| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.949579831932773, | |
| "eval_steps": 500, | |
| "global_step": 1480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06722689075630252, | |
| "grad_norm": 3.575806140899658, | |
| "learning_rate": 0.00019997747161747695, | |
| "loss": 1.5631, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13445378151260504, | |
| "grad_norm": 0.8588754534721375, | |
| "learning_rate": 0.00019990989662046818, | |
| "loss": 1.1154, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.20168067226890757, | |
| "grad_norm": 0.7627600431442261, | |
| "learning_rate": 0.00019979730545608126, | |
| "loss": 0.9953, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2689075630252101, | |
| "grad_norm": 0.6456862092018127, | |
| "learning_rate": 0.00019963974885425266, | |
| "loss": 0.9508, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 0.6212641000747681, | |
| "learning_rate": 0.00019943729780489027, | |
| "loss": 0.9451, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.40336134453781514, | |
| "grad_norm": 0.6262118816375732, | |
| "learning_rate": 0.00019919004352588767, | |
| "loss": 0.9302, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.5837720036506653, | |
| "learning_rate": 0.00019889809742202455, | |
| "loss": 0.909, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5378151260504201, | |
| "grad_norm": 0.6171330809593201, | |
| "learning_rate": 0.00019856159103477086, | |
| "loss": 0.9109, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6050420168067226, | |
| "grad_norm": 0.6107902526855469, | |
| "learning_rate": 0.0001981806759830189, | |
| "loss": 0.8952, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 0.5945306420326233, | |
| "learning_rate": 0.00019775552389476864, | |
| "loss": 0.8968, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7394957983193278, | |
| "grad_norm": 0.5847243666648865, | |
| "learning_rate": 0.00019728632632979746, | |
| "loss": 0.8696, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8067226890756303, | |
| "grad_norm": 0.5536367893218994, | |
| "learning_rate": 0.0001967732946933499, | |
| "loss": 0.8575, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8739495798319328, | |
| "grad_norm": 0.6354172825813293, | |
| "learning_rate": 0.00019621666014088494, | |
| "loss": 0.8429, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.5658561587333679, | |
| "learning_rate": 0.00019561667347392508, | |
| "loss": 0.8376, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0084033613445378, | |
| "grad_norm": 0.5443424582481384, | |
| "learning_rate": 0.0001949736050270532, | |
| "loss": 0.834, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0756302521008403, | |
| "grad_norm": 0.6456009745597839, | |
| "learning_rate": 0.00019428774454610843, | |
| "loss": 0.7672, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.7190167903900146, | |
| "learning_rate": 0.0001935594010576362, | |
| "loss": 0.7292, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.2100840336134453, | |
| "grad_norm": 0.628425657749176, | |
| "learning_rate": 0.00019278890272965096, | |
| "loss": 0.7646, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2773109243697478, | |
| "grad_norm": 0.6629491448402405, | |
| "learning_rate": 0.0001919765967237739, | |
| "loss": 0.7627, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3445378151260505, | |
| "grad_norm": 0.6584550738334656, | |
| "learning_rate": 0.0001911228490388136, | |
| "loss": 0.7632, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.7824658751487732, | |
| "learning_rate": 0.00019022804434585852, | |
| "loss": 0.7644, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4789915966386555, | |
| "grad_norm": 0.6643643379211426, | |
| "learning_rate": 0.00018929258581495685, | |
| "loss": 0.7455, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5462184873949578, | |
| "grad_norm": 0.615719199180603, | |
| "learning_rate": 0.00018831689493346095, | |
| "loss": 0.7807, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.6134453781512605, | |
| "grad_norm": 0.6157252192497253, | |
| "learning_rate": 0.00018730141131611882, | |
| "loss": 0.7355, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.680672268907563, | |
| "grad_norm": 0.6899864673614502, | |
| "learning_rate": 0.00018624659250699805, | |
| "loss": 0.7544, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7478991596638656, | |
| "grad_norm": 0.5952817797660828, | |
| "learning_rate": 0.00018515291377333112, | |
| "loss": 0.7484, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.815126050420168, | |
| "grad_norm": 0.639409601688385, | |
| "learning_rate": 0.00018402086789137546, | |
| "loss": 0.7586, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.6305294036865234, | |
| "learning_rate": 0.00018285096492438424, | |
| "loss": 0.7429, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.949579831932773, | |
| "grad_norm": 0.6956226825714111, | |
| "learning_rate": 0.00018164373199278856, | |
| "loss": 0.7446, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.0168067226890756, | |
| "grad_norm": 0.6213023662567139, | |
| "learning_rate": 0.00018039971303669407, | |
| "loss": 0.7035, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.0840336134453783, | |
| "grad_norm": 0.7370445728302002, | |
| "learning_rate": 0.00017911946857079888, | |
| "loss": 0.6189, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.1512605042016806, | |
| "grad_norm": 0.7824037075042725, | |
| "learning_rate": 0.00017780357543184397, | |
| "loss": 0.6206, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.2184873949579833, | |
| "grad_norm": 0.792977511882782, | |
| "learning_rate": 0.00017645262651870926, | |
| "loss": 0.6176, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.7439864873886108, | |
| "learning_rate": 0.00017506723052527242, | |
| "loss": 0.621, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.833249032497406, | |
| "learning_rate": 0.00017364801166615124, | |
| "loss": 0.6165, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.4201680672268906, | |
| "grad_norm": 0.743938684463501, | |
| "learning_rate": 0.00017219560939545246, | |
| "loss": 0.6062, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.4873949579831933, | |
| "grad_norm": 0.8116819858551025, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.6283, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.5546218487394956, | |
| "grad_norm": 0.8074822425842285, | |
| "learning_rate": 0.00016919388689775464, | |
| "loss": 0.651, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.6218487394957983, | |
| "grad_norm": 0.8415045738220215, | |
| "learning_rate": 0.0001676459191498087, | |
| "loss": 0.6388, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.689075630252101, | |
| "grad_norm": 0.727366030216217, | |
| "learning_rate": 0.00016606747233900815, | |
| "loss": 0.6067, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7563025210084033, | |
| "grad_norm": 0.7842895984649658, | |
| "learning_rate": 0.00016445925766242391, | |
| "loss": 0.6408, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.7419390082359314, | |
| "learning_rate": 0.00016282199972956425, | |
| "loss": 0.6388, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.8907563025210083, | |
| "grad_norm": 0.7776601910591125, | |
| "learning_rate": 0.00016115643623588915, | |
| "loss": 0.6318, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.957983193277311, | |
| "grad_norm": 0.8350630402565002, | |
| "learning_rate": 0.00015946331763042867, | |
| "loss": 0.6174, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.0252100840336134, | |
| "grad_norm": 0.7818737030029297, | |
| "learning_rate": 0.0001577434067776548, | |
| "loss": 0.5877, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.092436974789916, | |
| "grad_norm": 0.9094923138618469, | |
| "learning_rate": 0.00015599747861375955, | |
| "loss": 0.4781, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.1596638655462184, | |
| "grad_norm": 0.9884558320045471, | |
| "learning_rate": 0.00015422631979749354, | |
| "loss": 0.475, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.226890756302521, | |
| "grad_norm": 0.9660643339157104, | |
| "learning_rate": 0.00015243072835572318, | |
| "loss": 0.4689, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.9638343453407288, | |
| "learning_rate": 0.00015061151332386566, | |
| "loss": 0.4726, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.361344537815126, | |
| "grad_norm": 0.9662260413169861, | |
| "learning_rate": 0.00014876949438136347, | |
| "loss": 0.5037, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.9045869708061218, | |
| "learning_rate": 0.0001469055014823637, | |
| "loss": 0.4875, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.495798319327731, | |
| "grad_norm": 0.9099442958831787, | |
| "learning_rate": 0.00014502037448176734, | |
| "loss": 0.5047, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.5630252100840334, | |
| "grad_norm": 0.9331455826759338, | |
| "learning_rate": 0.00014311496275681783, | |
| "loss": 0.4929, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.630252100840336, | |
| "grad_norm": 1.0234043598175049, | |
| "learning_rate": 0.0001411901248243993, | |
| "loss": 0.5033, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.697478991596639, | |
| "grad_norm": 0.9315100312232971, | |
| "learning_rate": 0.00013924672795421637, | |
| "loss": 0.4925, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 1.0947026014328003, | |
| "learning_rate": 0.00013728564777803088, | |
| "loss": 0.5003, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.831932773109244, | |
| "grad_norm": 0.9220852255821228, | |
| "learning_rate": 0.0001353077678951301, | |
| "loss": 0.51, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.899159663865546, | |
| "grad_norm": 1.065261721611023, | |
| "learning_rate": 0.00013331397947420576, | |
| "loss": 0.5091, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.966386554621849, | |
| "grad_norm": 0.8783352971076965, | |
| "learning_rate": 0.00013130518085182225, | |
| "loss": 0.5036, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.033613445378151, | |
| "grad_norm": 0.9211850166320801, | |
| "learning_rate": 0.00012928227712765504, | |
| "loss": 0.4408, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.100840336134453, | |
| "grad_norm": 1.0706185102462769, | |
| "learning_rate": 0.0001272461797566823, | |
| "loss": 0.3445, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.168067226890757, | |
| "grad_norm": 1.152872920036316, | |
| "learning_rate": 0.00012519780613851254, | |
| "loss": 0.3399, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 1.1418325901031494, | |
| "learning_rate": 0.00012313807920403419, | |
| "loss": 0.3516, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.302521008403361, | |
| "grad_norm": 1.178936243057251, | |
| "learning_rate": 0.00012106792699957263, | |
| "loss": 0.3605, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.369747899159664, | |
| "grad_norm": 0.9696270823478699, | |
| "learning_rate": 0.00011898828226874284, | |
| "loss": 0.3519, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.436974789915967, | |
| "grad_norm": 1.0649495124816895, | |
| "learning_rate": 0.00011690008203218493, | |
| "loss": 0.3534, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.504201680672269, | |
| "grad_norm": 1.114478588104248, | |
| "learning_rate": 0.00011480426716537315, | |
| "loss": 0.3624, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 1.2927486896514893, | |
| "learning_rate": 0.00011270178197468789, | |
| "loss": 0.3614, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 4.6386554621848735, | |
| "grad_norm": 1.2928136587142944, | |
| "learning_rate": 0.00011059357377194161, | |
| "loss": 0.3616, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 1.0552875995635986, | |
| "learning_rate": 0.00010848059244755093, | |
| "loss": 0.3666, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.773109243697479, | |
| "grad_norm": 1.0149446725845337, | |
| "learning_rate": 0.00010636379004254664, | |
| "loss": 0.3747, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 4.840336134453781, | |
| "grad_norm": 1.09207022190094, | |
| "learning_rate": 0.00010424412031961484, | |
| "loss": 0.3589, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 4.907563025210084, | |
| "grad_norm": 0.8984358906745911, | |
| "learning_rate": 0.00010212253833336237, | |
| "loss": 0.3671, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 4.974789915966387, | |
| "grad_norm": 1.0369324684143066, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3785, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.042016806722689, | |
| "grad_norm": 1.0416890382766724, | |
| "learning_rate": 9.787746166663764e-05, | |
| "loss": 0.3085, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 5.109243697478991, | |
| "grad_norm": 0.8845974802970886, | |
| "learning_rate": 9.57558796803852e-05, | |
| "loss": 0.2516, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 1.058775544166565, | |
| "learning_rate": 9.363620995745337e-05, | |
| "loss": 0.2539, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 5.243697478991597, | |
| "grad_norm": 1.0327303409576416, | |
| "learning_rate": 9.151940755244912e-05, | |
| "loss": 0.2488, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 5.310924369747899, | |
| "grad_norm": 0.9804151654243469, | |
| "learning_rate": 8.94064262280584e-05, | |
| "loss": 0.2538, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 5.378151260504202, | |
| "grad_norm": 0.9500736594200134, | |
| "learning_rate": 8.729821802531212e-05, | |
| "loss": 0.2536, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.445378151260504, | |
| "grad_norm": 1.1535006761550903, | |
| "learning_rate": 8.519573283462687e-05, | |
| "loss": 0.2544, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 5.512605042016807, | |
| "grad_norm": 0.8586835861206055, | |
| "learning_rate": 8.309991796781511e-05, | |
| "loss": 0.2636, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 5.579831932773109, | |
| "grad_norm": 0.9877124428749084, | |
| "learning_rate": 8.101171773125716e-05, | |
| "loss": 0.2548, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 1.0656546354293823, | |
| "learning_rate": 7.89320730004274e-05, | |
| "loss": 0.2681, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 1.0215117931365967, | |
| "learning_rate": 7.686192079596586e-05, | |
| "loss": 0.2639, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.781512605042017, | |
| "grad_norm": 0.9939188361167908, | |
| "learning_rate": 7.48021938614875e-05, | |
| "loss": 0.2583, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 5.848739495798319, | |
| "grad_norm": 1.1241389513015747, | |
| "learning_rate": 7.275382024331772e-05, | |
| "loss": 0.2575, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 5.915966386554622, | |
| "grad_norm": 1.2222434282302856, | |
| "learning_rate": 7.071772287234497e-05, | |
| "loss": 0.2628, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 5.983193277310924, | |
| "grad_norm": 1.15309476852417, | |
| "learning_rate": 6.869481914817779e-05, | |
| "loss": 0.2662, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 6.050420168067227, | |
| "grad_norm": 0.7435505986213684, | |
| "learning_rate": 6.668602052579424e-05, | |
| "loss": 0.2097, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.9408479928970337, | |
| "learning_rate": 6.469223210486992e-05, | |
| "loss": 0.1896, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 6.184873949579832, | |
| "grad_norm": 0.7466050386428833, | |
| "learning_rate": 6.271435222196916e-05, | |
| "loss": 0.1881, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 6.2521008403361344, | |
| "grad_norm": 0.7878915667533875, | |
| "learning_rate": 6.0753272045783625e-05, | |
| "loss": 0.1921, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 6.319327731092437, | |
| "grad_norm": 0.7960318922996521, | |
| "learning_rate": 5.880987517560075e-05, | |
| "loss": 0.1957, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 6.38655462184874, | |
| "grad_norm": 0.8870383501052856, | |
| "learning_rate": 5.688503724318217e-05, | |
| "loss": 0.1954, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.453781512605042, | |
| "grad_norm": 0.8262315392494202, | |
| "learning_rate": 5.497962551823266e-05, | |
| "loss": 0.1946, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.5210084033613445, | |
| "grad_norm": 0.8804875612258911, | |
| "learning_rate": 5.309449851763633e-05, | |
| "loss": 0.1932, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 0.9018622040748596, | |
| "learning_rate": 5.123050561863657e-05, | |
| "loss": 0.1982, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 6.65546218487395, | |
| "grad_norm": 0.7731183171272278, | |
| "learning_rate": 4.938848667613436e-05, | |
| "loss": 0.2002, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 6.722689075630252, | |
| "grad_norm": 0.9117534160614014, | |
| "learning_rate": 4.756927164427685e-05, | |
| "loss": 0.1983, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.7899159663865545, | |
| "grad_norm": 1.0696296691894531, | |
| "learning_rate": 4.57736802025065e-05, | |
| "loss": 0.1984, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 0.7894629240036011, | |
| "learning_rate": 4.4002521386240466e-05, | |
| "loss": 0.2018, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 6.92436974789916, | |
| "grad_norm": 1.0283385515213013, | |
| "learning_rate": 4.2256593222345185e-05, | |
| "loss": 0.1979, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 6.991596638655462, | |
| "grad_norm": 0.8787071108818054, | |
| "learning_rate": 4.053668236957134e-05, | |
| "loss": 0.1975, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.7579413056373596, | |
| "learning_rate": 3.884356376411089e-05, | |
| "loss": 0.1682, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 7.126050420168067, | |
| "grad_norm": 0.7519951462745667, | |
| "learning_rate": 3.717800027043576e-05, | |
| "loss": 0.1653, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 7.19327731092437, | |
| "grad_norm": 0.7462877631187439, | |
| "learning_rate": 3.554074233757608e-05, | |
| "loss": 0.1643, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 7.260504201680672, | |
| "grad_norm": 0.7198042273521423, | |
| "learning_rate": 3.393252766099187e-05, | |
| "loss": 0.16, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 7.3277310924369745, | |
| "grad_norm": 0.6849682927131653, | |
| "learning_rate": 3.2354080850191324e-05, | |
| "loss": 0.1628, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 7.394957983193278, | |
| "grad_norm": 0.7326401472091675, | |
| "learning_rate": 3.080611310224539e-05, | |
| "loss": 0.1614, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.46218487394958, | |
| "grad_norm": 0.8447051048278809, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.1608, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 0.6565689444541931, | |
| "learning_rate": 2.7804390604547557e-05, | |
| "loss": 0.1605, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 7.5966386554621845, | |
| "grad_norm": 0.6809631586074829, | |
| "learning_rate": 2.6351988333848788e-05, | |
| "loss": 0.1656, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 7.663865546218488, | |
| "grad_norm": 0.7617256045341492, | |
| "learning_rate": 2.493276947472756e-05, | |
| "loss": 0.1594, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 7.73109243697479, | |
| "grad_norm": 0.7275475263595581, | |
| "learning_rate": 2.354737348129077e-05, | |
| "loss": 0.1649, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.798319327731092, | |
| "grad_norm": 0.7503297924995422, | |
| "learning_rate": 2.2196424568156073e-05, | |
| "loss": 0.1622, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 7.865546218487395, | |
| "grad_norm": 0.6865005493164062, | |
| "learning_rate": 2.0880531429201145e-05, | |
| "loss": 0.1644, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 7.932773109243698, | |
| "grad_norm": 0.8243385553359985, | |
| "learning_rate": 1.9600286963305957e-05, | |
| "loss": 0.1618, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.7706124782562256, | |
| "learning_rate": 1.835626800721144e-05, | |
| "loss": 0.1626, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 8.067226890756302, | |
| "grad_norm": 0.5460121631622314, | |
| "learning_rate": 1.7149035075615794e-05, | |
| "loss": 0.1438, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.134453781512605, | |
| "grad_norm": 0.6267870664596558, | |
| "learning_rate": 1.5979132108624574e-05, | |
| "loss": 0.1474, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 8.201680672268907, | |
| "grad_norm": 0.671909511089325, | |
| "learning_rate": 1.4847086226668872e-05, | |
| "loss": 0.1447, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 8.268907563025211, | |
| "grad_norm": 0.5568744540214539, | |
| "learning_rate": 1.3753407493001968e-05, | |
| "loss": 0.1453, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 8.336134453781513, | |
| "grad_norm": 0.6606815457344055, | |
| "learning_rate": 1.2698588683881186e-05, | |
| "loss": 0.1448, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 8.403361344537815, | |
| "grad_norm": 0.7296724319458008, | |
| "learning_rate": 1.1683105066539068e-05, | |
| "loss": 0.1473, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 0.6032551527023315, | |
| "learning_rate": 1.0707414185043163e-05, | |
| "loss": 0.1421, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 8.53781512605042, | |
| "grad_norm": 0.6723304986953735, | |
| "learning_rate": 9.771955654141496e-06, | |
| "loss": 0.1484, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 8.605042016806722, | |
| "grad_norm": 0.6945676207542419, | |
| "learning_rate": 8.87715096118642e-06, | |
| "loss": 0.1457, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 8.672268907563025, | |
| "grad_norm": 0.6513916254043579, | |
| "learning_rate": 8.023403276226126e-06, | |
| "loss": 0.1495, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 8.739495798319329, | |
| "grad_norm": 0.660636842250824, | |
| "learning_rate": 7.211097270349066e-06, | |
| "loss": 0.1448, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.806722689075631, | |
| "grad_norm": 0.625892698764801, | |
| "learning_rate": 6.440598942363796e-06, | |
| "loss": 0.148, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 8.873949579831933, | |
| "grad_norm": 0.6458612084388733, | |
| "learning_rate": 5.71225545389158e-06, | |
| "loss": 0.1458, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.6204715967178345, | |
| "learning_rate": 5.026394972946813e-06, | |
| "loss": 0.1483, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 9.008403361344538, | |
| "grad_norm": 0.5286262035369873, | |
| "learning_rate": 4.383326526074916e-06, | |
| "loss": 0.1445, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 9.07563025210084, | |
| "grad_norm": 0.6546939611434937, | |
| "learning_rate": 3.783339859115065e-06, | |
| "loss": 0.1417, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 0.5531789064407349, | |
| "learning_rate": 3.226705306650113e-06, | |
| "loss": 0.1408, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 9.210084033613445, | |
| "grad_norm": 0.5917507410049438, | |
| "learning_rate": 2.7136736702025433e-06, | |
| "loss": 0.1414, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 9.277310924369749, | |
| "grad_norm": 0.5545393228530884, | |
| "learning_rate": 2.2444761052313856e-06, | |
| "loss": 0.1413, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 9.344537815126051, | |
| "grad_norm": 0.599034309387207, | |
| "learning_rate": 1.8193240169810943e-06, | |
| "loss": 0.1376, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.5788201093673706, | |
| "learning_rate": 1.4384089652291543e-06, | |
| "loss": 0.1391, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.478991596638656, | |
| "grad_norm": 0.5973650217056274, | |
| "learning_rate": 1.1019025779754666e-06, | |
| "loss": 0.1388, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 9.546218487394958, | |
| "grad_norm": 0.532236635684967, | |
| "learning_rate": 8.099564741123166e-07, | |
| "loss": 0.1373, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 9.61344537815126, | |
| "grad_norm": 0.6429956555366516, | |
| "learning_rate": 5.627021951097545e-07, | |
| "loss": 0.1376, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 9.680672268907562, | |
| "grad_norm": 0.6416929960250854, | |
| "learning_rate": 3.6025114574734785e-07, | |
| "loss": 0.1372, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 9.747899159663866, | |
| "grad_norm": 0.8270869255065918, | |
| "learning_rate": 2.0269454391874666e-07, | |
| "loss": 0.1394, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.815126050420169, | |
| "grad_norm": 0.6364122629165649, | |
| "learning_rate": 9.010337953185843e-08, | |
| "loss": 0.1412, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 0.6194050312042236, | |
| "learning_rate": 2.2528382523057113e-08, | |
| "loss": 0.1399, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 9.949579831932773, | |
| "grad_norm": 0.5904578566551208, | |
| "learning_rate": 0.0, | |
| "loss": 0.1422, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 9.949579831932773, | |
| "step": 1480, | |
| "total_flos": 7.279937555791872e+16, | |
| "train_loss": 0.41079871348432595, | |
| "train_runtime": 5840.149, | |
| "train_samples_per_second": 1.019, | |
| "train_steps_per_second": 0.253 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7.279937555791872e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |