| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1535, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03257328990228013, | |
| "grad_norm": 8.907567024230957, | |
| "learning_rate": 2.5974025974025972e-05, | |
| "loss": 1.8912, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06514657980456026, | |
| "grad_norm": 9.222850799560547, | |
| "learning_rate": 5.1948051948051944e-05, | |
| "loss": 0.5715, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09771986970684039, | |
| "grad_norm": 2.1737639904022217, | |
| "learning_rate": 7.792207792207793e-05, | |
| "loss": 0.2636, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13029315960912052, | |
| "grad_norm": 2.1530239582061768, | |
| "learning_rate": 0.00010389610389610389, | |
| "loss": 0.1988, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16286644951140064, | |
| "grad_norm": 1.3454252481460571, | |
| "learning_rate": 0.00012987012987012987, | |
| "loss": 0.1389, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19543973941368079, | |
| "grad_norm": 1.806857943534851, | |
| "learning_rate": 0.00015584415584415587, | |
| "loss": 0.1527, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2280130293159609, | |
| "grad_norm": 1.9475533962249756, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.1769, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26058631921824105, | |
| "grad_norm": 0.9884098768234253, | |
| "learning_rate": 0.0001999979107245606, | |
| "loss": 0.1225, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2931596091205212, | |
| "grad_norm": 0.8951364755630493, | |
| "learning_rate": 0.0001999607704786645, | |
| "loss": 0.108, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3257328990228013, | |
| "grad_norm": 1.403916835784912, | |
| "learning_rate": 0.00019987722173728587, | |
| "loss": 0.1324, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3583061889250814, | |
| "grad_norm": 1.019700288772583, | |
| "learning_rate": 0.00019974730328935534, | |
| "loss": 0.1047, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.39087947882736157, | |
| "grad_norm": 0.8862571120262146, | |
| "learning_rate": 0.00019957107545173192, | |
| "loss": 0.1011, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4234527687296417, | |
| "grad_norm": 0.7065503597259521, | |
| "learning_rate": 0.00019934862004120004, | |
| "loss": 0.0847, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4560260586319218, | |
| "grad_norm": 1.27033531665802, | |
| "learning_rate": 0.00019908004033648453, | |
| "loss": 0.0785, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48859934853420195, | |
| "grad_norm": 0.9348053932189941, | |
| "learning_rate": 0.00019876546103030196, | |
| "loss": 0.0835, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5211726384364821, | |
| "grad_norm": 0.8986678719520569, | |
| "learning_rate": 0.00019840502817146966, | |
| "loss": 0.083, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5537459283387622, | |
| "grad_norm": 1.101909875869751, | |
| "learning_rate": 0.00019799890909710013, | |
| "loss": 0.0893, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5863192182410424, | |
| "grad_norm": 0.7529295086860657, | |
| "learning_rate": 0.00019754729235491207, | |
| "loss": 0.0877, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6188925081433225, | |
| "grad_norm": 0.7967862486839294, | |
| "learning_rate": 0.0001970503876156937, | |
| "loss": 0.0888, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6514657980456026, | |
| "grad_norm": 1.1783355474472046, | |
| "learning_rate": 0.00019650842557595967, | |
| "loss": 0.0756, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6840390879478827, | |
| "grad_norm": 1.0442906618118286, | |
| "learning_rate": 0.00019592165785084603, | |
| "loss": 0.0999, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7166123778501629, | |
| "grad_norm": 4.202133655548096, | |
| "learning_rate": 0.00019529035685729391, | |
| "loss": 0.0924, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.749185667752443, | |
| "grad_norm": 2.0415215492248535, | |
| "learning_rate": 0.00019461481568757506, | |
| "loss": 0.1073, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7817589576547231, | |
| "grad_norm": 0.6503229737281799, | |
| "learning_rate": 0.00019389534797321884, | |
| "loss": 0.0718, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8143322475570033, | |
| "grad_norm": 2.1912670135498047, | |
| "learning_rate": 0.00019313228773940345, | |
| "loss": 0.0771, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8469055374592834, | |
| "grad_norm": 1.1402329206466675, | |
| "learning_rate": 0.00019232598924987903, | |
| "loss": 0.0942, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8794788273615635, | |
| "grad_norm": 1.558457612991333, | |
| "learning_rate": 0.0001914768268424946, | |
| "loss": 0.0822, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9120521172638436, | |
| "grad_norm": 0.685741662979126, | |
| "learning_rate": 0.00019058519475540538, | |
| "loss": 0.0798, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9446254071661238, | |
| "grad_norm": 0.8257420659065247, | |
| "learning_rate": 0.00018965150694404094, | |
| "loss": 0.0555, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9771986970684039, | |
| "grad_norm": 0.7239089608192444, | |
| "learning_rate": 0.00018867619688891937, | |
| "loss": 0.079, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.009771986970684, | |
| "grad_norm": 0.46614518761634827, | |
| "learning_rate": 0.0001876597173943965, | |
| "loss": 0.0696, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0423452768729642, | |
| "grad_norm": 0.6465802788734436, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.0675, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0749185667752443, | |
| "grad_norm": 1.07686448097229, | |
| "learning_rate": 0.00018550515665355247, | |
| "loss": 0.0594, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1074918566775245, | |
| "grad_norm": 0.37035682797431946, | |
| "learning_rate": 0.000184368075698865, | |
| "loss": 0.0508, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1400651465798046, | |
| "grad_norm": 0.5353094935417175, | |
| "learning_rate": 0.00018319182542364117, | |
| "loss": 0.0752, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1726384364820848, | |
| "grad_norm": 0.5493381023406982, | |
| "learning_rate": 0.00018197695192216702, | |
| "loss": 0.0568, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.205211726384365, | |
| "grad_norm": 0.3111603260040283, | |
| "learning_rate": 0.00018072401922022117, | |
| "loss": 0.0427, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.237785016286645, | |
| "grad_norm": 0.4221360385417938, | |
| "learning_rate": 0.0001794336090132164, | |
| "loss": 0.0633, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2703583061889252, | |
| "grad_norm": 0.8296531438827515, | |
| "learning_rate": 0.00017810632039613736, | |
| "loss": 0.0585, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3029315960912053, | |
| "grad_norm": 0.4350226819515228, | |
| "learning_rate": 0.00017674276958540072, | |
| "loss": 0.0476, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3355048859934853, | |
| "grad_norm": 0.8966237306594849, | |
| "learning_rate": 0.00017534358963276607, | |
| "loss": 0.0532, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3680781758957654, | |
| "grad_norm": 0.4114079177379608, | |
| "learning_rate": 0.00017390943013143083, | |
| "loss": 0.0593, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4006514657980456, | |
| "grad_norm": 0.5560611486434937, | |
| "learning_rate": 0.0001724409569144455, | |
| "loss": 0.0514, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4332247557003257, | |
| "grad_norm": 0.5275083184242249, | |
| "learning_rate": 0.0001709388517455893, | |
| "loss": 0.0453, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4657980456026058, | |
| "grad_norm": 0.541585385799408, | |
| "learning_rate": 0.00016940381200284972, | |
| "loss": 0.0441, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.498371335504886, | |
| "grad_norm": 0.3919298052787781, | |
| "learning_rate": 0.0001678365503546528, | |
| "loss": 0.0348, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5309446254071661, | |
| "grad_norm": 0.626324474811554, | |
| "learning_rate": 0.0001662377944289948, | |
| "loss": 0.0426, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5635179153094463, | |
| "grad_norm": 4.0510358810424805, | |
| "learning_rate": 0.0001646082864756282, | |
| "loss": 0.0422, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5960912052117264, | |
| "grad_norm": 0.6301591396331787, | |
| "learning_rate": 0.00016294878302145987, | |
| "loss": 0.0435, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6286644951140063, | |
| "grad_norm": 0.6829243302345276, | |
| "learning_rate": 0.0001612600545193203, | |
| "loss": 0.0566, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6612377850162865, | |
| "grad_norm": 0.721428632736206, | |
| "learning_rate": 0.00015954288499026782, | |
| "loss": 0.051, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6938110749185666, | |
| "grad_norm": 0.810246467590332, | |
| "learning_rate": 0.0001577980716595934, | |
| "loss": 0.0509, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7263843648208468, | |
| "grad_norm": 0.8535786867141724, | |
| "learning_rate": 0.00015602642458669528, | |
| "loss": 0.0527, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.758957654723127, | |
| "grad_norm": 0.7078109383583069, | |
| "learning_rate": 0.0001542287662889948, | |
| "loss": 0.0517, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.791530944625407, | |
| "grad_norm": 0.9758318662643433, | |
| "learning_rate": 0.00015240593136006897, | |
| "loss": 0.0425, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8241042345276872, | |
| "grad_norm": 1.1173475980758667, | |
| "learning_rate": 0.0001505587660821759, | |
| "loss": 0.0462, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8566775244299674, | |
| "grad_norm": 0.44922390580177307, | |
| "learning_rate": 0.0001486881280333539, | |
| "loss": 0.0519, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8892508143322475, | |
| "grad_norm": 0.5552352666854858, | |
| "learning_rate": 0.00014679488568927616, | |
| "loss": 0.0518, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9218241042345277, | |
| "grad_norm": 0.6189628839492798, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.0397, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9543973941368078, | |
| "grad_norm": 0.5633209943771362, | |
| "learning_rate": 0.0001429441140821209, | |
| "loss": 0.0383, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.986970684039088, | |
| "grad_norm": 0.44652634859085083, | |
| "learning_rate": 0.00014098837260555084, | |
| "loss": 0.0416, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.019543973941368, | |
| "grad_norm": 0.6565033197402954, | |
| "learning_rate": 0.0001390136015767295, | |
| "loss": 0.0422, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0521172638436482, | |
| "grad_norm": 0.5799785256385803, | |
| "learning_rate": 0.00013702071781684517, | |
| "loss": 0.0484, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0846905537459284, | |
| "grad_norm": 0.46765848994255066, | |
| "learning_rate": 0.00013501064655623094, | |
| "loss": 0.0303, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1172638436482085, | |
| "grad_norm": 0.5322710275650024, | |
| "learning_rate": 0.00013298432100481079, | |
| "loss": 0.0321, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.1498371335504887, | |
| "grad_norm": 0.9900329113006592, | |
| "learning_rate": 0.0001309426819188409, | |
| "loss": 0.0377, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.182410423452769, | |
| "grad_norm": 0.6508163809776306, | |
| "learning_rate": 0.0001288866771641474, | |
| "loss": 0.0341, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.214983713355049, | |
| "grad_norm": 0.6541309356689453, | |
| "learning_rate": 0.00012681726127606376, | |
| "loss": 0.0535, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.247557003257329, | |
| "grad_norm": 0.3469844162464142, | |
| "learning_rate": 0.000124735395016271, | |
| "loss": 0.0314, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2801302931596092, | |
| "grad_norm": 0.5612144470214844, | |
| "learning_rate": 0.00012264204492674815, | |
| "loss": 0.0334, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3127035830618894, | |
| "grad_norm": 0.6845806837081909, | |
| "learning_rate": 0.0001205381828810382, | |
| "loss": 0.0417, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3452768729641695, | |
| "grad_norm": 0.6910083889961243, | |
| "learning_rate": 0.00011842478563303952, | |
| "loss": 0.0295, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3778501628664497, | |
| "grad_norm": 0.5824999213218689, | |
| "learning_rate": 0.00011630283436353098, | |
| "loss": 0.0367, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.41042345276873, | |
| "grad_norm": 0.9331006407737732, | |
| "learning_rate": 0.00011417331422464205, | |
| "loss": 0.0349, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.44299674267101, | |
| "grad_norm": 0.33145973086357117, | |
| "learning_rate": 0.00011203721388247923, | |
| "loss": 0.0291, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.47557003257329, | |
| "grad_norm": 0.35993027687072754, | |
| "learning_rate": 0.00010989552505812072, | |
| "loss": 0.0415, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5081433224755703, | |
| "grad_norm": 0.446781188249588, | |
| "learning_rate": 0.0001077492420671931, | |
| "loss": 0.0319, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5407166123778504, | |
| "grad_norm": 0.31257572770118713, | |
| "learning_rate": 0.00010559936135824322, | |
| "loss": 0.0269, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5732899022801305, | |
| "grad_norm": 0.4158976078033447, | |
| "learning_rate": 0.00010344688105012005, | |
| "loss": 0.0364, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.6058631921824107, | |
| "grad_norm": 0.4493694305419922, | |
| "learning_rate": 0.00010129280046858086, | |
| "loss": 0.0377, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.6384364820846904, | |
| "grad_norm": 0.4073673188686371, | |
| "learning_rate": 9.913811968233716e-05, | |
| "loss": 0.0203, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6710097719869705, | |
| "grad_norm": 0.47181200981140137, | |
| "learning_rate": 9.69838390387558e-05, | |
| "loss": 0.0479, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7035830618892507, | |
| "grad_norm": 0.3475622236728668, | |
| "learning_rate": 9.483095869943055e-05, | |
| "loss": 0.0256, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.736156351791531, | |
| "grad_norm": 0.3033672571182251, | |
| "learning_rate": 9.268047817583998e-05, | |
| "loss": 0.0196, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.768729641693811, | |
| "grad_norm": 0.38617008924484253, | |
| "learning_rate": 9.053339586530723e-05, | |
| "loss": 0.0246, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.801302931596091, | |
| "grad_norm": 0.4176099896430969, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.034, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.8338762214983713, | |
| "grad_norm": 0.5051398277282715, | |
| "learning_rate": 8.625341112152487e-05, | |
| "loss": 0.0267, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8664495114006514, | |
| "grad_norm": 0.3694643974304199, | |
| "learning_rate": 8.412249574431428e-05, | |
| "loss": 0.0252, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8990228013029316, | |
| "grad_norm": 0.47114214301109314, | |
| "learning_rate": 8.199895176971488e-05, | |
| "loss": 0.0331, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9315960912052117, | |
| "grad_norm": 0.3039054572582245, | |
| "learning_rate": 7.988376508929676e-05, | |
| "loss": 0.0232, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.964169381107492, | |
| "grad_norm": 0.31323984265327454, | |
| "learning_rate": 7.777791771461332e-05, | |
| "loss": 0.0245, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.996742671009772, | |
| "grad_norm": 0.36317452788352966, | |
| "learning_rate": 7.568238732128585e-05, | |
| "loss": 0.0315, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.029315960912052, | |
| "grad_norm": 0.3417069911956787, | |
| "learning_rate": 7.359814679510065e-05, | |
| "loss": 0.031, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0618892508143323, | |
| "grad_norm": 0.5224521160125732, | |
| "learning_rate": 7.152616378033042e-05, | |
| "loss": 0.0198, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0944625407166124, | |
| "grad_norm": 0.45148587226867676, | |
| "learning_rate": 6.94674002304887e-05, | |
| "loss": 0.0243, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.1270358306188926, | |
| "grad_norm": 0.6386067271232605, | |
| "learning_rate": 6.742281196172663e-05, | |
| "loss": 0.025, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1596091205211727, | |
| "grad_norm": 0.3056974709033966, | |
| "learning_rate": 6.539334820907888e-05, | |
| "loss": 0.0162, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.192182410423453, | |
| "grad_norm": 0.38449615240097046, | |
| "learning_rate": 6.337995118576521e-05, | |
| "loss": 0.0236, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.224755700325733, | |
| "grad_norm": 0.7297264933586121, | |
| "learning_rate": 6.138355564575169e-05, | |
| "loss": 0.0211, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.257328990228013, | |
| "grad_norm": 0.2978608310222626, | |
| "learning_rate": 5.940508844977537e-05, | |
| "loss": 0.0238, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2899022801302933, | |
| "grad_norm": 0.2972109913825989, | |
| "learning_rate": 5.744546813503328e-05, | |
| "loss": 0.0341, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.3224755700325734, | |
| "grad_norm": 3.2639620304107666, | |
| "learning_rate": 5.550560448873575e-05, | |
| "loss": 0.0264, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.3550488599348536, | |
| "grad_norm": 0.28048136830329895, | |
| "learning_rate": 5.358639812572244e-05, | |
| "loss": 0.0199, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3876221498371337, | |
| "grad_norm": 0.5666626691818237, | |
| "learning_rate": 5.168874007033615e-05, | |
| "loss": 0.0222, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.420195439739414, | |
| "grad_norm": 0.2740839421749115, | |
| "learning_rate": 4.9813511342749805e-05, | |
| "loss": 0.0207, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.4527687296416936, | |
| "grad_norm": 0.2190520465373993, | |
| "learning_rate": 4.7961582549937675e-05, | |
| "loss": 0.0214, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.4853420195439737, | |
| "grad_norm": 0.2338235080242157, | |
| "learning_rate": 4.6133813481481246e-05, | |
| "loss": 0.0165, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.517915309446254, | |
| "grad_norm": 0.30623218417167664, | |
| "learning_rate": 4.433105271039721e-05, | |
| "loss": 0.0192, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.550488599348534, | |
| "grad_norm": 0.40116289258003235, | |
| "learning_rate": 4.255413719917294e-05, | |
| "loss": 0.023, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.583061889250814, | |
| "grad_norm": 0.2895350158214569, | |
| "learning_rate": 4.080389191119241e-05, | |
| "loss": 0.0146, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.6156351791530943, | |
| "grad_norm": 0.29448819160461426, | |
| "learning_rate": 3.9081129427732774e-05, | |
| "loss": 0.0166, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.6482084690553744, | |
| "grad_norm": 0.21917672455310822, | |
| "learning_rate": 3.7386649570709644e-05, | |
| "loss": 0.0143, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.6807817589576546, | |
| "grad_norm": 0.29176798462867737, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.0164, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.7133550488599347, | |
| "grad_norm": 0.23363909125328064, | |
| "learning_rate": 3.408567100493787e-05, | |
| "loss": 0.0134, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.745928338762215, | |
| "grad_norm": 0.23968514800071716, | |
| "learning_rate": 3.248070483188426e-05, | |
| "loss": 0.0253, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.778501628664495, | |
| "grad_norm": 0.32823416590690613, | |
| "learning_rate": 3.090708564515124e-05, | |
| "loss": 0.0184, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.811074918566775, | |
| "grad_norm": 0.25938963890075684, | |
| "learning_rate": 2.936554402433087e-05, | |
| "loss": 0.0181, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.8436482084690553, | |
| "grad_norm": 0.29427650570869446, | |
| "learning_rate": 2.7856795656457257e-05, | |
| "loss": 0.0154, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8762214983713354, | |
| "grad_norm": 0.2785288393497467, | |
| "learning_rate": 2.6381541003736486e-05, | |
| "loss": 0.0213, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.9087947882736156, | |
| "grad_norm": 0.21833765506744385, | |
| "learning_rate": 2.494046497834518e-05, | |
| "loss": 0.019, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.9413680781758957, | |
| "grad_norm": 0.3013763725757599, | |
| "learning_rate": 2.3534236624448302e-05, | |
| "loss": 0.0194, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.973941368078176, | |
| "grad_norm": 0.5469600558280945, | |
| "learning_rate": 2.2163508807583998e-05, | |
| "loss": 0.0225, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.006514657980456, | |
| "grad_norm": 0.2737639248371124, | |
| "learning_rate": 2.082891791155954e-05, | |
| "loss": 0.0142, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.039087947882736, | |
| "grad_norm": 0.43073591589927673, | |
| "learning_rate": 1.9531083542999317e-05, | |
| "loss": 0.0164, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.071661237785016, | |
| "grad_norm": 0.14076876640319824, | |
| "learning_rate": 1.8270608243681953e-05, | |
| "loss": 0.0115, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.1042345276872965, | |
| "grad_norm": 0.2852550446987152, | |
| "learning_rate": 1.7048077210799772e-05, | |
| "loss": 0.0124, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.136807817589577, | |
| "grad_norm": 0.24066734313964844, | |
| "learning_rate": 1.5864058025271246e-05, | |
| "loss": 0.0133, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.169381107491857, | |
| "grad_norm": 0.24509824812412262, | |
| "learning_rate": 1.47191003882317e-05, | |
| "loss": 0.0167, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.201954397394137, | |
| "grad_norm": 0.2472413182258606, | |
| "learning_rate": 1.3613735865825305e-05, | |
| "loss": 0.0166, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.234527687296417, | |
| "grad_norm": 0.2663990557193756, | |
| "learning_rate": 1.2548477642416256e-05, | |
| "loss": 0.0164, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.267100977198697, | |
| "grad_norm": 0.1704978048801422, | |
| "learning_rate": 1.1523820282334219e-05, | |
| "loss": 0.011, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.299674267100977, | |
| "grad_norm": 0.26135244965553284, | |
| "learning_rate": 1.0540239500264516e-05, | |
| "loss": 0.0162, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.3322475570032575, | |
| "grad_norm": 0.2877485454082489, | |
| "learning_rate": 9.598191940389256e-06, | |
| "loss": 0.0181, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.364820846905538, | |
| "grad_norm": 0.5725845098495483, | |
| "learning_rate": 8.698114964382598e-06, | |
| "loss": 0.0283, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.397394136807818, | |
| "grad_norm": 0.17230850458145142, | |
| "learning_rate": 7.840426448358085e-06, | |
| "loss": 0.0099, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.429967426710098, | |
| "grad_norm": 0.42446282505989075, | |
| "learning_rate": 7.025524588862542e-06, | |
| "loss": 0.022, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.462540716612378, | |
| "grad_norm": 0.15192359685897827, | |
| "learning_rate": 6.253787718006498e-06, | |
| "loss": 0.0179, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.495114006514658, | |
| "grad_norm": 0.27675601840019226, | |
| "learning_rate": 5.525574127817046e-06, | |
| "loss": 0.0128, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.527687296416938, | |
| "grad_norm": 0.273651123046875, | |
| "learning_rate": 4.841221903894633e-06, | |
| "loss": 0.0143, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.5602605863192185, | |
| "grad_norm": 0.14251072704792023, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.0137, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.592833876221499, | |
| "grad_norm": 0.2250894010066986, | |
| "learning_rate": 3.605351932801693e-06, | |
| "loss": 0.0248, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.625407166123779, | |
| "grad_norm": 0.22826635837554932, | |
| "learning_rate": 3.0544079593795573e-06, | |
| "loss": 0.018, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.657980456026059, | |
| "grad_norm": 0.20432162284851074, | |
| "learning_rate": 2.548472633337007e-06, | |
| "loss": 0.0181, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.690553745928339, | |
| "grad_norm": 0.31331172585487366, | |
| "learning_rate": 2.0877808437928637e-06, | |
| "loss": 0.0144, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.723127035830619, | |
| "grad_norm": 0.2865249812602997, | |
| "learning_rate": 1.6725464747811447e-06, | |
| "loss": 0.0201, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.755700325732899, | |
| "grad_norm": 0.1912713199853897, | |
| "learning_rate": 1.3029623059517493e-06, | |
| "loss": 0.0116, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.7882736156351795, | |
| "grad_norm": 0.1277429610490799, | |
| "learning_rate": 9.791999230692629e-07, | |
| "loss": 0.0119, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.82084690553746, | |
| "grad_norm": 0.13365145027637482, | |
| "learning_rate": 7.014096383512802e-07, | |
| "loss": 0.0179, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.85342019543974, | |
| "grad_norm": 0.3035740554332733, | |
| "learning_rate": 4.6972042068341714e-07, | |
| "loss": 0.0136, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.88599348534202, | |
| "grad_norm": 0.2395792156457901, | |
| "learning_rate": 2.8423983574328295e-07, | |
| "loss": 0.0183, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.918566775244299, | |
| "grad_norm": 0.15957923233509064, | |
| "learning_rate": 1.4505399606130621e-07, | |
| "loss": 0.011, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.95114006514658, | |
| "grad_norm": 0.18425774574279785, | |
| "learning_rate": 5.2227521041470216e-08, | |
| "loss": 0.0129, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.9837133550488595, | |
| "grad_norm": 0.3651920557022095, | |
| "learning_rate": 5.803506960722072e-09, | |
| "loss": 0.0144, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1535, | |
| "total_flos": 5.197521391244208e+16, | |
| "train_loss": 0.059920015540398684, | |
| "train_runtime": 725.7561, | |
| "train_samples_per_second": 33.841, | |
| "train_steps_per_second": 2.115 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1535, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.197521391244208e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |