| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 4054, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00493339911198816, | |
| "grad_norm": 32.17447280883789, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.0607, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00986679822397632, | |
| "grad_norm": 30.161712646484375, | |
| "learning_rate": 4.995882062263219e-05, | |
| "loss": 3.7481, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01480019733596448, | |
| "grad_norm": 38.9721565246582, | |
| "learning_rate": 4.9884697743370127e-05, | |
| "loss": 3.9839, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01973359644795264, | |
| "grad_norm": 29.38846778869629, | |
| "learning_rate": 4.9802338988634494e-05, | |
| "loss": 3.7565, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0246669955599408, | |
| "grad_norm": 32.933860778808594, | |
| "learning_rate": 4.9728216109372424e-05, | |
| "loss": 3.7562, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02960039467192896, | |
| "grad_norm": 24.409399032592773, | |
| "learning_rate": 4.96458573546368e-05, | |
| "loss": 3.3008, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03453379378391712, | |
| "grad_norm": 25.322418212890625, | |
| "learning_rate": 4.956349859990117e-05, | |
| "loss": 3.0558, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03946719289590528, | |
| "grad_norm": 28.60622215270996, | |
| "learning_rate": 4.948113984516554e-05, | |
| "loss": 3.4003, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04440059200789344, | |
| "grad_norm": 15.609489440917969, | |
| "learning_rate": 4.9398781090429915e-05, | |
| "loss": 3.0121, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0493339911198816, | |
| "grad_norm": 22.34126091003418, | |
| "learning_rate": 4.931642233569428e-05, | |
| "loss": 3.1712, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05426739023186976, | |
| "grad_norm": 19.47034454345703, | |
| "learning_rate": 4.923406358095866e-05, | |
| "loss": 3.2003, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05920078934385792, | |
| "grad_norm": 20.92117691040039, | |
| "learning_rate": 4.9151704826223025e-05, | |
| "loss": 3.2641, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06413418845584608, | |
| "grad_norm": 10.978850364685059, | |
| "learning_rate": 4.90693460714874e-05, | |
| "loss": 3.0391, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06906758756783424, | |
| "grad_norm": 12.89031982421875, | |
| "learning_rate": 4.8986987316751774e-05, | |
| "loss": 2.7204, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.0740009866798224, | |
| "grad_norm": 9.097904205322266, | |
| "learning_rate": 4.890462856201615e-05, | |
| "loss": 2.9097, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07893438579181056, | |
| "grad_norm": 10.250645637512207, | |
| "learning_rate": 4.8822269807280516e-05, | |
| "loss": 2.8115, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08386778490379872, | |
| "grad_norm": 9.343478202819824, | |
| "learning_rate": 4.873991105254489e-05, | |
| "loss": 3.2778, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08880118401578688, | |
| "grad_norm": 10.233019828796387, | |
| "learning_rate": 4.8657552297809264e-05, | |
| "loss": 2.8652, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09373458312777504, | |
| "grad_norm": 9.129461288452148, | |
| "learning_rate": 4.857519354307363e-05, | |
| "loss": 2.7035, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0986679822397632, | |
| "grad_norm": 6.655858039855957, | |
| "learning_rate": 4.8492834788338006e-05, | |
| "loss": 2.8944, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10360138135175136, | |
| "grad_norm": 8.336820602416992, | |
| "learning_rate": 4.8410476033602374e-05, | |
| "loss": 2.4224, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10853478046373952, | |
| "grad_norm": 16.6269474029541, | |
| "learning_rate": 4.832811727886675e-05, | |
| "loss": 2.9059, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.11346817957572768, | |
| "grad_norm": 14.81032657623291, | |
| "learning_rate": 4.8245758524131116e-05, | |
| "loss": 3.026, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11840157868771584, | |
| "grad_norm": 5.8194193840026855, | |
| "learning_rate": 4.816339976939549e-05, | |
| "loss": 2.6034, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.123334977799704, | |
| "grad_norm": 12.556567192077637, | |
| "learning_rate": 4.808104101465986e-05, | |
| "loss": 2.9635, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12826837691169216, | |
| "grad_norm": 11.739566802978516, | |
| "learning_rate": 4.799868225992423e-05, | |
| "loss": 3.1042, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.13320177602368033, | |
| "grad_norm": 9.672442436218262, | |
| "learning_rate": 4.791632350518861e-05, | |
| "loss": 3.0241, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.13813517513566848, | |
| "grad_norm": 5.8154497146606445, | |
| "learning_rate": 4.7833964750452975e-05, | |
| "loss": 2.7709, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.14306857424765665, | |
| "grad_norm": 20.954362869262695, | |
| "learning_rate": 4.775160599571735e-05, | |
| "loss": 2.8549, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1480019733596448, | |
| "grad_norm": 9.915385246276855, | |
| "learning_rate": 4.7669247240981717e-05, | |
| "loss": 2.9146, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15293537247163297, | |
| "grad_norm": 7.093989372253418, | |
| "learning_rate": 4.758688848624609e-05, | |
| "loss": 2.6771, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.15786877158362111, | |
| "grad_norm": 8.882306098937988, | |
| "learning_rate": 4.750452973151046e-05, | |
| "loss": 2.6295, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1628021706956093, | |
| "grad_norm": 17.28571128845215, | |
| "learning_rate": 4.742217097677483e-05, | |
| "loss": 2.492, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.16773556980759743, | |
| "grad_norm": 18.01361846923828, | |
| "learning_rate": 4.73398122220392e-05, | |
| "loss": 3.288, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1726689689195856, | |
| "grad_norm": 10.41086196899414, | |
| "learning_rate": 4.7257453467303575e-05, | |
| "loss": 2.6676, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.17760236803157375, | |
| "grad_norm": 12.40413761138916, | |
| "learning_rate": 4.717509471256795e-05, | |
| "loss": 2.8085, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.18253576714356193, | |
| "grad_norm": 13.229728698730469, | |
| "learning_rate": 4.709273595783232e-05, | |
| "loss": 2.2978, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.18746916625555007, | |
| "grad_norm": 7.240847587585449, | |
| "learning_rate": 4.701037720309669e-05, | |
| "loss": 2.7408, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.19240256536753825, | |
| "grad_norm": 10.95034122467041, | |
| "learning_rate": 4.692801844836106e-05, | |
| "loss": 2.7323, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1973359644795264, | |
| "grad_norm": 11.429134368896484, | |
| "learning_rate": 4.6845659693625434e-05, | |
| "loss": 2.6121, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.20226936359151457, | |
| "grad_norm": 14.395548820495605, | |
| "learning_rate": 4.67633009388898e-05, | |
| "loss": 3.0402, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2072027627035027, | |
| "grad_norm": 5.770746231079102, | |
| "learning_rate": 4.668094218415418e-05, | |
| "loss": 2.4771, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2121361618154909, | |
| "grad_norm": 8.25928783416748, | |
| "learning_rate": 4.659858342941855e-05, | |
| "loss": 2.1011, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.21706956092747903, | |
| "grad_norm": 14.545477867126465, | |
| "learning_rate": 4.6516224674682924e-05, | |
| "loss": 3.0616, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.2220029600394672, | |
| "grad_norm": 11.545022964477539, | |
| "learning_rate": 4.643386591994729e-05, | |
| "loss": 2.2621, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.22693635915145535, | |
| "grad_norm": 25.751686096191406, | |
| "learning_rate": 4.6351507165211666e-05, | |
| "loss": 2.7998, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.23186975826344353, | |
| "grad_norm": 8.96506118774414, | |
| "learning_rate": 4.626914841047604e-05, | |
| "loss": 2.2543, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.23680315737543167, | |
| "grad_norm": 8.918927192687988, | |
| "learning_rate": 4.618678965574041e-05, | |
| "loss": 2.2573, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.24173655648741985, | |
| "grad_norm": 8.464729309082031, | |
| "learning_rate": 4.610443090100478e-05, | |
| "loss": 2.7079, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.246669955599408, | |
| "grad_norm": 9.733293533325195, | |
| "learning_rate": 4.602207214626915e-05, | |
| "loss": 3.423, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25160335471139617, | |
| "grad_norm": 11.454011917114258, | |
| "learning_rate": 4.5939713391533525e-05, | |
| "loss": 3.1205, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2565367538233843, | |
| "grad_norm": 8.79787826538086, | |
| "learning_rate": 4.585735463679789e-05, | |
| "loss": 2.4529, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.26147015293537246, | |
| "grad_norm": 6.300372123718262, | |
| "learning_rate": 4.577499588206227e-05, | |
| "loss": 2.3711, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.26640355204736066, | |
| "grad_norm": 5.140571117401123, | |
| "learning_rate": 4.5692637127326634e-05, | |
| "loss": 2.3833, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2713369511593488, | |
| "grad_norm": 11.252778053283691, | |
| "learning_rate": 4.561027837259101e-05, | |
| "loss": 2.3281, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.27627035027133695, | |
| "grad_norm": 10.254585266113281, | |
| "learning_rate": 4.552791961785538e-05, | |
| "loss": 3.0383, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2812037493833251, | |
| "grad_norm": 8.8512544631958, | |
| "learning_rate": 4.544556086311975e-05, | |
| "loss": 2.3813, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2861371484953133, | |
| "grad_norm": 4.089067459106445, | |
| "learning_rate": 4.5363202108384125e-05, | |
| "loss": 2.5559, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.29107054760730144, | |
| "grad_norm": 5.320517539978027, | |
| "learning_rate": 4.528084335364849e-05, | |
| "loss": 2.7508, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2960039467192896, | |
| "grad_norm": 6.5414862632751465, | |
| "learning_rate": 4.519848459891287e-05, | |
| "loss": 2.3911, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.30093734583127774, | |
| "grad_norm": 7.56878137588501, | |
| "learning_rate": 4.5116125844177235e-05, | |
| "loss": 2.9435, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.30587074494326594, | |
| "grad_norm": 10.586882591247559, | |
| "learning_rate": 4.503376708944161e-05, | |
| "loss": 2.6231, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.3108041440552541, | |
| "grad_norm": 9.455619812011719, | |
| "learning_rate": 4.495140833470598e-05, | |
| "loss": 2.8628, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.31573754316724223, | |
| "grad_norm": 5.5599565505981445, | |
| "learning_rate": 4.486904957997035e-05, | |
| "loss": 2.2629, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3206709422792304, | |
| "grad_norm": 5.07106351852417, | |
| "learning_rate": 4.4786690825234726e-05, | |
| "loss": 2.1062, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3256043413912186, | |
| "grad_norm": 8.023340225219727, | |
| "learning_rate": 4.470433207049909e-05, | |
| "loss": 2.4445, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3305377405032067, | |
| "grad_norm": 12.941463470458984, | |
| "learning_rate": 4.462197331576347e-05, | |
| "loss": 2.3829, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.33547113961519487, | |
| "grad_norm": 9.589372634887695, | |
| "learning_rate": 4.4539614561027835e-05, | |
| "loss": 2.8525, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.340404538727183, | |
| "grad_norm": 29.511812210083008, | |
| "learning_rate": 4.445725580629221e-05, | |
| "loss": 2.4082, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3453379378391712, | |
| "grad_norm": 17.187841415405273, | |
| "learning_rate": 4.437489705155658e-05, | |
| "loss": 2.3018, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.35027133695115936, | |
| "grad_norm": 6.561307430267334, | |
| "learning_rate": 4.429253829682096e-05, | |
| "loss": 2.4964, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3552047360631475, | |
| "grad_norm": 8.075435638427734, | |
| "learning_rate": 4.4210179542085326e-05, | |
| "loss": 2.4871, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.36013813517513565, | |
| "grad_norm": 8.039571762084961, | |
| "learning_rate": 4.41278207873497e-05, | |
| "loss": 2.4563, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.36507153428712386, | |
| "grad_norm": 24.152629852294922, | |
| "learning_rate": 4.404546203261407e-05, | |
| "loss": 2.549, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.370004933399112, | |
| "grad_norm": 2.6132395267486572, | |
| "learning_rate": 4.396310327787844e-05, | |
| "loss": 2.6625, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.37493833251110015, | |
| "grad_norm": 11.909361839294434, | |
| "learning_rate": 4.388074452314282e-05, | |
| "loss": 3.0034, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.3798717316230883, | |
| "grad_norm": 6.0602240562438965, | |
| "learning_rate": 4.3798385768407185e-05, | |
| "loss": 2.8432, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.3848051307350765, | |
| "grad_norm": 6.063085079193115, | |
| "learning_rate": 4.371602701367156e-05, | |
| "loss": 2.4882, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.38973852984706464, | |
| "grad_norm": 12.514052391052246, | |
| "learning_rate": 4.363366825893593e-05, | |
| "loss": 2.5734, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3946719289590528, | |
| "grad_norm": 5.626593589782715, | |
| "learning_rate": 4.35513095042003e-05, | |
| "loss": 1.8927, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.39960532807104093, | |
| "grad_norm": 6.073607444763184, | |
| "learning_rate": 4.346895074946467e-05, | |
| "loss": 2.4321, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.40453872718302913, | |
| "grad_norm": 7.1377763748168945, | |
| "learning_rate": 4.338659199472904e-05, | |
| "loss": 3.0152, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4094721262950173, | |
| "grad_norm": 5.029105186462402, | |
| "learning_rate": 4.330423323999341e-05, | |
| "loss": 2.3915, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4144055254070054, | |
| "grad_norm": 5.954276084899902, | |
| "learning_rate": 4.3221874485257785e-05, | |
| "loss": 2.5008, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.41933892451899357, | |
| "grad_norm": 7.565403938293457, | |
| "learning_rate": 4.313951573052216e-05, | |
| "loss": 2.5035, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4242723236309818, | |
| "grad_norm": 7.7029314041137695, | |
| "learning_rate": 4.305715697578653e-05, | |
| "loss": 2.4186, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4292057227429699, | |
| "grad_norm": 8.457364082336426, | |
| "learning_rate": 4.29747982210509e-05, | |
| "loss": 2.5165, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.43413912185495807, | |
| "grad_norm": 10.918882369995117, | |
| "learning_rate": 4.289243946631527e-05, | |
| "loss": 2.4591, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4390725209669462, | |
| "grad_norm": 9.553812026977539, | |
| "learning_rate": 4.2810080711579644e-05, | |
| "loss": 2.6194, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4440059200789344, | |
| "grad_norm": 9.679872512817383, | |
| "learning_rate": 4.272772195684401e-05, | |
| "loss": 2.9103, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.44893931919092256, | |
| "grad_norm": 5.805199146270752, | |
| "learning_rate": 4.2645363202108386e-05, | |
| "loss": 2.202, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4538727183029107, | |
| "grad_norm": 7.699798107147217, | |
| "learning_rate": 4.256300444737275e-05, | |
| "loss": 3.0242, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.45880611741489885, | |
| "grad_norm": 5.121403217315674, | |
| "learning_rate": 4.248064569263713e-05, | |
| "loss": 2.2204, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.46373951652688705, | |
| "grad_norm": 5.217260837554932, | |
| "learning_rate": 4.23982869379015e-05, | |
| "loss": 2.6708, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.4686729156388752, | |
| "grad_norm": 9.238679885864258, | |
| "learning_rate": 4.231592818316587e-05, | |
| "loss": 2.7795, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.47360631475086334, | |
| "grad_norm": 4.4800496101379395, | |
| "learning_rate": 4.2233569428430244e-05, | |
| "loss": 2.5765, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.4785397138628515, | |
| "grad_norm": 11.48628044128418, | |
| "learning_rate": 4.215121067369461e-05, | |
| "loss": 2.7761, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.4834731129748397, | |
| "grad_norm": 8.910355567932129, | |
| "learning_rate": 4.2068851918958986e-05, | |
| "loss": 2.8094, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.48840651208682784, | |
| "grad_norm": 7.16322660446167, | |
| "learning_rate": 4.198649316422336e-05, | |
| "loss": 2.3914, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.493339911198816, | |
| "grad_norm": 15.859028816223145, | |
| "learning_rate": 4.1904134409487735e-05, | |
| "loss": 2.5077, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.49827331031080413, | |
| "grad_norm": 10.420004844665527, | |
| "learning_rate": 4.18217756547521e-05, | |
| "loss": 2.4869, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5032067094227923, | |
| "grad_norm": 5.477315902709961, | |
| "learning_rate": 4.173941690001648e-05, | |
| "loss": 2.8359, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5081401085347804, | |
| "grad_norm": 5.169394493103027, | |
| "learning_rate": 4.1657058145280845e-05, | |
| "loss": 2.4023, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5130735076467686, | |
| "grad_norm": 8.969653129577637, | |
| "learning_rate": 4.157469939054522e-05, | |
| "loss": 2.7727, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5180069067587568, | |
| "grad_norm": 7.325026512145996, | |
| "learning_rate": 4.149234063580959e-05, | |
| "loss": 2.3592, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5229403058707449, | |
| "grad_norm": 5.182565212249756, | |
| "learning_rate": 4.140998188107396e-05, | |
| "loss": 2.4861, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5278737049827331, | |
| "grad_norm": 6.98890495300293, | |
| "learning_rate": 4.1327623126338335e-05, | |
| "loss": 1.9875, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5328071040947213, | |
| "grad_norm": 7.717536449432373, | |
| "learning_rate": 4.12452643716027e-05, | |
| "loss": 1.7559, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5377405032067094, | |
| "grad_norm": 8.016460418701172, | |
| "learning_rate": 4.116290561686708e-05, | |
| "loss": 2.8117, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5426739023186976, | |
| "grad_norm": 6.484166622161865, | |
| "learning_rate": 4.1080546862131445e-05, | |
| "loss": 2.4036, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5476073014306857, | |
| "grad_norm": 4.496994495391846, | |
| "learning_rate": 4.099818810739582e-05, | |
| "loss": 2.4948, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5525407005426739, | |
| "grad_norm": 6.8036065101623535, | |
| "learning_rate": 4.091582935266019e-05, | |
| "loss": 2.4636, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5574740996546621, | |
| "grad_norm": 8.330781936645508, | |
| "learning_rate": 4.083347059792456e-05, | |
| "loss": 2.6983, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5624074987666502, | |
| "grad_norm": 6.868759632110596, | |
| "learning_rate": 4.0751111843188936e-05, | |
| "loss": 2.5327, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.5673408978786384, | |
| "grad_norm": 7.583151340484619, | |
| "learning_rate": 4.0668753088453303e-05, | |
| "loss": 2.4815, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.5722742969906266, | |
| "grad_norm": 7.823330402374268, | |
| "learning_rate": 4.058639433371768e-05, | |
| "loss": 2.2532, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5772076961026147, | |
| "grad_norm": 6.149056434631348, | |
| "learning_rate": 4.0504035578982045e-05, | |
| "loss": 2.762, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.5821410952146029, | |
| "grad_norm": 9.300018310546875, | |
| "learning_rate": 4.042167682424642e-05, | |
| "loss": 2.505, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.587074494326591, | |
| "grad_norm": 10.552204132080078, | |
| "learning_rate": 4.033931806951079e-05, | |
| "loss": 2.6128, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5920078934385792, | |
| "grad_norm": 10.5072603225708, | |
| "learning_rate": 4.025695931477516e-05, | |
| "loss": 2.6206, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5969412925505674, | |
| "grad_norm": 5.355686664581299, | |
| "learning_rate": 4.017460056003953e-05, | |
| "loss": 2.8745, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6018746916625555, | |
| "grad_norm": 4.0471649169921875, | |
| "learning_rate": 4.0092241805303904e-05, | |
| "loss": 2.7503, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6068080907745437, | |
| "grad_norm": 6.991344928741455, | |
| "learning_rate": 4.000988305056828e-05, | |
| "loss": 2.686, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6117414898865319, | |
| "grad_norm": 4.674221515655518, | |
| "learning_rate": 3.9927524295832646e-05, | |
| "loss": 2.5496, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.61667488899852, | |
| "grad_norm": 23.224143981933594, | |
| "learning_rate": 3.984516554109702e-05, | |
| "loss": 3.0394, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6216082881105082, | |
| "grad_norm": 8.688756942749023, | |
| "learning_rate": 3.976280678636139e-05, | |
| "loss": 2.7851, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6265416872224963, | |
| "grad_norm": 6.090677261352539, | |
| "learning_rate": 3.968044803162576e-05, | |
| "loss": 2.2354, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6314750863344845, | |
| "grad_norm": 9.011785507202148, | |
| "learning_rate": 3.959808927689014e-05, | |
| "loss": 2.1909, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6364084854464727, | |
| "grad_norm": 9.320941925048828, | |
| "learning_rate": 3.951573052215451e-05, | |
| "loss": 2.1964, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6413418845584608, | |
| "grad_norm": 6.079468250274658, | |
| "learning_rate": 3.943337176741888e-05, | |
| "loss": 2.1266, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.646275283670449, | |
| "grad_norm": 6.427783489227295, | |
| "learning_rate": 3.935101301268325e-05, | |
| "loss": 2.1273, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6512086827824372, | |
| "grad_norm": 5.442935943603516, | |
| "learning_rate": 3.926865425794762e-05, | |
| "loss": 2.3792, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.6561420818944252, | |
| "grad_norm": 5.503900527954102, | |
| "learning_rate": 3.9186295503211995e-05, | |
| "loss": 2.3938, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.6610754810064134, | |
| "grad_norm": 9.838761329650879, | |
| "learning_rate": 3.910393674847637e-05, | |
| "loss": 2.2793, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6660088801184015, | |
| "grad_norm": 7.890000343322754, | |
| "learning_rate": 3.902157799374074e-05, | |
| "loss": 2.3842, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.6709422792303897, | |
| "grad_norm": 7.321061134338379, | |
| "learning_rate": 3.893921923900511e-05, | |
| "loss": 2.3586, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6758756783423779, | |
| "grad_norm": 5.176478385925293, | |
| "learning_rate": 3.885686048426948e-05, | |
| "loss": 2.6807, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.680809077454366, | |
| "grad_norm": 4.9928975105285645, | |
| "learning_rate": 3.8774501729533854e-05, | |
| "loss": 1.8495, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6857424765663542, | |
| "grad_norm": 6.170323371887207, | |
| "learning_rate": 3.869214297479822e-05, | |
| "loss": 2.415, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6906758756783424, | |
| "grad_norm": 10.49080753326416, | |
| "learning_rate": 3.8609784220062596e-05, | |
| "loss": 1.8006, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6956092747903305, | |
| "grad_norm": 16.418981552124023, | |
| "learning_rate": 3.852742546532696e-05, | |
| "loss": 2.5657, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7005426739023187, | |
| "grad_norm": 4.002269268035889, | |
| "learning_rate": 3.844506671059134e-05, | |
| "loss": 2.5927, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7054760730143068, | |
| "grad_norm": 7.075521469116211, | |
| "learning_rate": 3.836270795585571e-05, | |
| "loss": 1.8786, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.710409472126295, | |
| "grad_norm": 6.2812981605529785, | |
| "learning_rate": 3.828034920112008e-05, | |
| "loss": 2.1746, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.7153428712382832, | |
| "grad_norm": 7.315672874450684, | |
| "learning_rate": 3.8197990446384454e-05, | |
| "loss": 2.4418, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7202762703502713, | |
| "grad_norm": 7.978590488433838, | |
| "learning_rate": 3.811563169164882e-05, | |
| "loss": 2.3702, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7252096694622595, | |
| "grad_norm": 8.856497764587402, | |
| "learning_rate": 3.8033272936913196e-05, | |
| "loss": 2.7627, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7301430685742477, | |
| "grad_norm": 5.009411811828613, | |
| "learning_rate": 3.7950914182177564e-05, | |
| "loss": 2.1163, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7350764676862358, | |
| "grad_norm": 6.4281134605407715, | |
| "learning_rate": 3.786855542744194e-05, | |
| "loss": 2.554, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.740009866798224, | |
| "grad_norm": 6.119235515594482, | |
| "learning_rate": 3.7786196672706306e-05, | |
| "loss": 2.1492, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7449432659102121, | |
| "grad_norm": 4.099040508270264, | |
| "learning_rate": 3.770383791797068e-05, | |
| "loss": 2.9967, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.7498766650222003, | |
| "grad_norm": 6.364965438842773, | |
| "learning_rate": 3.7621479163235055e-05, | |
| "loss": 2.7822, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.7548100641341885, | |
| "grad_norm": 13.07193374633789, | |
| "learning_rate": 3.753912040849942e-05, | |
| "loss": 2.1634, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.7597434632461766, | |
| "grad_norm": 4.285301685333252, | |
| "learning_rate": 3.74567616537638e-05, | |
| "loss": 2.49, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.7646768623581648, | |
| "grad_norm": 8.684725761413574, | |
| "learning_rate": 3.7374402899028164e-05, | |
| "loss": 2.4202, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.769610261470153, | |
| "grad_norm": 7.857943058013916, | |
| "learning_rate": 3.7292044144292545e-05, | |
| "loss": 2.5489, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7745436605821411, | |
| "grad_norm": 10.40462589263916, | |
| "learning_rate": 3.720968538955691e-05, | |
| "loss": 2.5975, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7794770596941293, | |
| "grad_norm": 3.614388942718506, | |
| "learning_rate": 3.712732663482129e-05, | |
| "loss": 2.5092, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7844104588061174, | |
| "grad_norm": 4.367038726806641, | |
| "learning_rate": 3.7044967880085655e-05, | |
| "loss": 2.4972, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7893438579181056, | |
| "grad_norm": 8.629862785339355, | |
| "learning_rate": 3.696260912535003e-05, | |
| "loss": 2.8085, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7942772570300938, | |
| "grad_norm": 4.455677032470703, | |
| "learning_rate": 3.68802503706144e-05, | |
| "loss": 2.2631, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7992106561420819, | |
| "grad_norm": 3.8020565509796143, | |
| "learning_rate": 3.679789161587877e-05, | |
| "loss": 2.1875, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8041440552540701, | |
| "grad_norm": 8.94434928894043, | |
| "learning_rate": 3.6715532861143146e-05, | |
| "loss": 2.4996, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8090774543660583, | |
| "grad_norm": 6.989756107330322, | |
| "learning_rate": 3.6633174106407514e-05, | |
| "loss": 2.0738, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.8140108534780464, | |
| "grad_norm": 6.769233226776123, | |
| "learning_rate": 3.655081535167189e-05, | |
| "loss": 2.4975, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8189442525900346, | |
| "grad_norm": 5.789883136749268, | |
| "learning_rate": 3.6468456596936256e-05, | |
| "loss": 2.1423, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.8238776517020227, | |
| "grad_norm": 8.308130264282227, | |
| "learning_rate": 3.638609784220063e-05, | |
| "loss": 2.364, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8288110508140109, | |
| "grad_norm": 7.824023246765137, | |
| "learning_rate": 3.6303739087465e-05, | |
| "loss": 2.001, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.833744449925999, | |
| "grad_norm": 6.625056266784668, | |
| "learning_rate": 3.622138033272937e-05, | |
| "loss": 2.4029, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.8386778490379871, | |
| "grad_norm": 3.973498582839966, | |
| "learning_rate": 3.613902157799374e-05, | |
| "loss": 2.766, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.8436112481499753, | |
| "grad_norm": 6.130688667297363, | |
| "learning_rate": 3.6056662823258114e-05, | |
| "loss": 2.4471, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.8485446472619635, | |
| "grad_norm": 4.856011867523193, | |
| "learning_rate": 3.597430406852249e-05, | |
| "loss": 2.035, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.8534780463739516, | |
| "grad_norm": 5.570549488067627, | |
| "learning_rate": 3.5891945313786856e-05, | |
| "loss": 1.8751, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.8584114454859398, | |
| "grad_norm": 8.646450996398926, | |
| "learning_rate": 3.580958655905123e-05, | |
| "loss": 2.0868, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.8633448445979279, | |
| "grad_norm": 5.265438556671143, | |
| "learning_rate": 3.57272278043156e-05, | |
| "loss": 2.5141, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.8682782437099161, | |
| "grad_norm": 11.451949119567871, | |
| "learning_rate": 3.564486904957997e-05, | |
| "loss": 2.3301, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.8732116428219043, | |
| "grad_norm": 8.040514945983887, | |
| "learning_rate": 3.556251029484434e-05, | |
| "loss": 2.4714, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8781450419338924, | |
| "grad_norm": 6.2226481437683105, | |
| "learning_rate": 3.5480151540108715e-05, | |
| "loss": 2.667, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8830784410458806, | |
| "grad_norm": 9.629585266113281, | |
| "learning_rate": 3.539779278537308e-05, | |
| "loss": 2.6848, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8880118401578688, | |
| "grad_norm": 8.996950149536133, | |
| "learning_rate": 3.5315434030637457e-05, | |
| "loss": 2.4991, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8929452392698569, | |
| "grad_norm": 7.078036785125732, | |
| "learning_rate": 3.523307527590183e-05, | |
| "loss": 2.5703, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.8978786383818451, | |
| "grad_norm": 6.913575649261475, | |
| "learning_rate": 3.51507165211662e-05, | |
| "loss": 2.5782, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9028120374938332, | |
| "grad_norm": 8.875394821166992, | |
| "learning_rate": 3.506835776643057e-05, | |
| "loss": 2.1843, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.9077454366058214, | |
| "grad_norm": 12.219450950622559, | |
| "learning_rate": 3.498599901169494e-05, | |
| "loss": 2.7252, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9126788357178096, | |
| "grad_norm": 8.12746524810791, | |
| "learning_rate": 3.490364025695932e-05, | |
| "loss": 2.2132, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9176122348297977, | |
| "grad_norm": 17.30431365966797, | |
| "learning_rate": 3.482128150222369e-05, | |
| "loss": 2.807, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.9225456339417859, | |
| "grad_norm": 4.950164794921875, | |
| "learning_rate": 3.4738922747488064e-05, | |
| "loss": 2.2689, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.9274790330537741, | |
| "grad_norm": 6.925759315490723, | |
| "learning_rate": 3.465656399275243e-05, | |
| "loss": 2.3312, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.9324124321657622, | |
| "grad_norm": 5.608890533447266, | |
| "learning_rate": 3.4574205238016806e-05, | |
| "loss": 2.305, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.9373458312777504, | |
| "grad_norm": 8.436676025390625, | |
| "learning_rate": 3.4491846483281173e-05, | |
| "loss": 2.4468, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.9422792303897385, | |
| "grad_norm": 14.684228897094727, | |
| "learning_rate": 3.440948772854555e-05, | |
| "loss": 2.4759, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.9472126295017267, | |
| "grad_norm": 9.368124961853027, | |
| "learning_rate": 3.432712897380992e-05, | |
| "loss": 2.5555, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.9521460286137149, | |
| "grad_norm": 8.614909172058105, | |
| "learning_rate": 3.424477021907429e-05, | |
| "loss": 2.351, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.957079427725703, | |
| "grad_norm": 10.137232780456543, | |
| "learning_rate": 3.4162411464338664e-05, | |
| "loss": 2.2095, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.9620128268376912, | |
| "grad_norm": 5.5300493240356445, | |
| "learning_rate": 3.408005270960303e-05, | |
| "loss": 2.2628, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.9669462259496794, | |
| "grad_norm": 5.127216815948486, | |
| "learning_rate": 3.3997693954867406e-05, | |
| "loss": 1.9502, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.9718796250616675, | |
| "grad_norm": 4.680507659912109, | |
| "learning_rate": 3.3915335200131774e-05, | |
| "loss": 1.8751, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.9768130241736557, | |
| "grad_norm": 6.784450531005859, | |
| "learning_rate": 3.383297644539615e-05, | |
| "loss": 2.477, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.9817464232856438, | |
| "grad_norm": 6.171282768249512, | |
| "learning_rate": 3.3750617690660516e-05, | |
| "loss": 2.4942, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.986679822397632, | |
| "grad_norm": 6.970947265625, | |
| "learning_rate": 3.366825893592489e-05, | |
| "loss": 2.2718, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9916132215096202, | |
| "grad_norm": 4.48728084564209, | |
| "learning_rate": 3.3585900181189265e-05, | |
| "loss": 2.6097, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.9965466206216083, | |
| "grad_norm": 13.015890121459961, | |
| "learning_rate": 3.350354142645363e-05, | |
| "loss": 2.7752, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.0014800197335965, | |
| "grad_norm": 7.251813888549805, | |
| "learning_rate": 3.342118267171801e-05, | |
| "loss": 2.0151, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.0064134188455847, | |
| "grad_norm": 5.12528133392334, | |
| "learning_rate": 3.3338823916982374e-05, | |
| "loss": 2.0101, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.0113468179575729, | |
| "grad_norm": 5.179903507232666, | |
| "learning_rate": 3.325646516224675e-05, | |
| "loss": 2.0539, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.0162802170695608, | |
| "grad_norm": 13.52995777130127, | |
| "learning_rate": 3.3174106407511116e-05, | |
| "loss": 1.9679, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.021213616181549, | |
| "grad_norm": 10.211468696594238, | |
| "learning_rate": 3.309174765277549e-05, | |
| "loss": 2.4251, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.0261470152935372, | |
| "grad_norm": 11.468674659729004, | |
| "learning_rate": 3.300938889803986e-05, | |
| "loss": 2.026, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.0310804144055254, | |
| "grad_norm": 9.60104751586914, | |
| "learning_rate": 3.292703014330423e-05, | |
| "loss": 2.0974, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.0360138135175136, | |
| "grad_norm": 16.93134307861328, | |
| "learning_rate": 3.284467138856861e-05, | |
| "loss": 1.8676, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.0409472126295016, | |
| "grad_norm": 7.654803276062012, | |
| "learning_rate": 3.2762312633832975e-05, | |
| "loss": 1.6796, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.0458806117414898, | |
| "grad_norm": 7.115079879760742, | |
| "learning_rate": 3.267995387909735e-05, | |
| "loss": 1.9246, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.050814010853478, | |
| "grad_norm": 5.4859185218811035, | |
| "learning_rate": 3.2597595124361724e-05, | |
| "loss": 2.0991, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.0557474099654662, | |
| "grad_norm": 16.93896484375, | |
| "learning_rate": 3.25152363696261e-05, | |
| "loss": 1.87, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.0606808090774544, | |
| "grad_norm": 8.788496017456055, | |
| "learning_rate": 3.2432877614890466e-05, | |
| "loss": 1.8518, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.0656142081894426, | |
| "grad_norm": 5.605971336364746, | |
| "learning_rate": 3.235051886015484e-05, | |
| "loss": 1.4863, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.0705476073014306, | |
| "grad_norm": 8.495088577270508, | |
| "learning_rate": 3.226816010541921e-05, | |
| "loss": 1.7369, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.0754810064134188, | |
| "grad_norm": 9.221963882446289, | |
| "learning_rate": 3.218580135068358e-05, | |
| "loss": 2.0833, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.080414405525407, | |
| "grad_norm": 5.741699695587158, | |
| "learning_rate": 3.210344259594795e-05, | |
| "loss": 2.4029, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.0853478046373952, | |
| "grad_norm": 9.234859466552734, | |
| "learning_rate": 3.2021083841212324e-05, | |
| "loss": 1.5287, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.0902812037493834, | |
| "grad_norm": 7.735116004943848, | |
| "learning_rate": 3.19387250864767e-05, | |
| "loss": 1.9516, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.0952146028613714, | |
| "grad_norm": 6.103670120239258, | |
| "learning_rate": 3.1856366331741066e-05, | |
| "loss": 1.9241, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.1001480019733596, | |
| "grad_norm": 5.348116397857666, | |
| "learning_rate": 3.177400757700544e-05, | |
| "loss": 1.6599, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.1050814010853478, | |
| "grad_norm": 11.653718948364258, | |
| "learning_rate": 3.169164882226981e-05, | |
| "loss": 1.9352, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.110014800197336, | |
| "grad_norm": 8.342635154724121, | |
| "learning_rate": 3.160929006753418e-05, | |
| "loss": 1.963, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.1149481993093242, | |
| "grad_norm": 6.9188103675842285, | |
| "learning_rate": 3.152693131279855e-05, | |
| "loss": 1.9066, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.1198815984213122, | |
| "grad_norm": 7.183344841003418, | |
| "learning_rate": 3.1444572558062925e-05, | |
| "loss": 1.9223, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.1248149975333004, | |
| "grad_norm": 6.872612476348877, | |
| "learning_rate": 3.136221380332729e-05, | |
| "loss": 1.5956, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.1297483966452886, | |
| "grad_norm": 7.547828674316406, | |
| "learning_rate": 3.127985504859167e-05, | |
| "loss": 1.7239, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.1346817957572768, | |
| "grad_norm": 9.748641014099121, | |
| "learning_rate": 3.119749629385604e-05, | |
| "loss": 1.455, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.139615194869265, | |
| "grad_norm": 13.310020446777344, | |
| "learning_rate": 3.111513753912041e-05, | |
| "loss": 1.9041, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.144548593981253, | |
| "grad_norm": 8.848477363586426, | |
| "learning_rate": 3.103277878438478e-05, | |
| "loss": 2.0644, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.1494819930932412, | |
| "grad_norm": 7.570590019226074, | |
| "learning_rate": 3.095042002964915e-05, | |
| "loss": 2.1607, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.1544153922052294, | |
| "grad_norm": 7.833923816680908, | |
| "learning_rate": 3.0868061274913525e-05, | |
| "loss": 1.5976, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.1593487913172176, | |
| "grad_norm": 6.807453632354736, | |
| "learning_rate": 3.078570252017789e-05, | |
| "loss": 1.8354, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.1642821904292058, | |
| "grad_norm": 5.707053184509277, | |
| "learning_rate": 3.070334376544227e-05, | |
| "loss": 2.0815, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.169215589541194, | |
| "grad_norm": 3.920215606689453, | |
| "learning_rate": 3.0620985010706635e-05, | |
| "loss": 2.1072, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.174148988653182, | |
| "grad_norm": 8.955766677856445, | |
| "learning_rate": 3.053862625597101e-05, | |
| "loss": 1.9202, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.1790823877651702, | |
| "grad_norm": 6.6745219230651855, | |
| "learning_rate": 3.045626750123538e-05, | |
| "loss": 1.9565, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.1840157868771584, | |
| "grad_norm": 5.365530967712402, | |
| "learning_rate": 3.037390874649975e-05, | |
| "loss": 1.8332, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.1889491859891466, | |
| "grad_norm": 4.238736629486084, | |
| "learning_rate": 3.0291549991764122e-05, | |
| "loss": 1.6349, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.1938825851011348, | |
| "grad_norm": 4.686330795288086, | |
| "learning_rate": 3.02091912370285e-05, | |
| "loss": 2.0218, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.1988159842131227, | |
| "grad_norm": 10.073396682739258, | |
| "learning_rate": 3.012683248229287e-05, | |
| "loss": 2.1104, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.203749383325111, | |
| "grad_norm": 7.226747035980225, | |
| "learning_rate": 3.0044473727557242e-05, | |
| "loss": 1.7695, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.2086827824370991, | |
| "grad_norm": 6.546811103820801, | |
| "learning_rate": 2.9962114972821613e-05, | |
| "loss": 1.5873, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.2136161815490873, | |
| "grad_norm": 7.798117637634277, | |
| "learning_rate": 2.9879756218085987e-05, | |
| "loss": 1.7036, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.2185495806610755, | |
| "grad_norm": 5.489384174346924, | |
| "learning_rate": 2.979739746335036e-05, | |
| "loss": 1.9574, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.2234829797730638, | |
| "grad_norm": 11.180368423461914, | |
| "learning_rate": 2.971503870861473e-05, | |
| "loss": 1.7365, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.2284163788850517, | |
| "grad_norm": 6.837683200836182, | |
| "learning_rate": 2.96326799538791e-05, | |
| "loss": 1.8239, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.23334977799704, | |
| "grad_norm": 6.739687919616699, | |
| "learning_rate": 2.955032119914347e-05, | |
| "loss": 1.6268, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2382831771090281, | |
| "grad_norm": 6.837540149688721, | |
| "learning_rate": 2.9467962444407842e-05, | |
| "loss": 1.9372, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.2432165762210163, | |
| "grad_norm": 12.321346282958984, | |
| "learning_rate": 2.9385603689672213e-05, | |
| "loss": 1.7772, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.2481499753330045, | |
| "grad_norm": 5.676684379577637, | |
| "learning_rate": 2.9303244934936584e-05, | |
| "loss": 2.2801, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.2530833744449925, | |
| "grad_norm": 8.29293155670166, | |
| "learning_rate": 2.9220886180200955e-05, | |
| "loss": 1.9417, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.2580167735569807, | |
| "grad_norm": 10.451891899108887, | |
| "learning_rate": 2.913852742546533e-05, | |
| "loss": 1.9312, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.262950172668969, | |
| "grad_norm": 6.803496837615967, | |
| "learning_rate": 2.90561686707297e-05, | |
| "loss": 2.1992, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.2678835717809571, | |
| "grad_norm": 7.187436580657959, | |
| "learning_rate": 2.8973809915994072e-05, | |
| "loss": 1.9083, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.2728169708929453, | |
| "grad_norm": 8.493294715881348, | |
| "learning_rate": 2.8891451161258443e-05, | |
| "loss": 1.796, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.2777503700049335, | |
| "grad_norm": 10.225448608398438, | |
| "learning_rate": 2.8809092406522814e-05, | |
| "loss": 2.2556, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.2826837691169215, | |
| "grad_norm": 5.342253684997559, | |
| "learning_rate": 2.8726733651787185e-05, | |
| "loss": 1.8728, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.2876171682289097, | |
| "grad_norm": 6.041278839111328, | |
| "learning_rate": 2.8644374897051556e-05, | |
| "loss": 1.9532, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.292550567340898, | |
| "grad_norm": 6.2567853927612305, | |
| "learning_rate": 2.8562016142315927e-05, | |
| "loss": 2.0493, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.297483966452886, | |
| "grad_norm": 9.334773063659668, | |
| "learning_rate": 2.8479657387580298e-05, | |
| "loss": 1.573, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.302417365564874, | |
| "grad_norm": 13.7177734375, | |
| "learning_rate": 2.8397298632844672e-05, | |
| "loss": 2.2434, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.3073507646768623, | |
| "grad_norm": 11.203829765319824, | |
| "learning_rate": 2.8314939878109043e-05, | |
| "loss": 2.157, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.3122841637888505, | |
| "grad_norm": 7.4039483070373535, | |
| "learning_rate": 2.8232581123373414e-05, | |
| "loss": 2.0443, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.3172175629008387, | |
| "grad_norm": 4.927941799163818, | |
| "learning_rate": 2.8150222368637785e-05, | |
| "loss": 2.3465, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.322150962012827, | |
| "grad_norm": 11.956111907958984, | |
| "learning_rate": 2.8067863613902156e-05, | |
| "loss": 1.9842, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.327084361124815, | |
| "grad_norm": 5.877150535583496, | |
| "learning_rate": 2.7985504859166527e-05, | |
| "loss": 2.1889, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.3320177602368033, | |
| "grad_norm": 5.822342395782471, | |
| "learning_rate": 2.7903146104430905e-05, | |
| "loss": 1.6998, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.3369511593487913, | |
| "grad_norm": 5.760201930999756, | |
| "learning_rate": 2.7820787349695276e-05, | |
| "loss": 1.7595, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.3418845584607795, | |
| "grad_norm": 4.266313552856445, | |
| "learning_rate": 2.7738428594959647e-05, | |
| "loss": 1.4794, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.3468179575727677, | |
| "grad_norm": 9.27362060546875, | |
| "learning_rate": 2.7656069840224018e-05, | |
| "loss": 1.9813, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.3517513566847559, | |
| "grad_norm": 6.628077507019043, | |
| "learning_rate": 2.757371108548839e-05, | |
| "loss": 1.7057, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.3566847557967439, | |
| "grad_norm": 4.694432735443115, | |
| "learning_rate": 2.7491352330752764e-05, | |
| "loss": 1.8465, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.361618154908732, | |
| "grad_norm": 6.031017780303955, | |
| "learning_rate": 2.7408993576017135e-05, | |
| "loss": 1.2289, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.3665515540207203, | |
| "grad_norm": 5.344706058502197, | |
| "learning_rate": 2.7326634821281506e-05, | |
| "loss": 1.9155, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.3714849531327085, | |
| "grad_norm": 5.284564018249512, | |
| "learning_rate": 2.7244276066545877e-05, | |
| "loss": 1.7855, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.3764183522446967, | |
| "grad_norm": 4.38336181640625, | |
| "learning_rate": 2.7161917311810248e-05, | |
| "loss": 2.0217, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.3813517513566849, | |
| "grad_norm": 10.120343208312988, | |
| "learning_rate": 2.707955855707462e-05, | |
| "loss": 2.19, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.3862851504686728, | |
| "grad_norm": 6.627511024475098, | |
| "learning_rate": 2.699719980233899e-05, | |
| "loss": 1.9309, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.391218549580661, | |
| "grad_norm": 2.8276286125183105, | |
| "learning_rate": 2.691484104760336e-05, | |
| "loss": 1.8075, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.3961519486926492, | |
| "grad_norm": 10.60366153717041, | |
| "learning_rate": 2.6832482292867732e-05, | |
| "loss": 1.4447, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.4010853478046374, | |
| "grad_norm": 5.9414286613464355, | |
| "learning_rate": 2.6750123538132106e-05, | |
| "loss": 1.8544, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.4060187469166254, | |
| "grad_norm": 5.864882946014404, | |
| "learning_rate": 2.6667764783396477e-05, | |
| "loss": 2.1942, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.4109521460286136, | |
| "grad_norm": 5.399135112762451, | |
| "learning_rate": 2.6585406028660848e-05, | |
| "loss": 1.9905, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.4158855451406018, | |
| "grad_norm": 4.467281341552734, | |
| "learning_rate": 2.650304727392522e-05, | |
| "loss": 1.7024, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.42081894425259, | |
| "grad_norm": 7.034926414489746, | |
| "learning_rate": 2.642068851918959e-05, | |
| "loss": 1.8154, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.4257523433645782, | |
| "grad_norm": 4.346402168273926, | |
| "learning_rate": 2.633832976445396e-05, | |
| "loss": 1.7937, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.4306857424765664, | |
| "grad_norm": 9.743393898010254, | |
| "learning_rate": 2.6255971009718332e-05, | |
| "loss": 1.7868, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.4356191415885546, | |
| "grad_norm": 7.675935745239258, | |
| "learning_rate": 2.6173612254982703e-05, | |
| "loss": 1.8108, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.4405525407005426, | |
| "grad_norm": 5.96721076965332, | |
| "learning_rate": 2.6091253500247074e-05, | |
| "loss": 1.8259, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.4454859398125308, | |
| "grad_norm": 6.192174911499023, | |
| "learning_rate": 2.600889474551145e-05, | |
| "loss": 1.7254, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.450419338924519, | |
| "grad_norm": 7.611389636993408, | |
| "learning_rate": 2.592653599077582e-05, | |
| "loss": 2.1246, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.4553527380365072, | |
| "grad_norm": 8.755757331848145, | |
| "learning_rate": 2.584417723604019e-05, | |
| "loss": 1.2897, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.4602861371484952, | |
| "grad_norm": 17.625520706176758, | |
| "learning_rate": 2.5761818481304562e-05, | |
| "loss": 1.8039, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.4652195362604834, | |
| "grad_norm": 6.4828057289123535, | |
| "learning_rate": 2.5679459726568933e-05, | |
| "loss": 1.7734, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.4701529353724716, | |
| "grad_norm": 4.072720527648926, | |
| "learning_rate": 2.5597100971833304e-05, | |
| "loss": 2.1754, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.4750863344844598, | |
| "grad_norm": 5.953226089477539, | |
| "learning_rate": 2.551474221709768e-05, | |
| "loss": 1.7654, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.480019733596448, | |
| "grad_norm": 5.490500450134277, | |
| "learning_rate": 2.5432383462362053e-05, | |
| "loss": 1.9371, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4849531327084362, | |
| "grad_norm": 5.547089576721191, | |
| "learning_rate": 2.5350024707626424e-05, | |
| "loss": 1.8543, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 1.4898865318204244, | |
| "grad_norm": 7.270637512207031, | |
| "learning_rate": 2.5267665952890795e-05, | |
| "loss": 1.6728, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 1.4948199309324124, | |
| "grad_norm": 10.142009735107422, | |
| "learning_rate": 2.5185307198155166e-05, | |
| "loss": 1.7656, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 1.4997533300444006, | |
| "grad_norm": 11.43400764465332, | |
| "learning_rate": 2.510294844341954e-05, | |
| "loss": 1.7731, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 1.5046867291563888, | |
| "grad_norm": 8.225043296813965, | |
| "learning_rate": 2.502058968868391e-05, | |
| "loss": 2.0266, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.5096201282683768, | |
| "grad_norm": 8.20263385772705, | |
| "learning_rate": 2.4938230933948282e-05, | |
| "loss": 1.9749, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 1.514553527380365, | |
| "grad_norm": 8.855592727661133, | |
| "learning_rate": 2.4855872179212653e-05, | |
| "loss": 1.9333, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 1.5194869264923532, | |
| "grad_norm": 5.7763752937316895, | |
| "learning_rate": 2.4773513424477024e-05, | |
| "loss": 1.5184, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 1.5244203256043414, | |
| "grad_norm": 8.338578224182129, | |
| "learning_rate": 2.4691154669741395e-05, | |
| "loss": 1.9215, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 1.5293537247163296, | |
| "grad_norm": 5.367077350616455, | |
| "learning_rate": 2.4608795915005766e-05, | |
| "loss": 1.973, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.5342871238283178, | |
| "grad_norm": 9.6682710647583, | |
| "learning_rate": 2.4526437160270137e-05, | |
| "loss": 1.6508, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 1.539220522940306, | |
| "grad_norm": 5.364202499389648, | |
| "learning_rate": 2.4444078405534508e-05, | |
| "loss": 1.9087, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 1.5441539220522942, | |
| "grad_norm": 9.031702995300293, | |
| "learning_rate": 2.4361719650798882e-05, | |
| "loss": 1.759, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 1.5490873211642822, | |
| "grad_norm": 6.895035743713379, | |
| "learning_rate": 2.4279360896063253e-05, | |
| "loss": 1.7942, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 1.5540207202762704, | |
| "grad_norm": 7.573614120483398, | |
| "learning_rate": 2.4197002141327624e-05, | |
| "loss": 2.0408, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.5589541193882586, | |
| "grad_norm": 14.733696937561035, | |
| "learning_rate": 2.4114643386591996e-05, | |
| "loss": 1.748, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 1.5638875185002465, | |
| "grad_norm": 7.862844944000244, | |
| "learning_rate": 2.4032284631856367e-05, | |
| "loss": 1.8507, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 1.5688209176122347, | |
| "grad_norm": 4.0979390144348145, | |
| "learning_rate": 2.3949925877120738e-05, | |
| "loss": 2.0245, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 1.573754316724223, | |
| "grad_norm": 6.260437965393066, | |
| "learning_rate": 2.386756712238511e-05, | |
| "loss": 1.7502, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 1.5786877158362111, | |
| "grad_norm": 8.188227653503418, | |
| "learning_rate": 2.3785208367649483e-05, | |
| "loss": 1.959, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.5836211149481993, | |
| "grad_norm": 5.43758487701416, | |
| "learning_rate": 2.3702849612913854e-05, | |
| "loss": 1.7763, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 1.5885545140601876, | |
| "grad_norm": 5.603811740875244, | |
| "learning_rate": 2.3620490858178225e-05, | |
| "loss": 1.7247, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 1.5934879131721758, | |
| "grad_norm": 4.974462509155273, | |
| "learning_rate": 2.35381321034426e-05, | |
| "loss": 1.8819, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 1.598421312284164, | |
| "grad_norm": 5.835883617401123, | |
| "learning_rate": 2.345577334870697e-05, | |
| "loss": 1.9156, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 1.603354711396152, | |
| "grad_norm": 10.42984676361084, | |
| "learning_rate": 2.337341459397134e-05, | |
| "loss": 1.6488, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.6082881105081401, | |
| "grad_norm": 10.849640846252441, | |
| "learning_rate": 2.3291055839235712e-05, | |
| "loss": 1.9802, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 1.6132215096201281, | |
| "grad_norm": 6.4310784339904785, | |
| "learning_rate": 2.3208697084500083e-05, | |
| "loss": 1.9685, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 1.6181549087321163, | |
| "grad_norm": 4.533667087554932, | |
| "learning_rate": 2.3126338329764454e-05, | |
| "loss": 1.5295, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 1.6230883078441045, | |
| "grad_norm": 7.374139785766602, | |
| "learning_rate": 2.3043979575028825e-05, | |
| "loss": 1.8251, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 1.6280217069560927, | |
| "grad_norm": 7.097417831420898, | |
| "learning_rate": 2.2961620820293196e-05, | |
| "loss": 1.7943, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.632955106068081, | |
| "grad_norm": 9.811117172241211, | |
| "learning_rate": 2.2879262065557567e-05, | |
| "loss": 1.5591, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 1.6378885051800691, | |
| "grad_norm": 2.922159194946289, | |
| "learning_rate": 2.2796903310821942e-05, | |
| "loss": 1.5074, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 1.6428219042920573, | |
| "grad_norm": 5.961956977844238, | |
| "learning_rate": 2.2714544556086313e-05, | |
| "loss": 2.1771, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 1.6477553034040455, | |
| "grad_norm": 6.768845558166504, | |
| "learning_rate": 2.2632185801350687e-05, | |
| "loss": 1.7893, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 1.6526887025160335, | |
| "grad_norm": 5.665631294250488, | |
| "learning_rate": 2.2549827046615058e-05, | |
| "loss": 1.8398, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.6576221016280217, | |
| "grad_norm": 8.640176773071289, | |
| "learning_rate": 2.246746829187943e-05, | |
| "loss": 1.9559, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 1.66255550074001, | |
| "grad_norm": 6.773179531097412, | |
| "learning_rate": 2.23851095371438e-05, | |
| "loss": 1.7461, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 1.6674888998519979, | |
| "grad_norm": 5.100921630859375, | |
| "learning_rate": 2.230275078240817e-05, | |
| "loss": 1.9243, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 1.672422298963986, | |
| "grad_norm": 8.315923690795898, | |
| "learning_rate": 2.2220392027672542e-05, | |
| "loss": 1.8641, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 1.6773556980759743, | |
| "grad_norm": 6.231566429138184, | |
| "learning_rate": 2.2138033272936913e-05, | |
| "loss": 1.9068, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.6822890971879625, | |
| "grad_norm": 4.701402187347412, | |
| "learning_rate": 2.2055674518201284e-05, | |
| "loss": 1.9061, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 1.6872224962999507, | |
| "grad_norm": 5.9488325119018555, | |
| "learning_rate": 2.197331576346566e-05, | |
| "loss": 2.0304, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 1.692155895411939, | |
| "grad_norm": 9.257286071777344, | |
| "learning_rate": 2.189095700873003e-05, | |
| "loss": 1.8544, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 1.697089294523927, | |
| "grad_norm": 6.57626485824585, | |
| "learning_rate": 2.18085982539944e-05, | |
| "loss": 1.9199, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 1.7020226936359153, | |
| "grad_norm": 6.381181240081787, | |
| "learning_rate": 2.1726239499258772e-05, | |
| "loss": 2.0876, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.7069560927479033, | |
| "grad_norm": 3.832770586013794, | |
| "learning_rate": 2.1643880744523143e-05, | |
| "loss": 1.7948, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 1.7118894918598915, | |
| "grad_norm": 20.409229278564453, | |
| "learning_rate": 2.1561521989787514e-05, | |
| "loss": 1.9628, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 1.7168228909718797, | |
| "grad_norm": 12.321199417114258, | |
| "learning_rate": 2.1479163235051885e-05, | |
| "loss": 2.149, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 1.7217562900838677, | |
| "grad_norm": 7.48840856552124, | |
| "learning_rate": 2.139680448031626e-05, | |
| "loss": 1.6974, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 1.7266896891958559, | |
| "grad_norm": 10.339478492736816, | |
| "learning_rate": 2.131444572558063e-05, | |
| "loss": 2.0896, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.731623088307844, | |
| "grad_norm": 8.04315185546875, | |
| "learning_rate": 2.1232086970845e-05, | |
| "loss": 1.6795, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 1.7365564874198323, | |
| "grad_norm": 9.33163070678711, | |
| "learning_rate": 2.1149728216109376e-05, | |
| "loss": 1.7942, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 1.7414898865318205, | |
| "grad_norm": 11.18399715423584, | |
| "learning_rate": 2.1067369461373747e-05, | |
| "loss": 1.4243, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 1.7464232856438087, | |
| "grad_norm": 9.912873268127441, | |
| "learning_rate": 2.0985010706638118e-05, | |
| "loss": 1.5473, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 1.7513566847557969, | |
| "grad_norm": 8.763083457946777, | |
| "learning_rate": 2.090265195190249e-05, | |
| "loss": 1.8, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.756290083867785, | |
| "grad_norm": 8.042789459228516, | |
| "learning_rate": 2.082029319716686e-05, | |
| "loss": 1.9486, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 1.761223482979773, | |
| "grad_norm": 5.028835296630859, | |
| "learning_rate": 2.073793444243123e-05, | |
| "loss": 1.8425, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 1.7661568820917612, | |
| "grad_norm": 5.263244152069092, | |
| "learning_rate": 2.0655575687695602e-05, | |
| "loss": 2.3425, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 1.7710902812037492, | |
| "grad_norm": 6.808823585510254, | |
| "learning_rate": 2.0573216932959973e-05, | |
| "loss": 1.7344, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 1.7760236803157374, | |
| "grad_norm": 7.849328994750977, | |
| "learning_rate": 2.0490858178224344e-05, | |
| "loss": 1.9272, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.7809570794277256, | |
| "grad_norm": 5.156803131103516, | |
| "learning_rate": 2.0408499423488718e-05, | |
| "loss": 1.9647, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 1.7858904785397138, | |
| "grad_norm": 8.30761432647705, | |
| "learning_rate": 2.032614066875309e-05, | |
| "loss": 1.5823, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 1.790823877651702, | |
| "grad_norm": 9.06478214263916, | |
| "learning_rate": 2.0243781914017464e-05, | |
| "loss": 2.03, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 1.7957572767636902, | |
| "grad_norm": 13.680024147033691, | |
| "learning_rate": 2.0161423159281835e-05, | |
| "loss": 1.9991, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 1.8006906758756784, | |
| "grad_norm": 6.494818210601807, | |
| "learning_rate": 2.0079064404546206e-05, | |
| "loss": 1.8554, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.8056240749876666, | |
| "grad_norm": 7.450560092926025, | |
| "learning_rate": 1.9996705649810577e-05, | |
| "loss": 1.9436, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 1.8105574740996546, | |
| "grad_norm": 6.066117763519287, | |
| "learning_rate": 1.9914346895074948e-05, | |
| "loss": 1.7343, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 1.8154908732116428, | |
| "grad_norm": 4.272149562835693, | |
| "learning_rate": 1.983198814033932e-05, | |
| "loss": 1.8145, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 1.820424272323631, | |
| "grad_norm": 7.599696636199951, | |
| "learning_rate": 1.974962938560369e-05, | |
| "loss": 1.9128, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 1.825357671435619, | |
| "grad_norm": 8.375687599182129, | |
| "learning_rate": 1.966727063086806e-05, | |
| "loss": 1.568, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.8302910705476072, | |
| "grad_norm": 12.533012390136719, | |
| "learning_rate": 1.9584911876132435e-05, | |
| "loss": 1.7363, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 1.8352244696595954, | |
| "grad_norm": 7.711116313934326, | |
| "learning_rate": 1.9502553121396806e-05, | |
| "loss": 1.9017, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 1.8401578687715836, | |
| "grad_norm": 4.572812080383301, | |
| "learning_rate": 1.9420194366661177e-05, | |
| "loss": 1.5797, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 1.8450912678835718, | |
| "grad_norm": 4.474880695343018, | |
| "learning_rate": 1.9337835611925548e-05, | |
| "loss": 2.0278, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 1.85002466699556, | |
| "grad_norm": 5.497013092041016, | |
| "learning_rate": 1.925547685718992e-05, | |
| "loss": 1.757, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.8549580661075482, | |
| "grad_norm": 8.392544746398926, | |
| "learning_rate": 1.917311810245429e-05, | |
| "loss": 1.6158, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 1.8598914652195364, | |
| "grad_norm": 8.011329650878906, | |
| "learning_rate": 1.9090759347718665e-05, | |
| "loss": 1.8829, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 1.8648248643315244, | |
| "grad_norm": 6.119569778442383, | |
| "learning_rate": 1.9008400592983036e-05, | |
| "loss": 2.5915, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 1.8697582634435126, | |
| "grad_norm": 4.356236934661865, | |
| "learning_rate": 1.8926041838247407e-05, | |
| "loss": 1.7601, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 1.8746916625555008, | |
| "grad_norm": 3.654446601867676, | |
| "learning_rate": 1.8843683083511778e-05, | |
| "loss": 1.5618, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.8796250616674888, | |
| "grad_norm": 10.40364933013916, | |
| "learning_rate": 1.8761324328776152e-05, | |
| "loss": 2.1173, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 1.884558460779477, | |
| "grad_norm": 11.642656326293945, | |
| "learning_rate": 1.8678965574040523e-05, | |
| "loss": 1.9532, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 1.8894918598914652, | |
| "grad_norm": 7.247965335845947, | |
| "learning_rate": 1.8596606819304894e-05, | |
| "loss": 1.8573, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 1.8944252590034534, | |
| "grad_norm": 8.925628662109375, | |
| "learning_rate": 1.8514248064569265e-05, | |
| "loss": 1.6098, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 1.8993586581154416, | |
| "grad_norm": 6.515012264251709, | |
| "learning_rate": 1.8431889309833636e-05, | |
| "loss": 1.7684, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.9042920572274298, | |
| "grad_norm": 9.089420318603516, | |
| "learning_rate": 1.8349530555098007e-05, | |
| "loss": 2.1597, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 1.909225456339418, | |
| "grad_norm": 6.92453670501709, | |
| "learning_rate": 1.8267171800362378e-05, | |
| "loss": 1.8394, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 1.9141588554514062, | |
| "grad_norm": 7.409922122955322, | |
| "learning_rate": 1.818481304562675e-05, | |
| "loss": 1.8576, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 1.9190922545633942, | |
| "grad_norm": 5.864316463470459, | |
| "learning_rate": 1.810245429089112e-05, | |
| "loss": 1.7566, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 1.9240256536753824, | |
| "grad_norm": 7.16778039932251, | |
| "learning_rate": 1.8020095536155494e-05, | |
| "loss": 1.7765, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.9289590527873703, | |
| "grad_norm": 7.447871208190918, | |
| "learning_rate": 1.7937736781419865e-05, | |
| "loss": 1.5466, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 1.9338924518993585, | |
| "grad_norm": 8.456513404846191, | |
| "learning_rate": 1.785537802668424e-05, | |
| "loss": 1.6053, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.9388258510113467, | |
| "grad_norm": 5.673243045806885, | |
| "learning_rate": 1.777301927194861e-05, | |
| "loss": 1.6813, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 1.943759250123335, | |
| "grad_norm": 13.05078125, | |
| "learning_rate": 1.7690660517212982e-05, | |
| "loss": 1.8932, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 1.9486926492353231, | |
| "grad_norm": 5.095945358276367, | |
| "learning_rate": 1.7608301762477353e-05, | |
| "loss": 1.9467, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.9536260483473114, | |
| "grad_norm": 7.766141414642334, | |
| "learning_rate": 1.7525943007741724e-05, | |
| "loss": 1.9318, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 1.9585594474592996, | |
| "grad_norm": 6.2664690017700195, | |
| "learning_rate": 1.7443584253006095e-05, | |
| "loss": 1.5792, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 1.9634928465712878, | |
| "grad_norm": 9.3766508102417, | |
| "learning_rate": 1.7361225498270466e-05, | |
| "loss": 1.7359, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 1.9684262456832757, | |
| "grad_norm": 8.860836029052734, | |
| "learning_rate": 1.7278866743534837e-05, | |
| "loss": 1.7544, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 1.973359644795264, | |
| "grad_norm": 6.054210662841797, | |
| "learning_rate": 1.719650798879921e-05, | |
| "loss": 1.5618, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.9782930439072521, | |
| "grad_norm": 11.120077133178711, | |
| "learning_rate": 1.7114149234063582e-05, | |
| "loss": 1.8174, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 1.9832264430192401, | |
| "grad_norm": 6.720516681671143, | |
| "learning_rate": 1.7031790479327953e-05, | |
| "loss": 1.613, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 1.9881598421312283, | |
| "grad_norm": 6.063734531402588, | |
| "learning_rate": 1.6949431724592324e-05, | |
| "loss": 1.9363, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 1.9930932412432165, | |
| "grad_norm": 5.5154523849487305, | |
| "learning_rate": 1.6867072969856695e-05, | |
| "loss": 1.6523, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 1.9980266403552047, | |
| "grad_norm": 10.039024353027344, | |
| "learning_rate": 1.6784714215121066e-05, | |
| "loss": 1.8938, | |
| "step": 4050 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 6081, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5059810462466048e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |