 0.7557, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1855406269993602, | |
| "grad_norm": 0.9889078696060014, | |
| "learning_rate": 1.9566964208274254e-05, | |
| "loss": 0.777, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.19193857965451055, | |
| "grad_norm": 0.914516851971123, | |
| "learning_rate": 1.9499440924749878e-05, | |
| "loss": 0.7703, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19833653230966092, | |
| "grad_norm": 0.9666613857266174, | |
| "learning_rate": 1.9427161601695833e-05, | |
| "loss": 0.7609, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.20473448496481125, | |
| "grad_norm": 0.8774756951541983, | |
| "learning_rate": 1.9350162426854152e-05, | |
| "loss": 0.7475, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21113243761996162, | |
| "grad_norm": 0.9670708841788205, | |
| "learning_rate": 1.926848195103242e-05, | |
| "loss": 0.76, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.21753039027511195, | |
| "grad_norm": 0.8838923277068791, | |
| "learning_rate": 1.9182161068802742e-05, | |
| "loss": 0.7559, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22392834293026231, | |
| "grad_norm": 0.8979119023604246, | |
| "learning_rate": 1.909124299802724e-05, | |
| "loss": 0.7492, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.23032629558541268, | |
| "grad_norm": 0.8267942600640766, | |
| "learning_rate": 1.8995773258220374e-05, | |
| "loss": 0.7498, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.236724248240563, | |
| "grad_norm": 0.8110785643711048, | |
| "learning_rate": 1.8895799647758912e-05, | |
| "loss": 0.7397, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.24312220089571338, | |
| "grad_norm": 0.8742502964910401, | |
| "learning_rate": 1.879137221995095e-05, | |
| "loss": 0.746, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2495201535508637, | |
| "grad_norm": 0.8560236399558296, | |
| "learning_rate": 1.868254325797594e-05, | |
| "loss": 0.7563, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2559181062060141, | |
| "grad_norm": 0.8944327324689623, | |
| "learning_rate": 1.8569367248708343e-05, | |
| "loss": 0.7645, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.26231605886116444, | |
| "grad_norm": 0.8621529624234945, | |
| "learning_rate": 1.845190085543795e-05, | |
| "loss": 0.7552, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2687140115163148, | |
| "grad_norm": 0.8760942401693406, | |
| "learning_rate": 1.8330202889500518e-05, | |
| "loss": 0.7354, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2751119641714651, | |
| "grad_norm": 0.8712951389244179, | |
| "learning_rate": 1.8204334280833005e-05, | |
| "loss": 0.7556, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.28150991682661547, | |
| "grad_norm": 0.9117905123706684, | |
| "learning_rate": 1.807435804746807e-05, | |
| "loss": 0.7674, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.28790786948176583, | |
| "grad_norm": 0.9659529366014488, | |
| "learning_rate": 1.7940339263983112e-05, | |
| "loss": 0.7705, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2943058221369162, | |
| "grad_norm": 0.8392485914175913, | |
| "learning_rate": 1.7802345028919728e-05, | |
| "loss": 0.7522, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.30070377479206656, | |
| "grad_norm": 0.9180964413930104, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.7658, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.30710172744721687, | |
| "grad_norm": 0.8599442472058514, | |
| "learning_rate": 1.7514708515485002e-05, | |
| "loss": 0.7709, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.31349968010236723, | |
| "grad_norm": 0.874820715756307, | |
| "learning_rate": 1.736521024670737e-05, | |
| "loss": 0.7298, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3198976327575176, | |
| "grad_norm": 0.9507611932214313, | |
| "learning_rate": 1.7212024473438145e-05, | |
| "loss": 0.7829, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.32629558541266795, | |
| "grad_norm": 0.8751685918941091, | |
| "learning_rate": 1.705522789046377e-05, | |
| "loss": 0.7519, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3326935380678183, | |
| "grad_norm": 0.8011470314485565, | |
| "learning_rate": 1.6894899000377462e-05, | |
| "loss": 0.74, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3390914907229686, | |
| "grad_norm": 0.8521229042194711, | |
| "learning_rate": 1.67311180742757e-05, | |
| "loss": 0.7543, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.345489443378119, | |
| "grad_norm": 0.8130076328641188, | |
| "learning_rate": 1.65639671115693e-05, | |
| "loss": 0.7339, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.35188739603326935, | |
| "grad_norm": 0.9621951058914813, | |
| "learning_rate": 1.6393529798929103e-05, | |
| "loss": 0.7411, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3582853486884197, | |
| "grad_norm": 0.9254782095065907, | |
| "learning_rate": 1.621989146838704e-05, | |
| "loss": 0.7448, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3646833013435701, | |
| "grad_norm": 0.8675888456425931, | |
| "learning_rate": 1.6043139054613326e-05, | |
| "loss": 0.7701, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3710812539987204, | |
| "grad_norm": 0.8838846353868941, | |
| "learning_rate": 1.586336105139127e-05, | |
| "loss": 0.7365, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.37747920665387075, | |
| "grad_norm": 0.9360739406132822, | |
| "learning_rate": 1.568064746731156e-05, | |
| "loss": 0.7334, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3838771593090211, | |
| "grad_norm": 0.8975040372948366, | |
| "learning_rate": 1.5495089780708062e-05, | |
| "loss": 0.7294, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3902751119641715, | |
| "grad_norm": 0.8204316809700783, | |
| "learning_rate": 1.530678089385782e-05, | |
| "loss": 0.7436, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.39667306461932184, | |
| "grad_norm": 0.8979669622871939, | |
| "learning_rate": 1.5115815086468103e-05, | |
| "loss": 0.7596, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.40307101727447214, | |
| "grad_norm": 0.87770927627433, | |
| "learning_rate": 1.492228796847385e-05, | |
| "loss": 0.7468, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4094689699296225, | |
| "grad_norm": 0.8479491304098338, | |
| "learning_rate": 1.4726296432169095e-05, | |
| "loss": 0.7556, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.41586692258477287, | |
| "grad_norm": 0.8754654732089078, | |
| "learning_rate": 1.4527938603696376e-05, | |
| "loss": 0.7225, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.42226487523992323, | |
| "grad_norm": 0.881596577987191, | |
| "learning_rate": 1.4327313793918362e-05, | |
| "loss": 0.7446, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4286628278950736, | |
| "grad_norm": 0.9352817977484518, | |
| "learning_rate": 1.4124522448696407e-05, | |
| "loss": 0.7216, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4350607805502239, | |
| "grad_norm": 0.8472386929218481, | |
| "learning_rate": 1.3919666098600753e-05, | |
| "loss": 0.7351, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.44145873320537427, | |
| "grad_norm": 0.9626870416040649, | |
| "learning_rate": 1.3712847308077737e-05, | |
| "loss": 0.7335, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.44785668586052463, | |
| "grad_norm": 0.8583773106564518, | |
| "learning_rate": 1.350416962409934e-05, | |
| "loss": 0.7472, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.454254638515675, | |
| "grad_norm": 0.8162278696217269, | |
| "learning_rate": 1.3293737524320798e-05, | |
| "loss": 0.7425, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.46065259117082535, | |
| "grad_norm": 0.8587065087391589, | |
| "learning_rate": 1.3081656364772308e-05, | |
| "loss": 0.721, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.46705054382597566, | |
| "grad_norm": 0.9489656570168764, | |
| "learning_rate": 1.2868032327110904e-05, | |
| "loss": 0.7117, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.473448496481126, | |
| "grad_norm": 0.8866065865142619, | |
| "learning_rate": 1.2652972365459008e-05, | |
| "loss": 0.7154, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4798464491362764, | |
| "grad_norm": 0.9315838159542567, | |
| "learning_rate": 1.243658415285622e-05, | |
| "loss": 0.7684, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.48624440179142675, | |
| "grad_norm": 0.8553566142004, | |
| "learning_rate": 1.2218976027351177e-05, | |
| "loss": 0.7246, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4926423544465771, | |
| "grad_norm": 0.9294553052999376, | |
| "learning_rate": 1.2000256937760446e-05, | |
| "loss": 0.7312, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4990403071017274, | |
| "grad_norm": 0.8632512793370595, | |
| "learning_rate": 1.1780536389121668e-05, | |
| "loss": 0.7414, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5054382597568778, | |
| "grad_norm": 0.8379396676665268, | |
| "learning_rate": 1.155992438786818e-05, | |
| "loss": 0.7264, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5118362124120281, | |
| "grad_norm": 0.9222125293365617, | |
| "learning_rate": 1.1338531386752618e-05, | |
| "loss": 0.7685, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5182341650671785, | |
| "grad_norm": 0.8293885569783266, | |
| "learning_rate": 1.1116468229547079e-05, | |
| "loss": 0.7315, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5246321177223289, | |
| "grad_norm": 0.8181032830246795, | |
| "learning_rate": 1.0893846095547493e-05, | |
| "loss": 0.7124, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5310300703774792, | |
| "grad_norm": 0.8952645487752635, | |
| "learning_rate": 1.0670776443910024e-05, | |
| "loss": 0.7371, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5374280230326296, | |
| "grad_norm": 0.8781812680428416, | |
| "learning_rate": 1.0447370957847343e-05, | |
| "loss": 0.7374, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5438259756877799, | |
| "grad_norm": 0.8047369438911633, | |
| "learning_rate": 1.0223741488712732e-05, | |
| "loss": 0.727, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5502239283429302, | |
| "grad_norm": 0.859407154090335, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7421, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5566218809980806, | |
| "grad_norm": 0.8893479868039521, | |
| "learning_rate": 9.776258511287271e-06, | |
| "loss": 0.732, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.5630198336532309, | |
| "grad_norm": 0.949546721569847, | |
| "learning_rate": 9.55262904215266e-06, | |
| "loss": 0.7332, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5694177863083814, | |
| "grad_norm": 0.8485580508908309, | |
| "learning_rate": 9.329223556089976e-06, | |
| "loss": 0.7412, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5758157389635317, | |
| "grad_norm": 0.8283004065969708, | |
| "learning_rate": 9.10615390445251e-06, | |
| "loss": 0.7211, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.582213691618682, | |
| "grad_norm": 0.8817929492946212, | |
| "learning_rate": 8.883531770452924e-06, | |
| "loss": 0.7341, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.5886116442738324, | |
| "grad_norm": 0.8900670681272643, | |
| "learning_rate": 8.661468613247387e-06, | |
| "loss": 0.7441, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5950095969289827, | |
| "grad_norm": 0.8174845932665962, | |
| "learning_rate": 8.440075612131823e-06, | |
| "loss": 0.7265, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6014075495841331, | |
| "grad_norm": 0.8390159513525518, | |
| "learning_rate": 8.219463610878336e-06, | |
| "loss": 0.7052, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6078055022392834, | |
| "grad_norm": 0.8412199328086044, | |
| "learning_rate": 7.999743062239557e-06, | |
| "loss": 0.7249, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.6142034548944337, | |
| "grad_norm": 0.8369817069040336, | |
| "learning_rate": 7.781023972648826e-06, | |
| "loss": 0.732, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6206014075495841, | |
| "grad_norm": 0.8837012458062626, | |
| "learning_rate": 7.563415847143782e-06, | |
| "loss": 0.7203, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.6269993602047345, | |
| "grad_norm": 0.8540794083066343, | |
| "learning_rate": 7.347027634540993e-06, | |
| "loss": 0.7135, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6333973128598849, | |
| "grad_norm": 0.8525705119650993, | |
| "learning_rate": 7.131967672889101e-06, | |
| "loss": 0.7195, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6397952655150352, | |
| "grad_norm": 0.8786823273169972, | |
| "learning_rate": 6.918343635227694e-06, | |
| "loss": 0.7339, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6461932181701855, | |
| "grad_norm": 1.0560174180349937, | |
| "learning_rate": 6.706262475679205e-06, | |
| "loss": 0.7187, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6525911708253359, | |
| "grad_norm": 0.8749872839659149, | |
| "learning_rate": 6.495830375900665e-06, | |
| "loss": 0.7354, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6589891234804862, | |
| "grad_norm": 0.8972887639537709, | |
| "learning_rate": 6.287152691922264e-06, | |
| "loss": 0.7352, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.6653870761356366, | |
| "grad_norm": 0.758512097349371, | |
| "learning_rate": 6.080333901399252e-06, | |
| "loss": 0.7247, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6717850287907869, | |
| "grad_norm": 0.8424258585380499, | |
| "learning_rate": 5.875477551303596e-06, | |
| "loss": 0.7299, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6781829814459372, | |
| "grad_norm": 0.8587554713063377, | |
| "learning_rate": 5.672686206081638e-06, | |
| "loss": 0.7456, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6845809341010877, | |
| "grad_norm": 0.8220663211697773, | |
| "learning_rate": 5.47206139630363e-06, | |
| "loss": 0.7255, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.690978886756238, | |
| "grad_norm": 0.764989841037643, | |
| "learning_rate": 5.273703567830908e-06, | |
| "loss": 0.6999, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6973768394113884, | |
| "grad_norm": 0.841679038026195, | |
| "learning_rate": 5.077712031526153e-06, | |
| "loss": 0.7216, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.7037747920665387, | |
| "grad_norm": 0.9158140652111489, | |
| "learning_rate": 4.8841849135319015e-06, | |
| "loss": 0.7197, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.710172744721689, | |
| "grad_norm": 0.7720710563951564, | |
| "learning_rate": 4.693219106142186e-06, | |
| "loss": 0.7101, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.7165706973768394, | |
| "grad_norm": 0.9540569720579076, | |
| "learning_rate": 4.504910219291941e-06, | |
| "loss": 0.7292, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7229686500319897, | |
| "grad_norm": 0.8573826120101363, | |
| "learning_rate": 4.319352532688444e-06, | |
| "loss": 0.7105, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.7293666026871402, | |
| "grad_norm": 0.8106292253940656, | |
| "learning_rate": 4.13663894860873e-06, | |
| "loss": 0.7084, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7357645553422905, | |
| "grad_norm": 0.8820590050867799, | |
| "learning_rate": 3.956860945386677e-06, | |
| "loss": 0.713, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.7421625079974408, | |
| "grad_norm": 0.8091830896370955, | |
| "learning_rate": 3.7801085316129615e-06, | |
| "loss": 0.6979, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7485604606525912, | |
| "grad_norm": 0.9215560782786927, | |
| "learning_rate": 3.606470201070904e-06, | |
| "loss": 0.7238, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.7549584133077415, | |
| "grad_norm": 0.8256761378699967, | |
| "learning_rate": 3.4360328884307058e-06, | |
| "loss": 0.7377, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7613563659628919, | |
| "grad_norm": 0.8664264779515329, | |
| "learning_rate": 3.2688819257242963e-06, | |
| "loss": 0.7419, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.7677543186180422, | |
| "grad_norm": 0.8896365508552663, | |
| "learning_rate": 3.1051009996225434e-06, | |
| "loss": 0.7021, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7741522712731925, | |
| "grad_norm": 0.8877160233939523, | |
| "learning_rate": 2.9447721095362325e-06, | |
| "loss": 0.7194, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.780550223928343, | |
| "grad_norm": 0.87887437602247, | |
| "learning_rate": 2.7879755265618558e-06, | |
| "loss": 0.6994, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7869481765834933, | |
| "grad_norm": 0.9105056631876884, | |
| "learning_rate": 2.6347897532926293e-06, | |
| "loss": 0.7178, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.7933461292386437, | |
| "grad_norm": 0.8372924489670353, | |
| "learning_rate": 2.485291484515e-06, | |
| "loss": 0.7033, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.799744081893794, | |
| "grad_norm": 0.8723730239208861, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.7276, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.8061420345489443, | |
| "grad_norm": 0.8150967413326101, | |
| "learning_rate": 2.1976549710802754e-06, | |
| "loss": 0.7364, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8125399872040947, | |
| "grad_norm": 0.9104498719001516, | |
| "learning_rate": 2.0596607360168897e-06, | |
| "loss": 0.721, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.818937939859245, | |
| "grad_norm": 0.8059019041226441, | |
| "learning_rate": 1.9256419525319316e-06, | |
| "loss": 0.7003, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8253358925143954, | |
| "grad_norm": 0.8016220960820958, | |
| "learning_rate": 1.7956657191669969e-06, | |
| "loss": 0.7405, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.8317338451695457, | |
| "grad_norm": 0.8296186304272232, | |
| "learning_rate": 1.6697971104994847e-06, | |
| "loss": 0.7237, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.838131797824696, | |
| "grad_norm": 0.8575579265825671, | |
| "learning_rate": 1.5480991445620541e-06, | |
| "loss": 0.7252, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.8445297504798465, | |
| "grad_norm": 0.8443927649243143, | |
| "learning_rate": 1.4306327512916574e-06, | |
| "loss": 0.7178, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.8509277031349968, | |
| "grad_norm": 0.7908279914848837, | |
| "learning_rate": 1.3174567420240647e-06, | |
| "loss": 0.6936, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.8573256557901472, | |
| "grad_norm": 0.8218483754182432, | |
| "learning_rate": 1.2086277800490554e-06, | |
| "loss": 0.7007, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8637236084452975, | |
| "grad_norm": 0.8256511379428485, | |
| "learning_rate": 1.1042003522410882e-06, | |
| "loss": 0.6889, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.8701215611004478, | |
| "grad_norm": 0.8968851185464259, | |
| "learning_rate": 1.0042267417796292e-06, | |
| "loss": 0.7123, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8765195137555982, | |
| "grad_norm": 0.8823080752196552, | |
| "learning_rate": 9.08757001972762e-07, | |
| "loss": 0.7021, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.8829174664107485, | |
| "grad_norm": 0.8466732079799328, | |
| "learning_rate": 8.178389311972612e-07, | |
| "loss": 0.6956, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.889315419065899, | |
| "grad_norm": 0.8760114826953183, | |
| "learning_rate": 7.315180489675822e-07, | |
| "loss": 0.6976, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.8957133717210493, | |
| "grad_norm": 0.8384622643848926, | |
| "learning_rate": 6.498375731458529e-07, | |
| "loss": 0.7103, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9021113243761996, | |
| "grad_norm": 0.8194072342460205, | |
| "learning_rate": 5.728383983041696e-07, | |
| "loss": 0.7157, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.90850927703135, | |
| "grad_norm": 0.9294398002536329, | |
| "learning_rate": 5.005590752501244e-07, | |
| "loss": 0.7248, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9149072296865003, | |
| "grad_norm": 0.8943875627138937, | |
| "learning_rate": 4.3303579172574884e-07, | |
| "loss": 0.7171, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.9213051823416507, | |
| "grad_norm": 0.8691772678725829, | |
| "learning_rate": 3.7030235428956895e-07, | |
| "loss": 0.7317, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.927703134996801, | |
| "grad_norm": 0.9150925426459633, | |
| "learning_rate": 3.1239017139084725e-07, | |
| "loss": 0.7459, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.9341010876519513, | |
| "grad_norm": 0.8226066365805443, | |
| "learning_rate": 2.593282376444539e-07, | |
| "loss": 0.7004, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9404990403071017, | |
| "grad_norm": 0.827122609753977, | |
| "learning_rate": 2.11143119314281e-07, | |
| "loss": 0.7353, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.946896992962252, | |
| "grad_norm": 0.928354448329939, | |
| "learning_rate": 1.6785894101243205e-07, | |
| "loss": 0.7494, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.9532949456174025, | |
| "grad_norm": 0.8296013670372103, | |
| "learning_rate": 1.2949737362087156e-07, | |
| "loss": 0.7057, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.9596928982725528, | |
| "grad_norm": 0.8790221772454616, | |
| "learning_rate": 9.607762344156946e-08, | |
| "loss": 0.725, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9660908509277031, | |
| "grad_norm": 0.8987832603853396, | |
| "learning_rate": 6.761642258056977e-08, | |
| "loss": 0.7245, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.9724888035828535, | |
| "grad_norm": 0.7657247344848286, | |
| "learning_rate": 4.412802057081278e-08, | |
| "loss": 0.7057, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9788867562380038, | |
| "grad_norm": 0.7827327625691514, | |
| "learning_rate": 2.5624177237884017e-08, | |
| "loss": 0.72, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.9852847088931542, | |
| "grad_norm": 0.8361648349847708, | |
| "learning_rate": 1.2114156812284006e-08, | |
| "loss": 0.7083, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.9916826615483045, | |
| "grad_norm": 0.8873989576007488, | |
| "learning_rate": 3.6047232911462506e-09, | |
| "loss": 0.6976, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.9980806142034548, | |
| "grad_norm": 0.873748350676312, | |
| "learning_rate": 1.0013705174061195e-10, | |
| "loss": 0.711, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.999360204734485, | |
| "eval_loss": 0.7211461067199707, | |
| "eval_runtime": 620.1493, | |
| "eval_samples_per_second": 69.338, | |
| "eval_steps_per_second": 2.167, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.999360204734485, | |
| "step": 781, | |
| "total_flos": 80545737277440.0, | |
| "train_loss": 0.7457574072476386, | |
| "train_runtime": 7303.4461, | |
| "train_samples_per_second": 13.692, | |
| "train_steps_per_second": 0.107 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 781, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 80545737277440.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |