| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.7336585365853656, | |
| "eval_steps": 500, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01951219512195122, | |
| "grad_norm": 0.6663638353347778, | |
| "learning_rate": 0.00019934683213585893, | |
| "loss": 0.5255, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03902439024390244, | |
| "grad_norm": 0.7442628741264343, | |
| "learning_rate": 0.00019804049640757677, | |
| "loss": 0.3367, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05853658536585366, | |
| "grad_norm": 0.504467785358429, | |
| "learning_rate": 0.0001967341606792946, | |
| "loss": 0.3368, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07804878048780488, | |
| "grad_norm": 0.339693546295166, | |
| "learning_rate": 0.00019542782495101242, | |
| "loss": 0.3064, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 0.267976313829422, | |
| "learning_rate": 0.00019412148922273026, | |
| "loss": 0.3108, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11707317073170732, | |
| "grad_norm": 0.48640233278274536, | |
| "learning_rate": 0.00019281515349444807, | |
| "loss": 0.3069, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13658536585365855, | |
| "grad_norm": 0.7269986867904663, | |
| "learning_rate": 0.0001915088177661659, | |
| "loss": 0.3134, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15609756097560976, | |
| "grad_norm": 0.31376832723617554, | |
| "learning_rate": 0.00019020248203788375, | |
| "loss": 0.2915, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17560975609756097, | |
| "grad_norm": 0.6166387796401978, | |
| "learning_rate": 0.00018889614630960156, | |
| "loss": 0.2862, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 0.6621638536453247, | |
| "learning_rate": 0.0001875898105813194, | |
| "loss": 0.263, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2146341463414634, | |
| "grad_norm": 0.2815336287021637, | |
| "learning_rate": 0.00018628347485303724, | |
| "loss": 0.2747, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.23414634146341465, | |
| "grad_norm": 0.5862469673156738, | |
| "learning_rate": 0.00018497713912475508, | |
| "loss": 0.2833, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.25365853658536586, | |
| "grad_norm": 0.5362260937690735, | |
| "learning_rate": 0.00018367080339647292, | |
| "loss": 0.2613, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2731707317073171, | |
| "grad_norm": 0.7799074053764343, | |
| "learning_rate": 0.00018236446766819073, | |
| "loss": 0.2535, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 0.8866592645645142, | |
| "learning_rate": 0.00018105813193990857, | |
| "loss": 0.2603, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3121951219512195, | |
| "grad_norm": 0.9003716707229614, | |
| "learning_rate": 0.00017975179621162638, | |
| "loss": 0.27, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.33170731707317075, | |
| "grad_norm": 0.5946381092071533, | |
| "learning_rate": 0.00017844546048334422, | |
| "loss": 0.2572, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.35121951219512193, | |
| "grad_norm": 0.8860711455345154, | |
| "learning_rate": 0.00017713912475506206, | |
| "loss": 0.2839, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.37073170731707317, | |
| "grad_norm": 0.8693526983261108, | |
| "learning_rate": 0.0001758327890267799, | |
| "loss": 0.2477, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3902439024390244, | |
| "grad_norm": 0.9044304490089417, | |
| "learning_rate": 0.00017452645329849774, | |
| "loss": 0.2674, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4097560975609756, | |
| "grad_norm": 0.5563161969184875, | |
| "learning_rate": 0.00017322011757021555, | |
| "loss": 0.2436, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4292682926829268, | |
| "grad_norm": 1.1451846361160278, | |
| "learning_rate": 0.0001719137818419334, | |
| "loss": 0.25, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.44878048780487806, | |
| "grad_norm": 0.8895041942596436, | |
| "learning_rate": 0.00017060744611365123, | |
| "loss": 0.2542, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4682926829268293, | |
| "grad_norm": 0.8991382718086243, | |
| "learning_rate": 0.00016930111038536904, | |
| "loss": 0.2523, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 1.0106490850448608, | |
| "learning_rate": 0.00016799477465708688, | |
| "loss": 0.2554, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5073170731707317, | |
| "grad_norm": 0.5570860505104065, | |
| "learning_rate": 0.0001666884389288047, | |
| "loss": 0.2431, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.526829268292683, | |
| "grad_norm": 1.1715517044067383, | |
| "learning_rate": 0.00016538210320052253, | |
| "loss": 0.2383, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5463414634146342, | |
| "grad_norm": 0.9527117609977722, | |
| "learning_rate": 0.00016407576747224037, | |
| "loss": 0.2222, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5658536585365853, | |
| "grad_norm": 1.012949824333191, | |
| "learning_rate": 0.0001627694317439582, | |
| "loss": 0.2743, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.157406210899353, | |
| "learning_rate": 0.00016146309601567605, | |
| "loss": 0.261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6048780487804878, | |
| "grad_norm": 1.044209599494934, | |
| "learning_rate": 0.00016015676028739386, | |
| "loss": 0.2346, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.624390243902439, | |
| "grad_norm": 0.8777090311050415, | |
| "learning_rate": 0.0001588504245591117, | |
| "loss": 0.2734, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6439024390243903, | |
| "grad_norm": 1.159125804901123, | |
| "learning_rate": 0.00015754408883082954, | |
| "loss": 0.2655, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6634146341463415, | |
| "grad_norm": 0.8555986881256104, | |
| "learning_rate": 0.00015623775310254735, | |
| "loss": 0.2584, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6829268292682927, | |
| "grad_norm": 0.4655376374721527, | |
| "learning_rate": 0.0001549314173742652, | |
| "loss": 0.2553, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7024390243902439, | |
| "grad_norm": 0.5464235544204712, | |
| "learning_rate": 0.000153625081645983, | |
| "loss": 0.2493, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7219512195121951, | |
| "grad_norm": 0.5507020354270935, | |
| "learning_rate": 0.00015231874591770084, | |
| "loss": 0.2485, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7414634146341463, | |
| "grad_norm": 0.47853443026542664, | |
| "learning_rate": 0.00015101241018941868, | |
| "loss": 0.2348, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7609756097560976, | |
| "grad_norm": 0.22285830974578857, | |
| "learning_rate": 0.00014970607446113652, | |
| "loss": 0.2393, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7804878048780488, | |
| "grad_norm": 1.4561814069747925, | |
| "learning_rate": 0.00014839973873285436, | |
| "loss": 0.2432, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.6247652769088745, | |
| "learning_rate": 0.00014709340300457217, | |
| "loss": 0.242, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8195121951219512, | |
| "grad_norm": 1.0940589904785156, | |
| "learning_rate": 0.00014578706727629, | |
| "loss": 0.2632, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8390243902439024, | |
| "grad_norm": 1.2155462503433228, | |
| "learning_rate": 0.00014448073154800785, | |
| "loss": 0.2432, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8585365853658536, | |
| "grad_norm": 0.3646700978279114, | |
| "learning_rate": 0.00014317439581972566, | |
| "loss": 0.2342, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8780487804878049, | |
| "grad_norm": 0.3318726122379303, | |
| "learning_rate": 0.0001418680600914435, | |
| "loss": 0.2454, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8975609756097561, | |
| "grad_norm": 1.2063515186309814, | |
| "learning_rate": 0.00014056172436316132, | |
| "loss": 0.2475, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9170731707317074, | |
| "grad_norm": 0.7661004066467285, | |
| "learning_rate": 0.00013925538863487918, | |
| "loss": 0.2635, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "grad_norm": 0.2922651469707489, | |
| "learning_rate": 0.00013794905290659702, | |
| "loss": 0.2235, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9560975609756097, | |
| "grad_norm": 1.2844675779342651, | |
| "learning_rate": 0.00013664271717831483, | |
| "loss": 0.2294, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 0.38282549381256104, | |
| "learning_rate": 0.00013533638145003267, | |
| "loss": 0.2265, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9951219512195122, | |
| "grad_norm": 0.286447674036026, | |
| "learning_rate": 0.00013403004572175048, | |
| "loss": 0.2237, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.015609756097561, | |
| "grad_norm": 0.2928523123264313, | |
| "learning_rate": 0.00013272370999346832, | |
| "loss": 0.2358, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0351219512195122, | |
| "grad_norm": 1.7477126121520996, | |
| "learning_rate": 0.00013141737426518616, | |
| "loss": 0.194, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0546341463414635, | |
| "grad_norm": 1.4766732454299927, | |
| "learning_rate": 0.00013011103853690398, | |
| "loss": 0.1828, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0741463414634147, | |
| "grad_norm": 1.7536602020263672, | |
| "learning_rate": 0.00012880470280862181, | |
| "loss": 0.1936, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.093658536585366, | |
| "grad_norm": 1.5118074417114258, | |
| "learning_rate": 0.00012749836708033965, | |
| "loss": 0.1757, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1131707317073172, | |
| "grad_norm": 0.5326169729232788, | |
| "learning_rate": 0.0001261920313520575, | |
| "loss": 0.1982, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.1326829268292684, | |
| "grad_norm": 1.1812074184417725, | |
| "learning_rate": 0.00012488569562377533, | |
| "loss": 0.2055, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.1521951219512194, | |
| "grad_norm": 1.6071784496307373, | |
| "learning_rate": 0.00012357935989549314, | |
| "loss": 0.19, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1717073170731707, | |
| "grad_norm": 1.3864187002182007, | |
| "learning_rate": 0.00012227302416721098, | |
| "loss": 0.2059, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.191219512195122, | |
| "grad_norm": 0.2694201171398163, | |
| "learning_rate": 0.00012096668843892882, | |
| "loss": 0.1966, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2107317073170731, | |
| "grad_norm": 0.7169620990753174, | |
| "learning_rate": 0.00011966035271064664, | |
| "loss": 0.186, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2302439024390244, | |
| "grad_norm": 1.2233684062957764, | |
| "learning_rate": 0.00011835401698236447, | |
| "loss": 0.1971, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2497560975609756, | |
| "grad_norm": 0.7872809767723083, | |
| "learning_rate": 0.0001170476812540823, | |
| "loss": 0.1874, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2692682926829268, | |
| "grad_norm": 1.3358359336853027, | |
| "learning_rate": 0.00011574134552580014, | |
| "loss": 0.2072, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.288780487804878, | |
| "grad_norm": 1.397544264793396, | |
| "learning_rate": 0.00011443500979751798, | |
| "loss": 0.1797, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3082926829268293, | |
| "grad_norm": 0.4923507571220398, | |
| "learning_rate": 0.00011312867406923579, | |
| "loss": 0.1791, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3278048780487806, | |
| "grad_norm": 0.9196809530258179, | |
| "learning_rate": 0.00011182233834095363, | |
| "loss": 0.1931, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3473170731707316, | |
| "grad_norm": 1.4173572063446045, | |
| "learning_rate": 0.00011051600261267146, | |
| "loss": 0.1958, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3668292682926828, | |
| "grad_norm": 0.23955851793289185, | |
| "learning_rate": 0.0001092096668843893, | |
| "loss": 0.2053, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.386341463414634, | |
| "grad_norm": 1.5932824611663818, | |
| "learning_rate": 0.00010790333115610713, | |
| "loss": 0.1861, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4058536585365853, | |
| "grad_norm": 1.292710304260254, | |
| "learning_rate": 0.00010659699542782495, | |
| "loss": 0.1949, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4253658536585365, | |
| "grad_norm": 1.5422075986862183, | |
| "learning_rate": 0.0001052906596995428, | |
| "loss": 0.1902, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4448780487804878, | |
| "grad_norm": 1.8156495094299316, | |
| "learning_rate": 0.00010398432397126061, | |
| "loss": 0.1984, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.464390243902439, | |
| "grad_norm": 0.6971287727355957, | |
| "learning_rate": 0.00010267798824297845, | |
| "loss": 0.1865, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4839024390243902, | |
| "grad_norm": 1.6063241958618164, | |
| "learning_rate": 0.00010137165251469629, | |
| "loss": 0.1771, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5034146341463415, | |
| "grad_norm": 1.5314477682113647, | |
| "learning_rate": 0.00010006531678641412, | |
| "loss": 0.2034, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.5229268292682927, | |
| "grad_norm": 1.4807705879211426, | |
| "learning_rate": 9.875898105813195e-05, | |
| "loss": 0.1853, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.542439024390244, | |
| "grad_norm": 1.519795536994934, | |
| "learning_rate": 9.745264532984978e-05, | |
| "loss": 0.1994, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.5619512195121952, | |
| "grad_norm": 1.4948278665542603, | |
| "learning_rate": 9.61463096015676e-05, | |
| "loss": 0.1826, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5814634146341464, | |
| "grad_norm": 1.2426303625106812, | |
| "learning_rate": 9.483997387328543e-05, | |
| "loss": 0.1968, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6009756097560977, | |
| "grad_norm": 1.4434090852737427, | |
| "learning_rate": 9.353363814500327e-05, | |
| "loss": 0.2016, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6204878048780489, | |
| "grad_norm": 1.3460036516189575, | |
| "learning_rate": 9.222730241672111e-05, | |
| "loss": 0.1853, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 1.008080005645752, | |
| "learning_rate": 9.092096668843894e-05, | |
| "loss": 0.2051, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6595121951219514, | |
| "grad_norm": 0.591911792755127, | |
| "learning_rate": 8.961463096015676e-05, | |
| "loss": 0.2003, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6790243902439026, | |
| "grad_norm": 0.8856975436210632, | |
| "learning_rate": 8.830829523187459e-05, | |
| "loss": 0.1994, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.6985365853658536, | |
| "grad_norm": 0.24821911752223969, | |
| "learning_rate": 8.700195950359243e-05, | |
| "loss": 0.1801, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7180487804878048, | |
| "grad_norm": 0.5248575806617737, | |
| "learning_rate": 8.569562377531027e-05, | |
| "loss": 0.1815, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.737560975609756, | |
| "grad_norm": 0.600226104259491, | |
| "learning_rate": 8.438928804702809e-05, | |
| "loss": 0.1885, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7570731707317073, | |
| "grad_norm": 0.720282793045044, | |
| "learning_rate": 8.308295231874592e-05, | |
| "loss": 0.1956, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7765853658536586, | |
| "grad_norm": 0.780463695526123, | |
| "learning_rate": 8.177661659046376e-05, | |
| "loss": 0.1892, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.7960975609756098, | |
| "grad_norm": 0.5687152147293091, | |
| "learning_rate": 8.047028086218158e-05, | |
| "loss": 0.1905, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.8156097560975608, | |
| "grad_norm": 0.33384960889816284, | |
| "learning_rate": 7.916394513389942e-05, | |
| "loss": 0.1854, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.835121951219512, | |
| "grad_norm": 0.6061075329780579, | |
| "learning_rate": 7.785760940561725e-05, | |
| "loss": 0.1808, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.8546341463414633, | |
| "grad_norm": 0.2853333353996277, | |
| "learning_rate": 7.655127367733507e-05, | |
| "loss": 0.1873, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8741463414634145, | |
| "grad_norm": 0.5911210775375366, | |
| "learning_rate": 7.524493794905291e-05, | |
| "loss": 0.1965, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.8936585365853658, | |
| "grad_norm": 0.8281512260437012, | |
| "learning_rate": 7.393860222077074e-05, | |
| "loss": 0.2078, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.913170731707317, | |
| "grad_norm": 1.2774325609207153, | |
| "learning_rate": 7.263226649248858e-05, | |
| "loss": 0.1943, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.9326829268292682, | |
| "grad_norm": 0.544083833694458, | |
| "learning_rate": 7.13259307642064e-05, | |
| "loss": 0.1892, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.9521951219512195, | |
| "grad_norm": 0.6588655114173889, | |
| "learning_rate": 7.001959503592423e-05, | |
| "loss": 0.1873, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9717073170731707, | |
| "grad_norm": 0.6782810688018799, | |
| "learning_rate": 6.871325930764207e-05, | |
| "loss": 0.2181, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.991219512195122, | |
| "grad_norm": 0.27836665511131287, | |
| "learning_rate": 6.740692357935991e-05, | |
| "loss": 0.1895, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.0117073170731707, | |
| "grad_norm": 1.3147952556610107, | |
| "learning_rate": 6.610058785107773e-05, | |
| "loss": 0.2023, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.031219512195122, | |
| "grad_norm": 0.32955631613731384, | |
| "learning_rate": 6.479425212279556e-05, | |
| "loss": 0.1539, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.050731707317073, | |
| "grad_norm": 0.7191469073295593, | |
| "learning_rate": 6.34879163945134e-05, | |
| "loss": 0.1522, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.0702439024390245, | |
| "grad_norm": 0.9275949001312256, | |
| "learning_rate": 6.218158066623122e-05, | |
| "loss": 0.1481, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.0897560975609757, | |
| "grad_norm": 0.37640875577926636, | |
| "learning_rate": 6.087524493794906e-05, | |
| "loss": 0.1425, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.109268292682927, | |
| "grad_norm": 1.2650576829910278, | |
| "learning_rate": 5.956890920966689e-05, | |
| "loss": 0.149, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.128780487804878, | |
| "grad_norm": 1.5103838443756104, | |
| "learning_rate": 5.826257348138472e-05, | |
| "loss": 0.1566, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.1482926829268294, | |
| "grad_norm": 0.9958144426345825, | |
| "learning_rate": 5.6956237753102546e-05, | |
| "loss": 0.1519, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1678048780487806, | |
| "grad_norm": 0.3257206082344055, | |
| "learning_rate": 5.564990202482038e-05, | |
| "loss": 0.1691, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.187317073170732, | |
| "grad_norm": 0.6846854090690613, | |
| "learning_rate": 5.434356629653822e-05, | |
| "loss": 0.1466, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.206829268292683, | |
| "grad_norm": 0.8696099519729614, | |
| "learning_rate": 5.3037230568256044e-05, | |
| "loss": 0.1593, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.2263414634146343, | |
| "grad_norm": 0.20721249282360077, | |
| "learning_rate": 5.1730894839973876e-05, | |
| "loss": 0.1475, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.2458536585365856, | |
| "grad_norm": 0.5807740092277527, | |
| "learning_rate": 5.04245591116917e-05, | |
| "loss": 0.1525, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.265365853658537, | |
| "grad_norm": 0.6303423643112183, | |
| "learning_rate": 4.911822338340954e-05, | |
| "loss": 0.1414, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.2848780487804876, | |
| "grad_norm": 0.6723970174789429, | |
| "learning_rate": 4.781188765512737e-05, | |
| "loss": 0.1286, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.304390243902439, | |
| "grad_norm": 0.5512611269950867, | |
| "learning_rate": 4.6505551926845206e-05, | |
| "loss": 0.137, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.32390243902439, | |
| "grad_norm": 0.32534515857696533, | |
| "learning_rate": 4.519921619856303e-05, | |
| "loss": 0.1509, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.3434146341463413, | |
| "grad_norm": 0.37095433473587036, | |
| "learning_rate": 4.3892880470280864e-05, | |
| "loss": 0.1447, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.3629268292682926, | |
| "grad_norm": 0.35774731636047363, | |
| "learning_rate": 4.25865447419987e-05, | |
| "loss": 0.1455, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.382439024390244, | |
| "grad_norm": 0.6414694786071777, | |
| "learning_rate": 4.128020901371652e-05, | |
| "loss": 0.1445, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.401951219512195, | |
| "grad_norm": 0.39603760838508606, | |
| "learning_rate": 3.997387328543436e-05, | |
| "loss": 0.142, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.4214634146341463, | |
| "grad_norm": 0.4492717981338501, | |
| "learning_rate": 3.866753755715219e-05, | |
| "loss": 0.132, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.4409756097560975, | |
| "grad_norm": 0.455432265996933, | |
| "learning_rate": 3.736120182887003e-05, | |
| "loss": 0.1483, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.4604878048780487, | |
| "grad_norm": 0.34108078479766846, | |
| "learning_rate": 3.605486610058785e-05, | |
| "loss": 0.1353, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.8474462628364563, | |
| "learning_rate": 3.4748530372305685e-05, | |
| "loss": 0.1433, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.499512195121951, | |
| "grad_norm": 0.490139365196228, | |
| "learning_rate": 3.344219464402352e-05, | |
| "loss": 0.1405, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.5190243902439025, | |
| "grad_norm": 0.2619794011116028, | |
| "learning_rate": 3.213585891574134e-05, | |
| "loss": 0.1431, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.5385365853658537, | |
| "grad_norm": 0.35025978088378906, | |
| "learning_rate": 3.082952318745918e-05, | |
| "loss": 0.1314, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.558048780487805, | |
| "grad_norm": 0.6071056127548218, | |
| "learning_rate": 2.952318745917701e-05, | |
| "loss": 0.1335, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.577560975609756, | |
| "grad_norm": 0.410426527261734, | |
| "learning_rate": 2.821685173089484e-05, | |
| "loss": 0.1466, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.5970731707317074, | |
| "grad_norm": 0.4310607612133026, | |
| "learning_rate": 2.6910516002612673e-05, | |
| "loss": 0.1452, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.6165853658536586, | |
| "grad_norm": 0.48974138498306274, | |
| "learning_rate": 2.5604180274330502e-05, | |
| "loss": 0.1488, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.63609756097561, | |
| "grad_norm": 0.4195750653743744, | |
| "learning_rate": 2.4297844546048334e-05, | |
| "loss": 0.146, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.655609756097561, | |
| "grad_norm": 0.27998435497283936, | |
| "learning_rate": 2.2991508817766167e-05, | |
| "loss": 0.1335, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.6751219512195124, | |
| "grad_norm": 0.5254719853401184, | |
| "learning_rate": 2.1685173089484e-05, | |
| "loss": 0.1317, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.694634146341463, | |
| "grad_norm": 0.3470640480518341, | |
| "learning_rate": 2.0378837361201832e-05, | |
| "loss": 0.1584, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.7141463414634144, | |
| "grad_norm": 0.29102063179016113, | |
| "learning_rate": 1.907250163291966e-05, | |
| "loss": 0.1354, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.7336585365853656, | |
| "grad_norm": 0.2697664201259613, | |
| "learning_rate": 1.7766165904637493e-05, | |
| "loss": 0.1379, | |
| "step": 1400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1536, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.6443635349501133e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |