| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.993485342019544, | |
| "eval_steps": 500, | |
| "global_step": 1533, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03257328990228013, | |
| "grad_norm": 10.128602027893066, | |
| "learning_rate": 2.5974025974025972e-05, | |
| "loss": 1.7762, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06514657980456026, | |
| "grad_norm": 11.557284355163574, | |
| "learning_rate": 5.1948051948051944e-05, | |
| "loss": 0.5874, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09771986970684039, | |
| "grad_norm": 1.7021989822387695, | |
| "learning_rate": 7.792207792207793e-05, | |
| "loss": 0.2361, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13029315960912052, | |
| "grad_norm": 2.088670015335083, | |
| "learning_rate": 0.00010389610389610389, | |
| "loss": 0.176, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16286644951140064, | |
| "grad_norm": 2.070906400680542, | |
| "learning_rate": 0.00012987012987012987, | |
| "loss": 0.1546, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19543973941368079, | |
| "grad_norm": 1.9704210758209229, | |
| "learning_rate": 0.00015584415584415587, | |
| "loss": 0.1463, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2280130293159609, | |
| "grad_norm": 1.2319754362106323, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.1306, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26058631921824105, | |
| "grad_norm": 1.4384595155715942, | |
| "learning_rate": 0.0001999979049808708, | |
| "loss": 0.1218, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2931596091205212, | |
| "grad_norm": 1.3861066102981567, | |
| "learning_rate": 0.00019996066263830531, | |
| "loss": 0.1031, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3257328990228013, | |
| "grad_norm": 1.3069312572479248, | |
| "learning_rate": 0.00019987688427197897, | |
| "loss": 0.1125, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3583061889250814, | |
| "grad_norm": 1.1740531921386719, | |
| "learning_rate": 0.0001997466088843548, | |
| "loss": 0.0984, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.39087947882736157, | |
| "grad_norm": 0.831253170967102, | |
| "learning_rate": 0.00019956989712427577, | |
| "loss": 0.1013, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4234527687296417, | |
| "grad_norm": 1.1512802839279175, | |
| "learning_rate": 0.0001993468312587303, | |
| "loss": 0.0827, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4560260586319218, | |
| "grad_norm": 0.7890095710754395, | |
| "learning_rate": 0.00019907751513455302, | |
| "loss": 0.0747, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48859934853420195, | |
| "grad_norm": 0.8635206818580627, | |
| "learning_rate": 0.00019876207413008015, | |
| "loss": 0.0911, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5211726384364821, | |
| "grad_norm": 0.984118640422821, | |
| "learning_rate": 0.00019840065509677988, | |
| "loss": 0.0708, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5537459283387622, | |
| "grad_norm": 1.1899775266647339, | |
| "learning_rate": 0.00019799342629088702, | |
| "loss": 0.0759, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5863192182410424, | |
| "grad_norm": 0.8736202120780945, | |
| "learning_rate": 0.00019754057729507227, | |
| "loss": 0.0747, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6188925081433225, | |
| "grad_norm": 0.6165608763694763, | |
| "learning_rate": 0.0001970423189301833, | |
| "loss": 0.0697, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6514657980456026, | |
| "grad_norm": 1.2145845890045166, | |
| "learning_rate": 0.00019649888315709843, | |
| "loss": 0.0768, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6840390879478827, | |
| "grad_norm": 0.7146447896957397, | |
| "learning_rate": 0.00019591052296873888, | |
| "loss": 0.0575, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7166123778501629, | |
| "grad_norm": 1.0241068601608276, | |
| "learning_rate": 0.00019527751227228963, | |
| "loss": 0.0688, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.749185667752443, | |
| "grad_norm": 0.8579204678535461, | |
| "learning_rate": 0.00019460014576168358, | |
| "loss": 0.0544, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7817589576547231, | |
| "grad_norm": 0.5284867286682129, | |
| "learning_rate": 0.0001938787387804088, | |
| "loss": 0.0516, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8143322475570033, | |
| "grad_norm": 0.6213068962097168, | |
| "learning_rate": 0.00019311362717470268, | |
| "loss": 0.0501, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8469055374592834, | |
| "grad_norm": 0.7320097088813782, | |
| "learning_rate": 0.00019230516713720052, | |
| "loss": 0.0581, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8794788273615635, | |
| "grad_norm": 0.36288219690322876, | |
| "learning_rate": 0.00019145373504111279, | |
| "loss": 0.0478, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9120521172638436, | |
| "grad_norm": 0.6096987724304199, | |
| "learning_rate": 0.00019055972726500695, | |
| "loss": 0.0548, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9446254071661238, | |
| "grad_norm": 0.7391396760940552, | |
| "learning_rate": 0.0001896235600082759, | |
| "loss": 0.0444, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9771986970684039, | |
| "grad_norm": 0.7820376753807068, | |
| "learning_rate": 0.00018864566909737937, | |
| "loss": 0.0566, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.009771986970684, | |
| "grad_norm": 0.3828499913215637, | |
| "learning_rate": 0.00018762650978294758, | |
| "loss": 0.0533, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0423452768729642, | |
| "grad_norm": 0.49720415472984314, | |
| "learning_rate": 0.0001865665565278424, | |
| "loss": 0.0582, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0749185667752443, | |
| "grad_norm": 0.4968441128730774, | |
| "learning_rate": 0.00018546630278627437, | |
| "loss": 0.0473, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1074918566775245, | |
| "grad_norm": 0.812972366809845, | |
| "learning_rate": 0.00018432626077407829, | |
| "loss": 0.0382, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1400651465798046, | |
| "grad_norm": 0.4055006802082062, | |
| "learning_rate": 0.00018314696123025454, | |
| "loss": 0.043, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1726384364820848, | |
| "grad_norm": 0.6709649562835693, | |
| "learning_rate": 0.0001819289531698871, | |
| "loss": 0.0343, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.205211726384365, | |
| "grad_norm": 0.6343613862991333, | |
| "learning_rate": 0.0001806728036285532, | |
| "loss": 0.0388, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.237785016286645, | |
| "grad_norm": 0.44080787897109985, | |
| "learning_rate": 0.00017937909739834367, | |
| "loss": 0.0419, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2703583061889252, | |
| "grad_norm": 0.714772641658783, | |
| "learning_rate": 0.00017804843675561677, | |
| "loss": 0.0476, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3029315960912053, | |
| "grad_norm": 0.4127221703529358, | |
| "learning_rate": 0.00017668144118061262, | |
| "loss": 0.0354, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3355048859934853, | |
| "grad_norm": 0.44271162152290344, | |
| "learning_rate": 0.00017527874706905805, | |
| "loss": 0.0385, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3680781758957654, | |
| "grad_norm": 0.6225533485412598, | |
| "learning_rate": 0.00017384100743589697, | |
| "loss": 0.0665, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4006514657980456, | |
| "grad_norm": 0.4418353736400604, | |
| "learning_rate": 0.0001723688916112835, | |
| "loss": 0.0554, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4332247557003257, | |
| "grad_norm": 0.3992595672607422, | |
| "learning_rate": 0.00017086308492897983, | |
| "loss": 0.0522, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4657980456026058, | |
| "grad_norm": 0.45478880405426025, | |
| "learning_rate": 0.0001693242884073035, | |
| "loss": 0.0398, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.498371335504886, | |
| "grad_norm": 0.5688516497612, | |
| "learning_rate": 0.00016775321842277312, | |
| "loss": 0.032, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5309446254071661, | |
| "grad_norm": 0.4958943724632263, | |
| "learning_rate": 0.00016615060637660388, | |
| "loss": 0.0422, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5635179153094463, | |
| "grad_norm": 0.4975818991661072, | |
| "learning_rate": 0.00016451719835420877, | |
| "loss": 0.0317, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5960912052117264, | |
| "grad_norm": 0.5288175344467163, | |
| "learning_rate": 0.00016285375477786322, | |
| "loss": 0.0412, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6286644951140063, | |
| "grad_norm": 0.6416569352149963, | |
| "learning_rate": 0.0001611610500526957, | |
| "loss": 0.0275, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6612377850162865, | |
| "grad_norm": 0.42349156737327576, | |
| "learning_rate": 0.00015943987220616855, | |
| "loss": 0.029, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6938110749185666, | |
| "grad_norm": 0.2572161853313446, | |
| "learning_rate": 0.00015769102252121702, | |
| "loss": 0.0308, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7263843648208468, | |
| "grad_norm": 0.5944684147834778, | |
| "learning_rate": 0.0001559153151632171, | |
| "loss": 0.0354, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.758957654723127, | |
| "grad_norm": 0.5154451727867126, | |
| "learning_rate": 0.0001541135768009566, | |
| "loss": 0.0424, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.791530944625407, | |
| "grad_norm": 0.5298171043395996, | |
| "learning_rate": 0.00015228664622178467, | |
| "loss": 0.0263, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8241042345276872, | |
| "grad_norm": 0.3582160472869873, | |
| "learning_rate": 0.00015043537394112007, | |
| "loss": 0.0386, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8566775244299674, | |
| "grad_norm": 0.6350813508033752, | |
| "learning_rate": 0.0001485606218064993, | |
| "loss": 0.0444, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8892508143322475, | |
| "grad_norm": 0.6265957355499268, | |
| "learning_rate": 0.00014666326259634918, | |
| "loss": 0.0417, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9218241042345277, | |
| "grad_norm": 0.41339483857154846, | |
| "learning_rate": 0.00014474417961367065, | |
| "loss": 0.0309, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9543973941368078, | |
| "grad_norm": 0.41287094354629517, | |
| "learning_rate": 0.000142804266274823, | |
| "loss": 0.0308, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.986970684039088, | |
| "grad_norm": 0.5833337903022766, | |
| "learning_rate": 0.00014084442569359964, | |
| "loss": 0.0346, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.019543973941368, | |
| "grad_norm": 0.3714321255683899, | |
| "learning_rate": 0.00013886557026078955, | |
| "loss": 0.0337, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0521172638436482, | |
| "grad_norm": 0.2712138295173645, | |
| "learning_rate": 0.0001368686212194199, | |
| "loss": 0.0196, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0846905537459284, | |
| "grad_norm": 0.6208952069282532, | |
| "learning_rate": 0.00013485450823587725, | |
| "loss": 0.0288, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1172638436482085, | |
| "grad_norm": 0.34229573607444763, | |
| "learning_rate": 0.00013282416896710778, | |
| "loss": 0.0246, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.1498371335504887, | |
| "grad_norm": 0.4573960304260254, | |
| "learning_rate": 0.00013077854862409696, | |
| "loss": 0.0249, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.182410423452769, | |
| "grad_norm": 0.3675532042980194, | |
| "learning_rate": 0.0001287185995318333, | |
| "loss": 0.022, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.214983713355049, | |
| "grad_norm": 0.5372172594070435, | |
| "learning_rate": 0.0001266452806859594, | |
| "loss": 0.0343, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.247557003257329, | |
| "grad_norm": 0.46904900670051575, | |
| "learning_rate": 0.00012455955730631804, | |
| "loss": 0.0233, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2801302931596092, | |
| "grad_norm": 0.4394093155860901, | |
| "learning_rate": 0.00012246240038760043, | |
| "loss": 0.0209, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3127035830618894, | |
| "grad_norm": 0.3127492368221283, | |
| "learning_rate": 0.00012035478624730608, | |
| "loss": 0.0321, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3452768729641695, | |
| "grad_norm": 0.601370096206665, | |
| "learning_rate": 0.00011823769607122479, | |
| "loss": 0.0243, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3778501628664497, | |
| "grad_norm": 0.5871070623397827, | |
| "learning_rate": 0.00011611211545665184, | |
| "loss": 0.0337, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.41042345276873, | |
| "grad_norm": 0.3546801805496216, | |
| "learning_rate": 0.00011397903395354996, | |
| "loss": 0.0288, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.44299674267101, | |
| "grad_norm": 0.8319407105445862, | |
| "learning_rate": 0.0001118394446038708, | |
| "loss": 0.0337, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.47557003257329, | |
| "grad_norm": 0.5210663080215454, | |
| "learning_rate": 0.00010969434347925076, | |
| "loss": 0.026, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5081433224755703, | |
| "grad_norm": 0.5834184288978577, | |
| "learning_rate": 0.00010754472921729661, | |
| "loss": 0.0282, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5407166123778504, | |
| "grad_norm": 0.42890864610671997, | |
| "learning_rate": 0.00010539160255667623, | |
| "loss": 0.028, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5732899022801305, | |
| "grad_norm": 0.4473400413990021, | |
| "learning_rate": 0.00010323596587123145, | |
| "loss": 0.025, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.6058631921824107, | |
| "grad_norm": 0.5189303159713745, | |
| "learning_rate": 0.00010107882270332952, | |
| "loss": 0.0293, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.6384364820846904, | |
| "grad_norm": 0.43365001678466797, | |
| "learning_rate": 9.892117729667052e-05, | |
| "loss": 0.0175, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6710097719869705, | |
| "grad_norm": 0.28346696496009827, | |
| "learning_rate": 9.676403412876856e-05, | |
| "loss": 0.0334, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7035830618892507, | |
| "grad_norm": 0.3956477642059326, | |
| "learning_rate": 9.460839744332378e-05, | |
| "loss": 0.0271, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.736156351791531, | |
| "grad_norm": 0.30705949664115906, | |
| "learning_rate": 9.245527078270341e-05, | |
| "loss": 0.0217, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.768729641693811, | |
| "grad_norm": 0.40188854932785034, | |
| "learning_rate": 9.030565652074926e-05, | |
| "loss": 0.019, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.801302931596091, | |
| "grad_norm": 0.3447129428386688, | |
| "learning_rate": 8.816055539612924e-05, | |
| "loss": 0.028, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.8338762214983713, | |
| "grad_norm": 0.38768622279167175, | |
| "learning_rate": 8.602096604645009e-05, | |
| "loss": 0.0218, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8664495114006514, | |
| "grad_norm": 0.26912721991539, | |
| "learning_rate": 8.388788454334817e-05, | |
| "loss": 0.0173, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8990228013029316, | |
| "grad_norm": 0.33078861236572266, | |
| "learning_rate": 8.176230392877523e-05, | |
| "loss": 0.0233, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9315960912052117, | |
| "grad_norm": 0.24832488596439362, | |
| "learning_rate": 7.964521375269396e-05, | |
| "loss": 0.0171, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.964169381107492, | |
| "grad_norm": 0.6595136523246765, | |
| "learning_rate": 7.753759961239964e-05, | |
| "loss": 0.0272, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.996742671009772, | |
| "grad_norm": 0.2780207097530365, | |
| "learning_rate": 7.544044269368197e-05, | |
| "loss": 0.0338, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.029315960912052, | |
| "grad_norm": 0.7173179388046265, | |
| "learning_rate": 7.335471931404063e-05, | |
| "loss": 0.0365, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0618892508143323, | |
| "grad_norm": 0.33753442764282227, | |
| "learning_rate": 7.128140046816671e-05, | |
| "loss": 0.0195, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0944625407166124, | |
| "grad_norm": 0.35064950585365295, | |
| "learning_rate": 6.922145137590306e-05, | |
| "loss": 0.02, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.1270358306188926, | |
| "grad_norm": 0.39598166942596436, | |
| "learning_rate": 6.717583103289229e-05, | |
| "loss": 0.0203, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1596091205211727, | |
| "grad_norm": 0.18257524073123932, | |
| "learning_rate": 6.514549176412275e-05, | |
| "loss": 0.0134, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.192182410423453, | |
| "grad_norm": 0.4458347260951996, | |
| "learning_rate": 6.313137878058013e-05, | |
| "loss": 0.0236, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.224755700325733, | |
| "grad_norm": 0.22742605209350586, | |
| "learning_rate": 6.113442973921046e-05, | |
| "loss": 0.0208, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.257328990228013, | |
| "grad_norm": 0.1858537793159485, | |
| "learning_rate": 5.9155574306400395e-05, | |
| "loss": 0.0218, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2899022801302933, | |
| "grad_norm": 0.24626286327838898, | |
| "learning_rate": 5.7195733725176994e-05, | |
| "loss": 0.0232, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.3224755700325734, | |
| "grad_norm": 0.2719153165817261, | |
| "learning_rate": 5.525582038632934e-05, | |
| "loss": 0.0148, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.3550488599348536, | |
| "grad_norm": 0.218730166554451, | |
| "learning_rate": 5.333673740365083e-05, | |
| "loss": 0.0157, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3876221498371337, | |
| "grad_norm": 0.20292945206165314, | |
| "learning_rate": 5.1439378193500707e-05, | |
| "loss": 0.0143, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.420195439739414, | |
| "grad_norm": 0.2846449017524719, | |
| "learning_rate": 4.956462605887994e-05, | |
| "loss": 0.0177, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.4527687296416936, | |
| "grad_norm": 0.322721391916275, | |
| "learning_rate": 4.771335377821535e-05, | |
| "loss": 0.0224, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.4853420195439737, | |
| "grad_norm": 0.1719449758529663, | |
| "learning_rate": 4.588642319904343e-05, | |
| "loss": 0.0234, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.517915309446254, | |
| "grad_norm": 0.44704851508140564, | |
| "learning_rate": 4.408468483678293e-05, | |
| "loss": 0.019, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.550488599348534, | |
| "grad_norm": 0.4159814417362213, | |
| "learning_rate": 4.230897747878303e-05, | |
| "loss": 0.0156, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.583061889250814, | |
| "grad_norm": 0.19604472815990448, | |
| "learning_rate": 4.056012779383145e-05, | |
| "loss": 0.0158, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.6156351791530943, | |
| "grad_norm": 0.19116809964179993, | |
| "learning_rate": 3.883894994730428e-05, | |
| "loss": 0.0174, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.6482084690553744, | |
| "grad_norm": 0.3637801706790924, | |
| "learning_rate": 3.714624522213681e-05, | |
| "loss": 0.0162, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.6807817589576546, | |
| "grad_norm": 0.1877295821905136, | |
| "learning_rate": 3.548280164579126e-05, | |
| "loss": 0.0142, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.7133550488599347, | |
| "grad_norm": 0.1830226182937622, | |
| "learning_rate": 3.384939362339614e-05, | |
| "loss": 0.0119, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.745928338762215, | |
| "grad_norm": 0.15163740515708923, | |
| "learning_rate": 3.224678157722689e-05, | |
| "loss": 0.0181, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.778501628664495, | |
| "grad_norm": 0.2479788213968277, | |
| "learning_rate": 3.067571159269651e-05, | |
| "loss": 0.0138, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.811074918566775, | |
| "grad_norm": 0.6171669960021973, | |
| "learning_rate": 2.913691507102019e-05, | |
| "loss": 0.0197, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.8436482084690553, | |
| "grad_norm": 0.18519634008407593, | |
| "learning_rate": 2.763110838871651e-05, | |
| "loss": 0.0137, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8762214983713354, | |
| "grad_norm": 0.26303982734680176, | |
| "learning_rate": 2.6158992564103058e-05, | |
| "loss": 0.0172, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.9087947882736156, | |
| "grad_norm": 0.28331807255744934, | |
| "learning_rate": 2.4721252930941974e-05, | |
| "loss": 0.0168, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.9413680781758957, | |
| "grad_norm": 0.20530906319618225, | |
| "learning_rate": 2.3318558819387404e-05, | |
| "loss": 0.0199, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.973941368078176, | |
| "grad_norm": 0.16924133896827698, | |
| "learning_rate": 2.1951563244383233e-05, | |
| "loss": 0.0146, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.006514657980456, | |
| "grad_norm": 0.13186028599739075, | |
| "learning_rate": 2.0620902601656345e-05, | |
| "loss": 0.0124, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.039087947882736, | |
| "grad_norm": 0.24360792338848114, | |
| "learning_rate": 1.9327196371446776e-05, | |
| "loss": 0.0119, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.071661237785016, | |
| "grad_norm": 0.09876150637865067, | |
| "learning_rate": 1.807104683011289e-05, | |
| "loss": 0.012, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.1042345276872965, | |
| "grad_norm": 0.2283184826374054, | |
| "learning_rate": 1.6853038769745467e-05, | |
| "loss": 0.0142, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.136807817589577, | |
| "grad_norm": 0.32383596897125244, | |
| "learning_rate": 1.5673739225921758e-05, | |
| "loss": 0.012, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.169381107491857, | |
| "grad_norm": 0.2783248722553253, | |
| "learning_rate": 1.4533697213725662e-05, | |
| "loss": 0.0163, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.201954397394137, | |
| "grad_norm": 0.17678265273571014, | |
| "learning_rate": 1.3433443472157613e-05, | |
| "loss": 0.012, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.234527687296417, | |
| "grad_norm": 0.25102487206459045, | |
| "learning_rate": 1.237349021705243e-05, | |
| "loss": 0.0158, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.267100977198697, | |
| "grad_norm": 0.15461167693138123, | |
| "learning_rate": 1.1354330902620636e-05, | |
| "loss": 0.0126, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.299674267100977, | |
| "grad_norm": 0.24122057855129242, | |
| "learning_rate": 1.0376439991724096e-05, | |
| "loss": 0.0168, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.3322475570032575, | |
| "grad_norm": 0.14669205248355865, | |
| "learning_rate": 9.440272734993072e-06, | |
| "loss": 0.0179, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.364820846905538, | |
| "grad_norm": 0.32440969347953796, | |
| "learning_rate": 8.546264958887219e-06, | |
| "loss": 0.0197, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.397394136807818, | |
| "grad_norm": 0.14456795156002045, | |
| "learning_rate": 7.694832862799505e-06, | |
| "loss": 0.0111, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.429967426710098, | |
| "grad_norm": 0.17956456542015076, | |
| "learning_rate": 6.886372825297349e-06, | |
| "loss": 0.0085, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.462540716612378, | |
| "grad_norm": 0.30424752831459045, | |
| "learning_rate": 6.12126121959119e-06, | |
| "loss": 0.0207, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.495114006514658, | |
| "grad_norm": 0.18671758472919464, | |
| "learning_rate": 5.399854238316437e-06, | |
| "loss": 0.013, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.527687296416938, | |
| "grad_norm": 0.3565406799316406, | |
| "learning_rate": 4.722487727710368e-06, | |
| "loss": 0.0165, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.5602605863192185, | |
| "grad_norm": 0.26344749331474304, | |
| "learning_rate": 4.089477031261113e-06, | |
| "loss": 0.0148, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.592833876221499, | |
| "grad_norm": 0.18339155614376068, | |
| "learning_rate": 3.5011168429016083e-06, | |
| "loss": 0.0195, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.625407166123779, | |
| "grad_norm": 0.2598022222518921, | |
| "learning_rate": 2.95768106981672e-06, | |
| "loss": 0.0135, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.657980456026059, | |
| "grad_norm": 0.3853515684604645, | |
| "learning_rate": 2.4594227049277386e-06, | |
| "loss": 0.0177, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.690553745928339, | |
| "grad_norm": 0.13664180040359497, | |
| "learning_rate": 2.006573709112991e-06, | |
| "loss": 0.0086, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.723127035830619, | |
| "grad_norm": 0.1015399917960167, | |
| "learning_rate": 1.5993449032201458e-06, | |
| "loss": 0.0116, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.755700325732899, | |
| "grad_norm": 0.18885648250579834, | |
| "learning_rate": 1.237925869919887e-06, | |
| "loss": 0.0175, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.7882736156351795, | |
| "grad_norm": 0.18131224811077118, | |
| "learning_rate": 9.224848654469931e-07, | |
| "loss": 0.0088, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.82084690553746, | |
| "grad_norm": 0.194551482796669, | |
| "learning_rate": 6.531687412697496e-07, | |
| "loss": 0.014, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.85342019543974, | |
| "grad_norm": 0.23798178136348724, | |
| "learning_rate": 4.3010287572422537e-07, | |
| "loss": 0.0097, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.88599348534202, | |
| "grad_norm": 0.141094371676445, | |
| "learning_rate": 2.5339111564521844e-07, | |
| "loss": 0.0151, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.918566775244299, | |
| "grad_norm": 0.17839759588241577, | |
| "learning_rate": 1.2311572802105043e-07, | |
| "loss": 0.0097, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.95114006514658, | |
| "grad_norm": 0.3124100863933563, | |
| "learning_rate": 3.933736169471347e-08, | |
| "loss": 0.0098, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.9837133550488595, | |
| "grad_norm": 0.17664480209350586, | |
| "learning_rate": 2.0950191292112842e-09, | |
| "loss": 0.0142, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.993485342019544, | |
| "step": 1533, | |
| "total_flos": 5.185032946443418e+16, | |
| "train_loss": 0.05182663513868756, | |
| "train_runtime": 722.6992, | |
| "train_samples_per_second": 33.939, | |
| "train_steps_per_second": 2.121 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1533, | |
| "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": 5.185032946443418e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |