| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 0.4270687699317932, | |
| "learning_rate": 0.00019999602855426865, | |
| "loss": 1.013, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 0.4152718782424927, | |
| "learning_rate": 0.00019998411453252217, | |
| "loss": 0.8289, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 0.7277560234069824, | |
| "learning_rate": 0.0001999642588810784, | |
| "loss": 0.5959, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 0.5505673885345459, | |
| "learning_rate": 0.00019993646317705016, | |
| "loss": 0.459, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 0.528052031993866, | |
| "learning_rate": 0.00019990072962822007, | |
| "loss": 0.3775, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 0.6307681202888489, | |
| "learning_rate": 0.00019985706107286514, | |
| "loss": 0.3285, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2978723404255319, | |
| "grad_norm": 0.6954013109207153, | |
| "learning_rate": 0.00019980546097953132, | |
| "loss": 0.2855, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 0.6790465116500854, | |
| "learning_rate": 0.000199745933446758, | |
| "loss": 0.2782, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.3829787234042553, | |
| "grad_norm": 1.324937105178833, | |
| "learning_rate": 0.0001996784832027525, | |
| "loss": 0.2635, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.8779314756393433, | |
| "learning_rate": 0.00019960311560501454, | |
| "loss": 0.1861, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.46808510638297873, | |
| "grad_norm": 0.693745493888855, | |
| "learning_rate": 0.00019951983663991056, | |
| "loss": 0.2001, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 1.0649502277374268, | |
| "learning_rate": 0.00019942865292219838, | |
| "loss": 0.1378, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5531914893617021, | |
| "grad_norm": 0.6962260007858276, | |
| "learning_rate": 0.0001993295716945017, | |
| "loss": 0.1579, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 0.7934479713439941, | |
| "learning_rate": 0.00019922260082673497, | |
| "loss": 0.092, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 1.1331907510757446, | |
| "learning_rate": 0.000199107748815478, | |
| "loss": 0.1208, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 1.3689247369766235, | |
| "learning_rate": 0.00019898502478330152, | |
| "loss": 0.0874, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.723404255319149, | |
| "grad_norm": 0.5304535031318665, | |
| "learning_rate": 0.00019885443847804211, | |
| "loss": 0.0881, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 0.6805845499038696, | |
| "learning_rate": 0.0001987160002720283, | |
| "loss": 0.0584, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8085106382978723, | |
| "grad_norm": 0.2527499198913574, | |
| "learning_rate": 0.00019856972116125653, | |
| "loss": 0.08, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.799462616443634, | |
| "learning_rate": 0.0001984156127645178, | |
| "loss": 0.0556, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8936170212765957, | |
| "grad_norm": 0.936975359916687, | |
| "learning_rate": 0.0001982536873224748, | |
| "loss": 0.0945, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 0.8067993521690369, | |
| "learning_rate": 0.00019808395769668963, | |
| "loss": 0.0495, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9787234042553191, | |
| "grad_norm": 0.45767834782600403, | |
| "learning_rate": 0.00019790643736860227, | |
| "loss": 0.0617, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 0.9198794364929199, | |
| "learning_rate": 0.00019772114043845965, | |
| "loss": 0.0467, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.7327796816825867, | |
| "learning_rate": 0.0001975280816241959, | |
| "loss": 0.0391, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.8003076910972595, | |
| "learning_rate": 0.00019732727626026305, | |
| "loss": 0.0428, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.148936170212766, | |
| "grad_norm": 0.10251367837190628, | |
| "learning_rate": 0.0001971187402964132, | |
| "loss": 0.032, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 0.45093855261802673, | |
| "learning_rate": 0.00019690249029643162, | |
| "loss": 0.0673, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2340425531914894, | |
| "grad_norm": 0.4845767915248871, | |
| "learning_rate": 0.0001966785434368211, | |
| "loss": 0.033, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.31195056438446045, | |
| "learning_rate": 0.00019644691750543767, | |
| "loss": 0.0261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3191489361702127, | |
| "grad_norm": 0.14839951694011688, | |
| "learning_rate": 0.00019620763090007762, | |
| "loss": 0.0298, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 0.20573872327804565, | |
| "learning_rate": 0.00019596070262701626, | |
| "loss": 0.0155, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.4042553191489362, | |
| "grad_norm": 0.47702595591545105, | |
| "learning_rate": 0.00019570615229949842, | |
| "loss": 0.0369, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 0.7073186039924622, | |
| "learning_rate": 0.00019544400013618023, | |
| "loss": 0.0302, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.1539478451013565, | |
| "learning_rate": 0.00019517426695952358, | |
| "loss": 0.0223, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 0.5202814340591431, | |
| "learning_rate": 0.00019489697419414182, | |
| "loss": 0.0263, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.574468085106383, | |
| "grad_norm": 0.968192458152771, | |
| "learning_rate": 0.00019461214386509842, | |
| "loss": 0.044, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 0.5662522912025452, | |
| "learning_rate": 0.00019431979859615726, | |
| "loss": 0.0421, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.6595744680851063, | |
| "grad_norm": 0.42925137281417847, | |
| "learning_rate": 0.00019401996160798573, | |
| "loss": 0.0606, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.5803830027580261, | |
| "learning_rate": 0.00019371265671631037, | |
| "loss": 0.0392, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7446808510638299, | |
| "grad_norm": 0.4235450327396393, | |
| "learning_rate": 0.00019339790833002515, | |
| "loss": 0.0286, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 0.519207775592804, | |
| "learning_rate": 0.00019307574144925287, | |
| "loss": 0.0522, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.8297872340425532, | |
| "grad_norm": 0.2344844490289688, | |
| "learning_rate": 0.00019274618166335912, | |
| "loss": 0.0281, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 0.1990007758140564, | |
| "learning_rate": 0.00019240925514892, | |
| "loss": 0.0229, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.10929415374994278, | |
| "learning_rate": 0.00019206498866764288, | |
| "loss": 0.0258, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 0.4308103024959564, | |
| "learning_rate": 0.00019171340956424074, | |
| "loss": 0.0167, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.46525439620018005, | |
| "learning_rate": 0.0001913545457642601, | |
| "loss": 0.0283, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.28837406635284424, | |
| "learning_rate": 0.00019098842577186314, | |
| "loss": 0.0137, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.0851063829787235, | |
| "grad_norm": 0.08533861488103867, | |
| "learning_rate": 0.00019061507866756347, | |
| "loss": 0.0182, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.3499375581741333, | |
| "learning_rate": 0.00019023453410591635, | |
| "loss": 0.0221, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.1702127659574466, | |
| "grad_norm": 0.07716694474220276, | |
| "learning_rate": 0.00018984682231316333, | |
| "loss": 0.0075, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 0.3093757927417755, | |
| "learning_rate": 0.00018945197408483123, | |
| "loss": 0.0133, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.25531914893617, | |
| "grad_norm": 0.13492655754089355, | |
| "learning_rate": 0.00018905002078328632, | |
| "loss": 0.0184, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 0.07833054661750793, | |
| "learning_rate": 0.000188640994335243, | |
| "loss": 0.0109, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.08865915983915329, | |
| "learning_rate": 0.0001882249272292282, | |
| "loss": 0.0121, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 0.31314581632614136, | |
| "learning_rate": 0.00018780185251300046, | |
| "loss": 0.0242, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.425531914893617, | |
| "grad_norm": 0.10387410968542099, | |
| "learning_rate": 0.00018737180379092537, | |
| "loss": 0.0285, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 0.11755700409412384, | |
| "learning_rate": 0.0001869348152213061, | |
| "loss": 0.0281, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.5106382978723403, | |
| "grad_norm": 0.1438552439212799, | |
| "learning_rate": 0.0001864909215136705, | |
| "loss": 0.0216, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.08940370380878448, | |
| "learning_rate": 0.00018604015792601396, | |
| "loss": 0.0226, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.595744680851064, | |
| "grad_norm": 0.19193390011787415, | |
| "learning_rate": 0.00018558256026199896, | |
| "loss": 0.0184, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 0.4586654603481293, | |
| "learning_rate": 0.00018511816486811134, | |
| "loss": 0.0154, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.6808510638297873, | |
| "grad_norm": 0.05811255797743797, | |
| "learning_rate": 0.00018464700863077312, | |
| "loss": 0.0238, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 0.16280680894851685, | |
| "learning_rate": 0.00018416912897341295, | |
| "loss": 0.0253, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.07031189650297165, | |
| "learning_rate": 0.00018368456385349334, | |
| "loss": 0.0146, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 0.07851342856884003, | |
| "learning_rate": 0.0001831933517594957, | |
| "loss": 0.0301, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.851063829787234, | |
| "grad_norm": 0.11461298912763596, | |
| "learning_rate": 0.0001826955317078636, | |
| "loss": 0.0155, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 0.1839868575334549, | |
| "learning_rate": 0.00018219114323990345, | |
| "loss": 0.0099, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.9361702127659575, | |
| "grad_norm": 0.04199373722076416, | |
| "learning_rate": 0.00018168022641864377, | |
| "loss": 0.0192, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.3203773498535156, | |
| "learning_rate": 0.00018116282182565311, | |
| "loss": 0.0272, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.021276595744681, | |
| "grad_norm": 0.327921599149704, | |
| "learning_rate": 0.0001806389705578168, | |
| "loss": 0.0081, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.0638297872340425, | |
| "grad_norm": 0.29155433177948, | |
| "learning_rate": 0.00018010871422407236, | |
| "loss": 0.0128, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.106382978723404, | |
| "grad_norm": 0.06926452368497849, | |
| "learning_rate": 0.00017957209494210493, | |
| "loss": 0.0171, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.148936170212766, | |
| "grad_norm": 0.08231089264154434, | |
| "learning_rate": 0.0001790291553350016, | |
| "loss": 0.0098, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.1914893617021276, | |
| "grad_norm": 0.14707215130329132, | |
| "learning_rate": 0.0001784799385278661, | |
| "loss": 0.0092, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2340425531914896, | |
| "grad_norm": 0.2735896706581116, | |
| "learning_rate": 0.00017792448814439333, | |
| "loss": 0.0115, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.276595744680851, | |
| "grad_norm": 0.44960370659828186, | |
| "learning_rate": 0.00017736284830340436, | |
| "loss": 0.0195, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.3191489361702127, | |
| "grad_norm": 0.13413724303245544, | |
| "learning_rate": 0.00017679506361534215, | |
| "loss": 0.0187, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.3617021276595747, | |
| "grad_norm": 0.24698686599731445, | |
| "learning_rate": 0.00017622117917872823, | |
| "loss": 0.0125, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.404255319148936, | |
| "grad_norm": 0.48694342374801636, | |
| "learning_rate": 0.00017564124057658056, | |
| "loss": 0.0234, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.4468085106382977, | |
| "grad_norm": 0.6931429505348206, | |
| "learning_rate": 0.00017505529387279277, | |
| "loss": 0.0234, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.4893617021276597, | |
| "grad_norm": 0.13700473308563232, | |
| "learning_rate": 0.00017446338560847568, | |
| "loss": 0.0145, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.5319148936170213, | |
| "grad_norm": 0.3254775404930115, | |
| "learning_rate": 0.00017386556279826021, | |
| "loss": 0.0179, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.574468085106383, | |
| "grad_norm": 0.368379682302475, | |
| "learning_rate": 0.00017326187292656333, | |
| "loss": 0.013, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.617021276595745, | |
| "grad_norm": 0.2745888829231262, | |
| "learning_rate": 0.00017265236394381633, | |
| "loss": 0.0136, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.6595744680851063, | |
| "grad_norm": 0.0781714916229248, | |
| "learning_rate": 0.00017203708426265614, | |
| "loss": 0.0126, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.702127659574468, | |
| "grad_norm": 0.7518234848976135, | |
| "learning_rate": 0.00017141608275408006, | |
| "loss": 0.0134, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.74468085106383, | |
| "grad_norm": 0.4271518886089325, | |
| "learning_rate": 0.00017078940874356392, | |
| "loss": 0.0127, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.7872340425531914, | |
| "grad_norm": 0.34387120604515076, | |
| "learning_rate": 0.00017015711200714414, | |
| "loss": 0.0161, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.829787234042553, | |
| "grad_norm": 0.09641717374324799, | |
| "learning_rate": 0.00016951924276746425, | |
| "loss": 0.0185, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.872340425531915, | |
| "grad_norm": 0.44215765595436096, | |
| "learning_rate": 0.00016887585168978562, | |
| "loss": 0.0115, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.9148936170212765, | |
| "grad_norm": 0.06954783201217651, | |
| "learning_rate": 0.0001682269898779632, | |
| "loss": 0.0121, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.9574468085106385, | |
| "grad_norm": 0.21080243587493896, | |
| "learning_rate": 0.00016757270887038654, | |
| "loss": 0.0125, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3382236063480377, | |
| "learning_rate": 0.00016691306063588583, | |
| "loss": 0.0109, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.042553191489362, | |
| "grad_norm": 0.36123183369636536, | |
| "learning_rate": 0.00016624809756960444, | |
| "loss": 0.0093, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.085106382978723, | |
| "grad_norm": 0.45228853821754456, | |
| "learning_rate": 0.00016557787248883696, | |
| "loss": 0.0117, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.127659574468085, | |
| "grad_norm": 0.2724202275276184, | |
| "learning_rate": 0.00016490243862883413, | |
| "loss": 0.0126, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.170212765957447, | |
| "grad_norm": 0.17904357612133026, | |
| "learning_rate": 0.00016422184963857432, | |
| "loss": 0.0103, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.212765957446808, | |
| "grad_norm": 0.4267734885215759, | |
| "learning_rate": 0.00016353615957650236, | |
| "loss": 0.0105, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "grad_norm": 0.11816457659006119, | |
| "learning_rate": 0.00016284542290623567, | |
| "loss": 0.0097, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.297872340425532, | |
| "grad_norm": 0.04166145250201225, | |
| "learning_rate": 0.00016214969449223824, | |
| "loss": 0.0116, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.340425531914893, | |
| "grad_norm": 0.0687410831451416, | |
| "learning_rate": 0.00016144902959546286, | |
| "loss": 0.0088, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.382978723404255, | |
| "grad_norm": 0.37416237592697144, | |
| "learning_rate": 0.00016074348386896177, | |
| "loss": 0.019, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.425531914893617, | |
| "grad_norm": 0.06069188937544823, | |
| "learning_rate": 0.00016003311335346636, | |
| "loss": 0.0138, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.468085106382979, | |
| "grad_norm": 0.0817495658993721, | |
| "learning_rate": 0.00015931797447293552, | |
| "loss": 0.0084, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.51063829787234, | |
| "grad_norm": 0.09787007421255112, | |
| "learning_rate": 0.00015859812403007443, | |
| "loss": 0.009, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.553191489361702, | |
| "grad_norm": 0.06179153174161911, | |
| "learning_rate": 0.0001578736192018224, | |
| "loss": 0.008, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.595744680851064, | |
| "grad_norm": 0.3092339038848877, | |
| "learning_rate": 0.00015714451753481168, | |
| "loss": 0.0131, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.638297872340425, | |
| "grad_norm": 0.06405780464410782, | |
| "learning_rate": 0.0001564108769407962, | |
| "loss": 0.0122, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.680851063829787, | |
| "grad_norm": 0.21473410725593567, | |
| "learning_rate": 0.00015567275569205218, | |
| "loss": 0.0117, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.723404255319149, | |
| "grad_norm": 0.3080317974090576, | |
| "learning_rate": 0.00015493021241674918, | |
| "loss": 0.011, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.76595744680851, | |
| "grad_norm": 0.6501839756965637, | |
| "learning_rate": 0.0001541833060942937, | |
| "loss": 0.0124, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.808510638297872, | |
| "grad_norm": 0.04628886282444, | |
| "learning_rate": 0.00015343209605064422, | |
| "loss": 0.0082, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.851063829787234, | |
| "grad_norm": 0.059793177992105484, | |
| "learning_rate": 0.00015267664195359917, | |
| "loss": 0.0072, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.8936170212765955, | |
| "grad_norm": 0.052162788808345795, | |
| "learning_rate": 0.00015191700380805752, | |
| "loss": 0.0095, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.9361702127659575, | |
| "grad_norm": 0.06147542968392372, | |
| "learning_rate": 0.00015115324195125274, | |
| "loss": 0.0098, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.9787234042553195, | |
| "grad_norm": 0.6586833000183105, | |
| "learning_rate": 0.00015038541704796003, | |
| "loss": 0.0139, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 5.0212765957446805, | |
| "grad_norm": 0.2920892834663391, | |
| "learning_rate": 0.0001496135900856782, | |
| "loss": 0.0083, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 5.0638297872340425, | |
| "grad_norm": 0.32393306493759155, | |
| "learning_rate": 0.0001488378223697851, | |
| "loss": 0.0155, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 5.1063829787234045, | |
| "grad_norm": 0.49029844999313354, | |
| "learning_rate": 0.00014805817551866838, | |
| "loss": 0.0109, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.148936170212766, | |
| "grad_norm": 0.05497799441218376, | |
| "learning_rate": 0.00014727471145883127, | |
| "loss": 0.0095, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 5.191489361702128, | |
| "grad_norm": 0.4540445804595947, | |
| "learning_rate": 0.00014648749241997363, | |
| "loss": 0.0106, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 5.23404255319149, | |
| "grad_norm": 0.16598157584667206, | |
| "learning_rate": 0.00014569658093004935, | |
| "loss": 0.0072, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 5.276595744680851, | |
| "grad_norm": 0.07160704582929611, | |
| "learning_rate": 0.0001449020398102996, | |
| "loss": 0.0108, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.319148936170213, | |
| "grad_norm": 0.197789266705513, | |
| "learning_rate": 0.00014410393217026318, | |
| "loss": 0.0118, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.361702127659575, | |
| "grad_norm": 0.07983817905187607, | |
| "learning_rate": 0.00014330232140276366, | |
| "loss": 0.0076, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 5.404255319148936, | |
| "grad_norm": 0.0746329054236412, | |
| "learning_rate": 0.00014249727117887425, | |
| "loss": 0.0089, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 5.446808510638298, | |
| "grad_norm": 0.09392493963241577, | |
| "learning_rate": 0.00014168884544286053, | |
| "loss": 0.0103, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 5.48936170212766, | |
| "grad_norm": 0.18386998772621155, | |
| "learning_rate": 0.0001408771084071012, | |
| "loss": 0.0096, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.531914893617021, | |
| "grad_norm": 0.07400283962488174, | |
| "learning_rate": 0.00014006212454698797, | |
| "loss": 0.0083, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.574468085106383, | |
| "grad_norm": 0.06513144075870514, | |
| "learning_rate": 0.00013924395859580432, | |
| "loss": 0.0093, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.617021276595745, | |
| "grad_norm": 0.6950928568840027, | |
| "learning_rate": 0.00013842267553958371, | |
| "loss": 0.0073, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.659574468085106, | |
| "grad_norm": 0.05320321023464203, | |
| "learning_rate": 0.00013759834061194794, | |
| "loss": 0.0098, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.702127659574468, | |
| "grad_norm": 0.17569933831691742, | |
| "learning_rate": 0.00013677101928892554, | |
| "loss": 0.0075, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.74468085106383, | |
| "grad_norm": 0.052122730761766434, | |
| "learning_rate": 0.00013594077728375128, | |
| "loss": 0.0107, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.787234042553192, | |
| "grad_norm": 0.2108752280473709, | |
| "learning_rate": 0.00013510768054164653, | |
| "loss": 0.0119, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 5.829787234042553, | |
| "grad_norm": 0.047486983239650726, | |
| "learning_rate": 0.00013427179523458127, | |
| "loss": 0.0092, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 5.872340425531915, | |
| "grad_norm": 0.043320391327142715, | |
| "learning_rate": 0.0001334331877560182, | |
| "loss": 0.0081, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 5.914893617021277, | |
| "grad_norm": 0.09155077487230301, | |
| "learning_rate": 0.00013259192471563912, | |
| "loss": 0.0091, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 5.957446808510638, | |
| "grad_norm": 0.049143675714731216, | |
| "learning_rate": 0.00013174807293405428, | |
| "loss": 0.0089, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.07365540415048599, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.0068, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 6.042553191489362, | |
| "grad_norm": 0.11437718570232391, | |
| "learning_rate": 0.00013005287145248878, | |
| "loss": 0.0064, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 6.085106382978723, | |
| "grad_norm": 0.06010650098323822, | |
| "learning_rate": 0.0001292016564005219, | |
| "loss": 0.0074, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 6.127659574468085, | |
| "grad_norm": 0.04595167934894562, | |
| "learning_rate": 0.0001283481218926818, | |
| "loss": 0.0066, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 6.170212765957447, | |
| "grad_norm": 0.05351310595870018, | |
| "learning_rate": 0.00012749233572428804, | |
| "loss": 0.0097, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 6.212765957446808, | |
| "grad_norm": 0.13630953431129456, | |
| "learning_rate": 0.00012663436586950714, | |
| "loss": 0.0079, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 6.25531914893617, | |
| "grad_norm": 0.17622588574886322, | |
| "learning_rate": 0.00012577428047595344, | |
| "loss": 0.0084, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 6.297872340425532, | |
| "grad_norm": 0.050954531878232956, | |
| "learning_rate": 0.0001249121478592762, | |
| "loss": 0.0077, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 6.340425531914893, | |
| "grad_norm": 0.3051726818084717, | |
| "learning_rate": 0.0001240480364977335, | |
| "loss": 0.0085, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 6.382978723404255, | |
| "grad_norm": 0.30509302020072937, | |
| "learning_rate": 0.00012318201502675285, | |
| "loss": 0.0092, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.425531914893617, | |
| "grad_norm": 0.09164142608642578, | |
| "learning_rate": 0.00012231415223347972, | |
| "loss": 0.008, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 6.468085106382979, | |
| "grad_norm": 0.05406223237514496, | |
| "learning_rate": 0.0001214445170513139, | |
| "loss": 0.0078, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 6.51063829787234, | |
| "grad_norm": 0.05845744535326958, | |
| "learning_rate": 0.00012057317855443395, | |
| "loss": 0.0092, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 6.553191489361702, | |
| "grad_norm": 0.05021122843027115, | |
| "learning_rate": 0.00011970020595231101, | |
| "loss": 0.007, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 6.595744680851064, | |
| "grad_norm": 0.10315235704183578, | |
| "learning_rate": 0.00011882566858421135, | |
| "loss": 0.0068, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 6.638297872340425, | |
| "grad_norm": 0.08750782907009125, | |
| "learning_rate": 0.00011794963591368893, | |
| "loss": 0.009, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 6.680851063829787, | |
| "grad_norm": 0.05412838235497475, | |
| "learning_rate": 0.0001170721775230679, | |
| "loss": 0.0071, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 6.723404255319149, | |
| "grad_norm": 0.17292432487010956, | |
| "learning_rate": 0.00011619336310791586, | |
| "loss": 0.0091, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 6.76595744680851, | |
| "grad_norm": 0.05503688380122185, | |
| "learning_rate": 0.00011531326247150803, | |
| "loss": 0.0069, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.808510638297872, | |
| "grad_norm": 0.05121155083179474, | |
| "learning_rate": 0.00011443194551928266, | |
| "loss": 0.008, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.851063829787234, | |
| "grad_norm": 0.0626005157828331, | |
| "learning_rate": 0.00011354948225328877, | |
| "loss": 0.0065, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 6.8936170212765955, | |
| "grad_norm": 0.058921247720718384, | |
| "learning_rate": 0.0001126659427666257, | |
| "loss": 0.0078, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 6.9361702127659575, | |
| "grad_norm": 0.058523211628198624, | |
| "learning_rate": 0.00011178139723787597, | |
| "loss": 0.008, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 6.9787234042553195, | |
| "grad_norm": 0.18594586849212646, | |
| "learning_rate": 0.00011089591592553082, | |
| "loss": 0.0076, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 7.0212765957446805, | |
| "grad_norm": 0.053747180849313736, | |
| "learning_rate": 0.00011000956916240985, | |
| "loss": 0.0074, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 7.0638297872340425, | |
| "grad_norm": 0.03964696079492569, | |
| "learning_rate": 0.00010912242735007441, | |
| "loss": 0.0071, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 7.1063829787234045, | |
| "grad_norm": 0.05952566862106323, | |
| "learning_rate": 0.00010823456095323579, | |
| "loss": 0.0065, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 7.148936170212766, | |
| "grad_norm": 0.21424749493598938, | |
| "learning_rate": 0.00010734604049415822, | |
| "loss": 0.0075, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 7.191489361702128, | |
| "grad_norm": 0.03922433406114578, | |
| "learning_rate": 0.0001064569365470574, | |
| "loss": 0.0071, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 7.23404255319149, | |
| "grad_norm": 0.05505215749144554, | |
| "learning_rate": 0.00010556731973249485, | |
| "loss": 0.0061, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.276595744680851, | |
| "grad_norm": 0.03699268028140068, | |
| "learning_rate": 0.00010467726071176853, | |
| "loss": 0.0075, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 7.319148936170213, | |
| "grad_norm": 0.04546520113945007, | |
| "learning_rate": 0.00010378683018130047, | |
| "loss": 0.0072, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 7.361702127659575, | |
| "grad_norm": 0.056984953582286835, | |
| "learning_rate": 0.0001028960988670212, | |
| "loss": 0.007, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 7.404255319148936, | |
| "grad_norm": 0.24714773893356323, | |
| "learning_rate": 0.00010200513751875227, | |
| "loss": 0.0074, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 7.446808510638298, | |
| "grad_norm": 0.06558862328529358, | |
| "learning_rate": 0.00010111401690458654, | |
| "loss": 0.0064, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.48936170212766, | |
| "grad_norm": 0.06254340708255768, | |
| "learning_rate": 0.00010022280780526725, | |
| "loss": 0.0076, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 7.531914893617021, | |
| "grad_norm": 0.08606445789337158, | |
| "learning_rate": 9.93315810085658e-05, | |
| "loss": 0.0076, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 7.574468085106383, | |
| "grad_norm": 0.051956657320261, | |
| "learning_rate": 9.844040730365936e-05, | |
| "loss": 0.0073, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 7.617021276595745, | |
| "grad_norm": 0.24819940328598022, | |
| "learning_rate": 9.754935747550804e-05, | |
| "loss": 0.0077, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 7.659574468085106, | |
| "grad_norm": 0.039045918732881546, | |
| "learning_rate": 9.665850229923258e-05, | |
| "loss": 0.0071, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.702127659574468, | |
| "grad_norm": 0.04438428208231926, | |
| "learning_rate": 9.57679125344927e-05, | |
| "loss": 0.0074, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 7.74468085106383, | |
| "grad_norm": 0.31067439913749695, | |
| "learning_rate": 9.487765891986682e-05, | |
| "loss": 0.0087, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 7.787234042553192, | |
| "grad_norm": 0.06251853704452515, | |
| "learning_rate": 9.398781216723331e-05, | |
| "loss": 0.0069, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 7.829787234042553, | |
| "grad_norm": 0.04706185683608055, | |
| "learning_rate": 9.309844295615389e-05, | |
| "loss": 0.0072, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 7.872340425531915, | |
| "grad_norm": 0.03839968144893646, | |
| "learning_rate": 9.220962192825968e-05, | |
| "loss": 0.0069, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.914893617021277, | |
| "grad_norm": 0.03722318261861801, | |
| "learning_rate": 9.132141968164026e-05, | |
| "loss": 0.0069, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 7.957446808510638, | |
| "grad_norm": 0.0543668158352375, | |
| "learning_rate": 9.043390676523604e-05, | |
| "loss": 0.0076, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.051696889102458954, | |
| "learning_rate": 8.954715367323468e-05, | |
| "loss": 0.0066, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 8.042553191489361, | |
| "grad_norm": 0.08786992728710175, | |
| "learning_rate": 8.866123083947182e-05, | |
| "loss": 0.0062, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 8.085106382978724, | |
| "grad_norm": 0.3556790351867676, | |
| "learning_rate": 8.777620863183657e-05, | |
| "loss": 0.0079, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.127659574468085, | |
| "grad_norm": 0.08049122244119644, | |
| "learning_rate": 8.689215734668232e-05, | |
| "loss": 0.0064, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 8.170212765957446, | |
| "grad_norm": 0.05408492311835289, | |
| "learning_rate": 8.600914720324316e-05, | |
| "loss": 0.0077, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 8.212765957446809, | |
| "grad_norm": 0.034946855157613754, | |
| "learning_rate": 8.512724833805634e-05, | |
| "loss": 0.0067, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 8.25531914893617, | |
| "grad_norm": 0.05422681197524071, | |
| "learning_rate": 8.424653079939156e-05, | |
| "loss": 0.0062, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 8.297872340425531, | |
| "grad_norm": 0.06405791640281677, | |
| "learning_rate": 8.336706454168701e-05, | |
| "loss": 0.0064, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 8.340425531914894, | |
| "grad_norm": 0.08694509416818619, | |
| "learning_rate": 8.248891941999297e-05, | |
| "loss": 0.006, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 8.382978723404255, | |
| "grad_norm": 0.05589594319462776, | |
| "learning_rate": 8.161216518442334e-05, | |
| "loss": 0.0067, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 8.425531914893616, | |
| "grad_norm": 0.05914654955267906, | |
| "learning_rate": 8.073687147461547e-05, | |
| "loss": 0.0065, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 8.46808510638298, | |
| "grad_norm": 0.05237606540322304, | |
| "learning_rate": 7.98631078141987e-05, | |
| "loss": 0.0071, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 8.51063829787234, | |
| "grad_norm": 0.056138359010219574, | |
| "learning_rate": 7.89909436052722e-05, | |
| "loss": 0.0079, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.553191489361701, | |
| "grad_norm": 0.06285829097032547, | |
| "learning_rate": 7.812044812289249e-05, | |
| "loss": 0.0064, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 8.595744680851064, | |
| "grad_norm": 0.04014954715967178, | |
| "learning_rate": 7.72516905095709e-05, | |
| "loss": 0.0072, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 8.638297872340425, | |
| "grad_norm": 0.04423344135284424, | |
| "learning_rate": 7.638473976978177e-05, | |
| "loss": 0.0065, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 8.680851063829786, | |
| "grad_norm": 0.05287999287247658, | |
| "learning_rate": 7.55196647644814e-05, | |
| "loss": 0.0081, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 8.72340425531915, | |
| "grad_norm": 0.056493304669857025, | |
| "learning_rate": 7.465653420563845e-05, | |
| "loss": 0.0067, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 8.76595744680851, | |
| "grad_norm": 0.057281821966171265, | |
| "learning_rate": 7.379541665077643e-05, | |
| "loss": 0.0078, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 8.808510638297872, | |
| "grad_norm": 0.053731102496385574, | |
| "learning_rate": 7.293638049752812e-05, | |
| "loss": 0.0066, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 8.851063829787234, | |
| "grad_norm": 0.18697482347488403, | |
| "learning_rate": 7.207949397820278e-05, | |
| "loss": 0.0069, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 8.893617021276595, | |
| "grad_norm": 0.0387713797390461, | |
| "learning_rate": 7.122482515436661e-05, | |
| "loss": 0.007, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 8.936170212765958, | |
| "grad_norm": 0.04441935196518898, | |
| "learning_rate": 7.037244191143661e-05, | |
| "loss": 0.0067, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.97872340425532, | |
| "grad_norm": 0.05758345127105713, | |
| "learning_rate": 6.952241195328868e-05, | |
| "loss": 0.0065, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 9.02127659574468, | |
| "grad_norm": 0.050706226378679276, | |
| "learning_rate": 6.867480279687974e-05, | |
| "loss": 0.0063, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 9.063829787234043, | |
| "grad_norm": 0.05180887505412102, | |
| "learning_rate": 6.782968176688514e-05, | |
| "loss": 0.0062, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 9.106382978723405, | |
| "grad_norm": 0.05492401868104935, | |
| "learning_rate": 6.6987115990351e-05, | |
| "loss": 0.006, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 9.148936170212766, | |
| "grad_norm": 0.053439777344465256, | |
| "learning_rate": 6.614717239136246e-05, | |
| "loss": 0.0066, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 9.191489361702128, | |
| "grad_norm": 0.15650008618831635, | |
| "learning_rate": 6.530991768572794e-05, | |
| "loss": 0.006, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 9.23404255319149, | |
| "grad_norm": 0.04846300184726715, | |
| "learning_rate": 6.447541837568e-05, | |
| "loss": 0.0068, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 9.27659574468085, | |
| "grad_norm": 0.047157011926174164, | |
| "learning_rate": 6.364374074459307e-05, | |
| "loss": 0.006, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 9.319148936170214, | |
| "grad_norm": 0.047358132898807526, | |
| "learning_rate": 6.281495085171869e-05, | |
| "loss": 0.0058, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 9.361702127659575, | |
| "grad_norm": 0.11200093477964401, | |
| "learning_rate": 6.198911452693853e-05, | |
| "loss": 0.007, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.404255319148936, | |
| "grad_norm": 0.06481984257698059, | |
| "learning_rate": 6.116629736553552e-05, | |
| "loss": 0.0069, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 9.446808510638299, | |
| "grad_norm": 0.04743931442499161, | |
| "learning_rate": 6.0346564722983736e-05, | |
| "loss": 0.0072, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 9.48936170212766, | |
| "grad_norm": 0.07307924330234528, | |
| "learning_rate": 5.952998170975724e-05, | |
| "loss": 0.0062, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 9.53191489361702, | |
| "grad_norm": 0.05650079995393753, | |
| "learning_rate": 5.871661318615848e-05, | |
| "loss": 0.0061, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 9.574468085106384, | |
| "grad_norm": 0.057734813541173935, | |
| "learning_rate": 5.790652375716652e-05, | |
| "loss": 0.0068, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.617021276595745, | |
| "grad_norm": 0.05608903244137764, | |
| "learning_rate": 5.709977776730537e-05, | |
| "loss": 0.0071, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 9.659574468085106, | |
| "grad_norm": 0.07133158296346664, | |
| "learning_rate": 5.62964392955335e-05, | |
| "loss": 0.0067, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 9.702127659574469, | |
| "grad_norm": 0.2167043685913086, | |
| "learning_rate": 5.549657215015367e-05, | |
| "loss": 0.0067, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 9.74468085106383, | |
| "grad_norm": 0.06523007154464722, | |
| "learning_rate": 5.470023986374516e-05, | |
| "loss": 0.0068, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 9.787234042553191, | |
| "grad_norm": 0.04895370826125145, | |
| "learning_rate": 5.39075056881172e-05, | |
| "loss": 0.0058, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.829787234042554, | |
| "grad_norm": 0.056433409452438354, | |
| "learning_rate": 5.31184325892849e-05, | |
| "loss": 0.0067, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 9.872340425531915, | |
| "grad_norm": 0.049970466643571854, | |
| "learning_rate": 5.233308324246805e-05, | |
| "loss": 0.0058, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 9.914893617021276, | |
| "grad_norm": 0.08367093652486801, | |
| "learning_rate": 5.155152002711285e-05, | |
| "loss": 0.0063, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 9.957446808510639, | |
| "grad_norm": 0.07821632921695709, | |
| "learning_rate": 5.077380502193725e-05, | |
| "loss": 0.0071, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.04344159737229347, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.0078, | |
| "step": 2350 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3525, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "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": 9.44675050315776e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |