| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.001626016260163, | |
| "eval_steps": 500, | |
| "global_step": 3076, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 11.395153045654297, | |
| "learning_rate": 1.2987012987012986e-05, | |
| "loss": 1.6854, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 3.7914836406707764, | |
| "learning_rate": 2.5974025974025972e-05, | |
| "loss": 0.7918, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 2.904379367828369, | |
| "learning_rate": 3.8961038961038966e-05, | |
| "loss": 0.3595, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 1.9180514812469482, | |
| "learning_rate": 5.1948051948051944e-05, | |
| "loss": 0.2492, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 1.5458292961120605, | |
| "learning_rate": 6.493506493506494e-05, | |
| "loss": 0.192, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 2.6975796222686768, | |
| "learning_rate": 7.792207792207793e-05, | |
| "loss": 0.1904, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 1.1650338172912598, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.1408, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 1.7408723831176758, | |
| "learning_rate": 0.00010389610389610389, | |
| "loss": 0.142, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 1.0142450332641602, | |
| "learning_rate": 0.00011688311688311689, | |
| "loss": 0.1426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 1.9548156261444092, | |
| "learning_rate": 0.00012987012987012987, | |
| "loss": 0.1518, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 1.2884893417358398, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.1255, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 0.9044198393821716, | |
| "learning_rate": 0.00015584415584415587, | |
| "loss": 0.1339, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21138211382113822, | |
| "grad_norm": 0.9501888155937195, | |
| "learning_rate": 0.00016883116883116884, | |
| "loss": 0.1031, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 1.059137225151062, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.1248, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 1.6067619323730469, | |
| "learning_rate": 0.0001948051948051948, | |
| "loss": 0.1125, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 0.800540030002594, | |
| "learning_rate": 0.00019999791929590824, | |
| "loss": 0.1226, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2764227642276423, | |
| "grad_norm": 0.8457431793212891, | |
| "learning_rate": 0.00019998520419557735, | |
| "loss": 0.117, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 0.8381863236427307, | |
| "learning_rate": 0.00019996093140964042, | |
| "loss": 0.097, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3089430894308943, | |
| "grad_norm": 0.6342546939849854, | |
| "learning_rate": 0.00019992510374388481, | |
| "loss": 0.1145, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 1.3920717239379883, | |
| "learning_rate": 0.0001998777253397723, | |
| "loss": 0.1136, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34146341463414637, | |
| "grad_norm": 0.6389938592910767, | |
| "learning_rate": 0.0001998188016739602, | |
| "loss": 0.1014, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 1.111118197441101, | |
| "learning_rate": 0.0001997483395576683, | |
| "loss": 0.1186, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37398373983739835, | |
| "grad_norm": 0.8862912058830261, | |
| "learning_rate": 0.00019966634713589157, | |
| "loss": 0.1037, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3902439024390244, | |
| "grad_norm": 0.47783714532852173, | |
| "learning_rate": 0.00019957283388645874, | |
| "loss": 0.0986, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4065040650406504, | |
| "grad_norm": 0.7350404858589172, | |
| "learning_rate": 0.00019946781061893647, | |
| "loss": 0.0938, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42276422764227645, | |
| "grad_norm": 1.0116444826126099, | |
| "learning_rate": 0.0001993512894733801, | |
| "loss": 0.1017, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43902439024390244, | |
| "grad_norm": 0.6825156211853027, | |
| "learning_rate": 0.0001992232839189301, | |
| "loss": 0.0956, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.45528455284552843, | |
| "grad_norm": 1.1011830568313599, | |
| "learning_rate": 0.00019908380875225534, | |
| "loss": 0.0993, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4715447154471545, | |
| "grad_norm": 0.560863196849823, | |
| "learning_rate": 0.00019893288009584256, | |
| "loss": 0.0784, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 0.8265031576156616, | |
| "learning_rate": 0.00019877051539613264, | |
| "loss": 0.0931, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5040650406504065, | |
| "grad_norm": 0.5396478772163391, | |
| "learning_rate": 0.0001985967334215041, | |
| "loss": 0.0804, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5203252032520326, | |
| "grad_norm": 0.5586859583854675, | |
| "learning_rate": 0.0001984115542601034, | |
| "loss": 0.0889, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5365853658536586, | |
| "grad_norm": 0.9515529870986938, | |
| "learning_rate": 0.0001982149993175229, | |
| "loss": 0.0915, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5528455284552846, | |
| "grad_norm": 0.9865338206291199, | |
| "learning_rate": 0.00019800709131432667, | |
| "loss": 0.0765, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5691056910569106, | |
| "grad_norm": 0.46598589420318604, | |
| "learning_rate": 0.00019778785428342386, | |
| "loss": 0.0844, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.2223691940307617, | |
| "learning_rate": 0.00019755731356729093, | |
| "loss": 0.0797, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6016260162601627, | |
| "grad_norm": 0.568372905254364, | |
| "learning_rate": 0.00019731549581504193, | |
| "loss": 0.0713, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6178861788617886, | |
| "grad_norm": 0.9440287947654724, | |
| "learning_rate": 0.00019706242897934824, | |
| "loss": 0.0859, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6341463414634146, | |
| "grad_norm": 0.9875515699386597, | |
| "learning_rate": 0.00019679814231320734, | |
| "loss": 0.0959, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6504065040650406, | |
| "grad_norm": 1.18289315700531, | |
| "learning_rate": 0.00019652266636656132, | |
| "loss": 0.0794, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.8284242749214172, | |
| "learning_rate": 0.00019623603298276544, | |
| "loss": 0.0807, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6829268292682927, | |
| "grad_norm": 0.5042060017585754, | |
| "learning_rate": 0.00019593827529490741, | |
| "loss": 0.0798, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6991869918699187, | |
| "grad_norm": 0.4532330632209778, | |
| "learning_rate": 0.00019562942772197723, | |
| "loss": 0.0704, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7154471544715447, | |
| "grad_norm": 0.6993445754051208, | |
| "learning_rate": 0.00019530952596488857, | |
| "loss": 0.0829, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7317073170731707, | |
| "grad_norm": 0.7966378927230835, | |
| "learning_rate": 0.00019497860700235206, | |
| "loss": 0.0716, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7479674796747967, | |
| "grad_norm": 0.6459921002388, | |
| "learning_rate": 0.00019463670908660075, | |
| "loss": 0.0695, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7642276422764228, | |
| "grad_norm": 0.53463214635849, | |
| "learning_rate": 0.00019428387173896837, | |
| "loss": 0.071, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7804878048780488, | |
| "grad_norm": 0.5773355960845947, | |
| "learning_rate": 0.0001939201357453208, | |
| "loss": 0.0619, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7967479674796748, | |
| "grad_norm": 0.5577520728111267, | |
| "learning_rate": 0.00019354554315134166, | |
| "loss": 0.0643, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8130081300813008, | |
| "grad_norm": 0.6475623846054077, | |
| "learning_rate": 0.0001931601372576719, | |
| "loss": 0.06, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8292682926829268, | |
| "grad_norm": 0.6802220344543457, | |
| "learning_rate": 0.00019276396261490465, | |
| "loss": 0.0617, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8455284552845529, | |
| "grad_norm": 0.41168370842933655, | |
| "learning_rate": 0.0001923570650184354, | |
| "loss": 0.0529, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8617886178861789, | |
| "grad_norm": 0.5927830338478088, | |
| "learning_rate": 0.00019193949150316826, | |
| "loss": 0.0747, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8780487804878049, | |
| "grad_norm": 0.45662933588027954, | |
| "learning_rate": 0.00019151129033807908, | |
| "loss": 0.0644, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8943089430894309, | |
| "grad_norm": 0.5423726439476013, | |
| "learning_rate": 0.0001910725110206358, | |
| "loss": 0.0541, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9105691056910569, | |
| "grad_norm": 0.4179861843585968, | |
| "learning_rate": 0.00019062320427107698, | |
| "loss": 0.0558, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.926829268292683, | |
| "grad_norm": 0.9079977869987488, | |
| "learning_rate": 0.0001901634220265486, | |
| "loss": 0.0749, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.943089430894309, | |
| "grad_norm": 0.7620092034339905, | |
| "learning_rate": 0.00018969321743510066, | |
| "loss": 0.0692, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.959349593495935, | |
| "grad_norm": 0.6436122059822083, | |
| "learning_rate": 0.00018921264484954344, | |
| "loss": 0.0571, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 1.0210789442062378, | |
| "learning_rate": 0.00018872175982116482, | |
| "loss": 0.0646, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.991869918699187, | |
| "grad_norm": 0.7681993246078491, | |
| "learning_rate": 0.00018822061909330864, | |
| "loss": 0.0716, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.008130081300813, | |
| "grad_norm": 0.6054890751838684, | |
| "learning_rate": 0.00018770928059481574, | |
| "loss": 0.0656, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.024390243902439, | |
| "grad_norm": 0.5673167705535889, | |
| "learning_rate": 0.00018718780343332755, | |
| "loss": 0.059, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.040650406504065, | |
| "grad_norm": 0.5374870896339417, | |
| "learning_rate": 0.00018665624788845387, | |
| "loss": 0.056, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.056910569105691, | |
| "grad_norm": 0.544373095035553, | |
| "learning_rate": 0.0001861146754048045, | |
| "loss": 0.0646, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0731707317073171, | |
| "grad_norm": 0.4903548061847687, | |
| "learning_rate": 0.00018556314858488707, | |
| "loss": 0.0618, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.089430894308943, | |
| "grad_norm": 0.4946286082267761, | |
| "learning_rate": 0.0001850017311818702, | |
| "loss": 0.0484, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.1056910569105691, | |
| "grad_norm": 0.46937233209609985, | |
| "learning_rate": 0.00018443048809221424, | |
| "loss": 0.0651, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.1219512195121952, | |
| "grad_norm": 0.6026718020439148, | |
| "learning_rate": 0.0001838494853481695, | |
| "loss": 0.0469, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.1382113821138211, | |
| "grad_norm": 0.5334330201148987, | |
| "learning_rate": 0.00018325879011014352, | |
| "loss": 0.0527, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1544715447154472, | |
| "grad_norm": 0.6381818652153015, | |
| "learning_rate": 0.00018265847065893737, | |
| "loss": 0.0621, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.170731707317073, | |
| "grad_norm": 0.9406611919403076, | |
| "learning_rate": 0.00018204859638785328, | |
| "loss": 0.055, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1869918699186992, | |
| "grad_norm": 0.4635262191295624, | |
| "learning_rate": 0.0001814292377946727, | |
| "loss": 0.0594, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.203252032520325, | |
| "grad_norm": 0.5755775570869446, | |
| "learning_rate": 0.00018080046647350756, | |
| "loss": 0.0497, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.2195121951219512, | |
| "grad_norm": 0.4675120413303375, | |
| "learning_rate": 0.00018016235510652425, | |
| "loss": 0.0498, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2357723577235773, | |
| "grad_norm": 0.6721792817115784, | |
| "learning_rate": 0.0001795149774555421, | |
| "loss": 0.0562, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.2520325203252032, | |
| "grad_norm": 0.416792631149292, | |
| "learning_rate": 0.00017885840835350674, | |
| "loss": 0.051, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2682926829268293, | |
| "grad_norm": 0.2711058259010315, | |
| "learning_rate": 0.00017819272369584016, | |
| "loss": 0.0549, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2845528455284554, | |
| "grad_norm": 0.6742275357246399, | |
| "learning_rate": 0.00017751800043166744, | |
| "loss": 0.0456, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.3008130081300813, | |
| "grad_norm": 0.43051138520240784, | |
| "learning_rate": 0.000176834316554922, | |
| "loss": 0.0488, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3170731707317074, | |
| "grad_norm": 0.7541930079460144, | |
| "learning_rate": 0.00017614175109532997, | |
| "loss": 0.0531, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.46846646070480347, | |
| "learning_rate": 0.00017544038410927476, | |
| "loss": 0.0632, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.3495934959349594, | |
| "grad_norm": 0.5363613963127136, | |
| "learning_rate": 0.00017473029667054298, | |
| "loss": 0.0546, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.3658536585365852, | |
| "grad_norm": 0.5677704811096191, | |
| "learning_rate": 0.00017401157086095317, | |
| "loss": 0.0536, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3821138211382114, | |
| "grad_norm": 0.4499371647834778, | |
| "learning_rate": 0.00017328428976086702, | |
| "loss": 0.0549, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3983739837398375, | |
| "grad_norm": 0.42111602425575256, | |
| "learning_rate": 0.00017254853743958642, | |
| "loss": 0.0576, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.4146341463414633, | |
| "grad_norm": 0.602150022983551, | |
| "learning_rate": 0.00017180439894563497, | |
| "loss": 0.0522, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.4308943089430894, | |
| "grad_norm": 0.5597222447395325, | |
| "learning_rate": 0.00017105196029692743, | |
| "loss": 0.0526, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.4471544715447155, | |
| "grad_norm": 0.3994801342487335, | |
| "learning_rate": 0.00017029130847082615, | |
| "loss": 0.0451, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.4634146341463414, | |
| "grad_norm": 0.36083677411079407, | |
| "learning_rate": 0.00016952253139408723, | |
| "loss": 0.0467, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4796747967479675, | |
| "grad_norm": 0.4908004105091095, | |
| "learning_rate": 0.00016874571793269665, | |
| "loss": 0.0519, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4959349593495934, | |
| "grad_norm": 0.3880288600921631, | |
| "learning_rate": 0.0001679609578815979, | |
| "loss": 0.0603, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.5121951219512195, | |
| "grad_norm": 0.30312085151672363, | |
| "learning_rate": 0.00016716834195431223, | |
| "loss": 0.0651, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.5284552845528454, | |
| "grad_norm": 0.5956284999847412, | |
| "learning_rate": 0.00016636796177245278, | |
| "loss": 0.0431, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.5447154471544715, | |
| "grad_norm": 0.4647449553012848, | |
| "learning_rate": 0.0001655599098551335, | |
| "loss": 0.0455, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.5609756097560976, | |
| "grad_norm": 0.3006104528903961, | |
| "learning_rate": 0.00016474427960827472, | |
| "loss": 0.0575, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5772357723577235, | |
| "grad_norm": 0.37439459562301636, | |
| "learning_rate": 0.0001639211653138059, | |
| "loss": 0.05, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5934959349593496, | |
| "grad_norm": 0.40915167331695557, | |
| "learning_rate": 0.00016309066211876708, | |
| "loss": 0.0439, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.6097560975609757, | |
| "grad_norm": 0.4699673652648926, | |
| "learning_rate": 0.00016225286602431063, | |
| "loss": 0.0475, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.6260162601626016, | |
| "grad_norm": 0.5752647519111633, | |
| "learning_rate": 0.00016140787387460405, | |
| "loss": 0.0501, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6422764227642277, | |
| "grad_norm": 0.294344037771225, | |
| "learning_rate": 0.0001605557833456354, | |
| "loss": 0.0659, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.6585365853658538, | |
| "grad_norm": 0.4021753668785095, | |
| "learning_rate": 0.0001596966929339224, | |
| "loss": 0.0401, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6747967479674797, | |
| "grad_norm": 0.5346744656562805, | |
| "learning_rate": 0.00015883070194512694, | |
| "loss": 0.0428, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6910569105691056, | |
| "grad_norm": 0.5238606929779053, | |
| "learning_rate": 0.0001579579104825761, | |
| "loss": 0.0402, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.7073170731707317, | |
| "grad_norm": 0.4844679534435272, | |
| "learning_rate": 0.00015707841943569037, | |
| "loss": 0.0445, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.7235772357723578, | |
| "grad_norm": 0.603135347366333, | |
| "learning_rate": 0.000156192330468322, | |
| "loss": 0.0544, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.7398373983739837, | |
| "grad_norm": 0.40627196431159973, | |
| "learning_rate": 0.00015529974600700275, | |
| "loss": 0.0461, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.7560975609756098, | |
| "grad_norm": 0.5749149918556213, | |
| "learning_rate": 0.0001544007692291044, | |
| "loss": 0.0434, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.7723577235772359, | |
| "grad_norm": 0.39188066124916077, | |
| "learning_rate": 0.0001534955040509119, | |
| "loss": 0.0414, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.7886178861788617, | |
| "grad_norm": 0.22225162386894226, | |
| "learning_rate": 0.00015258405511561115, | |
| "loss": 0.0363, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8048780487804879, | |
| "grad_norm": 0.394105464220047, | |
| "learning_rate": 0.0001516665277811932, | |
| "loss": 0.0407, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.821138211382114, | |
| "grad_norm": 0.2789241075515747, | |
| "learning_rate": 0.00015074302810827515, | |
| "loss": 0.035, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.8373983739837398, | |
| "grad_norm": 0.5585935115814209, | |
| "learning_rate": 0.00014981366284784058, | |
| "loss": 0.0433, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.8536585365853657, | |
| "grad_norm": 0.7364455461502075, | |
| "learning_rate": 0.00014887853942889927, | |
| "loss": 0.0371, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8699186991869918, | |
| "grad_norm": 0.46286317706108093, | |
| "learning_rate": 0.0001479377659460695, | |
| "loss": 0.0492, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.886178861788618, | |
| "grad_norm": 0.470198392868042, | |
| "learning_rate": 0.00014699145114708287, | |
| "loss": 0.038, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.9024390243902438, | |
| "grad_norm": 0.5939562320709229, | |
| "learning_rate": 0.00014603970442021347, | |
| "loss": 0.0507, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.91869918699187, | |
| "grad_norm": 0.4777078926563263, | |
| "learning_rate": 0.00014508263578163358, | |
| "loss": 0.0443, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.934959349593496, | |
| "grad_norm": 0.4176855981349945, | |
| "learning_rate": 0.00014412035586269624, | |
| "loss": 0.0463, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.951219512195122, | |
| "grad_norm": 0.493206262588501, | |
| "learning_rate": 0.0001431529758971471, | |
| "loss": 0.0557, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.967479674796748, | |
| "grad_norm": 0.44188496470451355, | |
| "learning_rate": 0.00014218060770826636, | |
| "loss": 0.0561, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.9837398373983741, | |
| "grad_norm": 0.5256350636482239, | |
| "learning_rate": 0.00014120336369594265, | |
| "loss": 0.0504, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.2594618499279022, | |
| "learning_rate": 0.0001402213568236804, | |
| "loss": 0.0403, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.016260162601626, | |
| "grad_norm": 0.36164024472236633, | |
| "learning_rate": 0.00013923470060554186, | |
| "loss": 0.047, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.032520325203252, | |
| "grad_norm": 0.44333750009536743, | |
| "learning_rate": 0.00013824350909302552, | |
| "loss": 0.0368, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.048780487804878, | |
| "grad_norm": 0.6669727563858032, | |
| "learning_rate": 0.00013724789686188267, | |
| "loss": 0.0305, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.065040650406504, | |
| "grad_norm": 0.32740581035614014, | |
| "learning_rate": 0.0001362479789988727, | |
| "loss": 0.0347, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.08130081300813, | |
| "grad_norm": 0.5777595043182373, | |
| "learning_rate": 0.00013524387108846032, | |
| "loss": 0.0476, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.097560975609756, | |
| "grad_norm": 0.30891892313957214, | |
| "learning_rate": 0.0001342356891994542, | |
| "loss": 0.0437, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.113821138211382, | |
| "grad_norm": 0.4573158919811249, | |
| "learning_rate": 0.00013322354987159045, | |
| "loss": 0.0421, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.130081300813008, | |
| "grad_norm": 0.42825648188591003, | |
| "learning_rate": 0.00013220757010206112, | |
| "loss": 0.0405, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.1463414634146343, | |
| "grad_norm": 0.26400211453437805, | |
| "learning_rate": 0.00013118786733199014, | |
| "loss": 0.0385, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.16260162601626, | |
| "grad_norm": 0.42198774218559265, | |
| "learning_rate": 0.00013016455943285786, | |
| "loss": 0.0464, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.178861788617886, | |
| "grad_norm": 0.41942107677459717, | |
| "learning_rate": 0.0001291377646928757, | |
| "loss": 0.0514, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.1951219512195124, | |
| "grad_norm": 0.2822849154472351, | |
| "learning_rate": 0.00012810760180331288, | |
| "loss": 0.0365, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.2113821138211383, | |
| "grad_norm": 0.41634827852249146, | |
| "learning_rate": 0.00012707418984477637, | |
| "loss": 0.0497, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.227642276422764, | |
| "grad_norm": 0.3224164545536041, | |
| "learning_rate": 0.0001260376482734458, | |
| "loss": 0.0379, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.2439024390243905, | |
| "grad_norm": 0.34206849336624146, | |
| "learning_rate": 0.00012499809690726522, | |
| "loss": 0.0338, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.2601626016260163, | |
| "grad_norm": 0.364622563123703, | |
| "learning_rate": 0.00012395565591209273, | |
| "loss": 0.0422, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.2764227642276422, | |
| "grad_norm": 0.28372102975845337, | |
| "learning_rate": 0.00012291044578781015, | |
| "loss": 0.0369, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.292682926829268, | |
| "grad_norm": 0.6621326804161072, | |
| "learning_rate": 0.00012186258735439379, | |
| "loss": 0.0547, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.3089430894308944, | |
| "grad_norm": 0.3451598286628723, | |
| "learning_rate": 0.00012081220173794863, | |
| "loss": 0.0411, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.3252032520325203, | |
| "grad_norm": 0.36769524216651917, | |
| "learning_rate": 0.00011975941035670664, | |
| "loss": 0.0428, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.341463414634146, | |
| "grad_norm": 0.38125452399253845, | |
| "learning_rate": 0.00011870433490699166, | |
| "loss": 0.0363, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.3577235772357725, | |
| "grad_norm": 0.44016361236572266, | |
| "learning_rate": 0.00011764709734915218, | |
| "loss": 0.038, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.3739837398373984, | |
| "grad_norm": 0.2804912328720093, | |
| "learning_rate": 0.0001165878198934632, | |
| "loss": 0.0385, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.3902439024390243, | |
| "grad_norm": 0.31978878378868103, | |
| "learning_rate": 0.00011552662498599958, | |
| "loss": 0.0269, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.40650406504065, | |
| "grad_norm": 0.2746686339378357, | |
| "learning_rate": 0.00011446363529448209, | |
| "loss": 0.0365, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.4227642276422765, | |
| "grad_norm": 0.35929563641548157, | |
| "learning_rate": 0.00011339897369409774, | |
| "loss": 0.0399, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.4390243902439024, | |
| "grad_norm": 0.25356271862983704, | |
| "learning_rate": 0.00011233276325329596, | |
| "loss": 0.0381, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4552845528455283, | |
| "grad_norm": 0.27866753935813904, | |
| "learning_rate": 0.00011126512721956289, | |
| "loss": 0.0273, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.4715447154471546, | |
| "grad_norm": 0.3891955018043518, | |
| "learning_rate": 0.00011019618900517455, | |
| "loss": 0.0367, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.4878048780487805, | |
| "grad_norm": 0.26467326283454895, | |
| "learning_rate": 0.0001091260721729312, | |
| "loss": 0.0311, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.5040650406504064, | |
| "grad_norm": 0.21822647750377655, | |
| "learning_rate": 0.0001080549004218742, | |
| "loss": 0.0296, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.5203252032520327, | |
| "grad_norm": 0.3661019206047058, | |
| "learning_rate": 0.00010698279757298715, | |
| "loss": 0.0395, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.5365853658536586, | |
| "grad_norm": 0.15822070837020874, | |
| "learning_rate": 0.00010590988755488292, | |
| "loss": 0.0265, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.5528455284552845, | |
| "grad_norm": 0.25284048914909363, | |
| "learning_rate": 0.00010483629438947826, | |
| "loss": 0.0307, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.569105691056911, | |
| "grad_norm": 0.3222500681877136, | |
| "learning_rate": 0.00010376214217765772, | |
| "loss": 0.0361, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.5853658536585367, | |
| "grad_norm": 0.3294614255428314, | |
| "learning_rate": 0.00010268755508492804, | |
| "loss": 0.0354, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.6016260162601625, | |
| "grad_norm": 0.4046204388141632, | |
| "learning_rate": 0.0001016126573270658, | |
| "loss": 0.0297, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.617886178861789, | |
| "grad_norm": 0.37592411041259766, | |
| "learning_rate": 0.0001005375731557584, | |
| "loss": 0.0326, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.6341463414634148, | |
| "grad_norm": 0.4793863594532013, | |
| "learning_rate": 9.946242684424162e-05, | |
| "loss": 0.0315, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.6504065040650406, | |
| "grad_norm": 0.31578221917152405, | |
| "learning_rate": 9.838734267293421e-05, | |
| "loss": 0.0316, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.4366059899330139, | |
| "learning_rate": 9.731244491507197e-05, | |
| "loss": 0.0286, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.682926829268293, | |
| "grad_norm": 0.3444521725177765, | |
| "learning_rate": 9.623785782234234e-05, | |
| "loss": 0.0278, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.6991869918699187, | |
| "grad_norm": 0.5460310578346252, | |
| "learning_rate": 9.516370561052174e-05, | |
| "loss": 0.0444, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.7154471544715446, | |
| "grad_norm": 0.360115110874176, | |
| "learning_rate": 9.409011244511712e-05, | |
| "loss": 0.0288, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.7317073170731705, | |
| "grad_norm": 0.3582008183002472, | |
| "learning_rate": 9.30172024270129e-05, | |
| "loss": 0.0311, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.747967479674797, | |
| "grad_norm": 0.4152023494243622, | |
| "learning_rate": 9.194509957812582e-05, | |
| "loss": 0.0463, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.7642276422764227, | |
| "grad_norm": 0.2088373750448227, | |
| "learning_rate": 9.087392782706883e-05, | |
| "loss": 0.0331, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7804878048780486, | |
| "grad_norm": 0.5220609307289124, | |
| "learning_rate": 8.980381099482546e-05, | |
| "loss": 0.0311, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.796747967479675, | |
| "grad_norm": 0.49769702553749084, | |
| "learning_rate": 8.873487278043712e-05, | |
| "loss": 0.0336, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.813008130081301, | |
| "grad_norm": 0.3503866195678711, | |
| "learning_rate": 8.766723674670407e-05, | |
| "loss": 0.0272, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.8292682926829267, | |
| "grad_norm": 0.24078375101089478, | |
| "learning_rate": 8.660102630590227e-05, | |
| "loss": 0.0365, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.845528455284553, | |
| "grad_norm": 0.27272090315818787, | |
| "learning_rate": 8.553636470551792e-05, | |
| "loss": 0.0252, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.861788617886179, | |
| "grad_norm": 0.25519323348999023, | |
| "learning_rate": 8.447337501400047e-05, | |
| "loss": 0.0286, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.8780487804878048, | |
| "grad_norm": 0.27832046151161194, | |
| "learning_rate": 8.341218010653684e-05, | |
| "loss": 0.0306, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.894308943089431, | |
| "grad_norm": 0.3626549243927002, | |
| "learning_rate": 8.235290265084785e-05, | |
| "loss": 0.0266, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.910569105691057, | |
| "grad_norm": 0.2994539737701416, | |
| "learning_rate": 8.129566509300835e-05, | |
| "loss": 0.0338, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.926829268292683, | |
| "grad_norm": 0.32438531517982483, | |
| "learning_rate": 8.02405896432934e-05, | |
| "loss": 0.0308, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.943089430894309, | |
| "grad_norm": 0.4146476686000824, | |
| "learning_rate": 7.91877982620514e-05, | |
| "loss": 0.0258, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.959349593495935, | |
| "grad_norm": 0.3459465503692627, | |
| "learning_rate": 7.813741264560622e-05, | |
| "loss": 0.025, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.975609756097561, | |
| "grad_norm": 0.3172384202480316, | |
| "learning_rate": 7.708955421218986e-05, | |
| "loss": 0.0327, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.991869918699187, | |
| "grad_norm": 0.5268030166625977, | |
| "learning_rate": 7.604434408790729e-05, | |
| "loss": 0.033, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.008130081300813, | |
| "grad_norm": 0.2980897128582001, | |
| "learning_rate": 7.50019030927348e-05, | |
| "loss": 0.0307, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.024390243902439, | |
| "grad_norm": 0.3045070767402649, | |
| "learning_rate": 7.396235172655422e-05, | |
| "loss": 0.025, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.040650406504065, | |
| "grad_norm": 0.2718939185142517, | |
| "learning_rate": 7.292581015522363e-05, | |
| "loss": 0.0229, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.0569105691056913, | |
| "grad_norm": 0.5520108938217163, | |
| "learning_rate": 7.189239819668711e-05, | |
| "loss": 0.0305, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.073170731707317, | |
| "grad_norm": 0.2900155186653137, | |
| "learning_rate": 7.086223530712433e-05, | |
| "loss": 0.0301, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.089430894308943, | |
| "grad_norm": 0.31466805934906006, | |
| "learning_rate": 6.983544056714215e-05, | |
| "loss": 0.0343, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.105691056910569, | |
| "grad_norm": 0.4351949989795685, | |
| "learning_rate": 6.881213266800988e-05, | |
| "loss": 0.0287, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.1219512195121952, | |
| "grad_norm": 0.29242777824401855, | |
| "learning_rate": 6.779242989793892e-05, | |
| "loss": 0.0214, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.138211382113821, | |
| "grad_norm": 0.27103492617607117, | |
| "learning_rate": 6.677645012840956e-05, | |
| "loss": 0.0262, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.154471544715447, | |
| "grad_norm": 0.46795374155044556, | |
| "learning_rate": 6.576431080054581e-05, | |
| "loss": 0.0264, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.1707317073170733, | |
| "grad_norm": 0.2594567537307739, | |
| "learning_rate": 6.475612891153969e-05, | |
| "loss": 0.0318, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.186991869918699, | |
| "grad_norm": 0.2888941168785095, | |
| "learning_rate": 6.375202100112729e-05, | |
| "loss": 0.0318, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.203252032520325, | |
| "grad_norm": 0.3916606307029724, | |
| "learning_rate": 6.275210313811739e-05, | |
| "loss": 0.0293, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.2195121951219514, | |
| "grad_norm": 0.22073742747306824, | |
| "learning_rate": 6.175649090697448e-05, | |
| "loss": 0.0321, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.2357723577235773, | |
| "grad_norm": 0.3153645694255829, | |
| "learning_rate": 6.0765299394458185e-05, | |
| "loss": 0.0265, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.252032520325203, | |
| "grad_norm": 0.5098276138305664, | |
| "learning_rate": 5.977864317631965e-05, | |
| "loss": 0.0255, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2682926829268295, | |
| "grad_norm": 0.2877456843852997, | |
| "learning_rate": 5.879663630405736e-05, | |
| "loss": 0.0296, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.2845528455284554, | |
| "grad_norm": 0.1817747801542282, | |
| "learning_rate": 5.781939229173365e-05, | |
| "loss": 0.0225, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.3008130081300813, | |
| "grad_norm": 0.31750866770744324, | |
| "learning_rate": 5.684702410285292e-05, | |
| "loss": 0.0237, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.317073170731707, | |
| "grad_norm": 0.2775154411792755, | |
| "learning_rate": 5.5879644137303797e-05, | |
| "loss": 0.0241, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.30817779898643494, | |
| "learning_rate": 5.491736421836646e-05, | |
| "loss": 0.0217, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.3495934959349594, | |
| "grad_norm": 0.2524275481700897, | |
| "learning_rate": 5.396029557978657e-05, | |
| "loss": 0.0278, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.3658536585365852, | |
| "grad_norm": 0.21641696989536285, | |
| "learning_rate": 5.3008548852917194e-05, | |
| "loss": 0.0266, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.3821138211382116, | |
| "grad_norm": 0.32717615365982056, | |
| "learning_rate": 5.2062234053930504e-05, | |
| "loss": 0.027, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.3983739837398375, | |
| "grad_norm": 0.41103601455688477, | |
| "learning_rate": 5.1121460571100776e-05, | |
| "loss": 0.0186, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.4146341463414633, | |
| "grad_norm": 0.14661918580532074, | |
| "learning_rate": 5.018633715215948e-05, | |
| "loss": 0.0211, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.430894308943089, | |
| "grad_norm": 0.22353650629520416, | |
| "learning_rate": 4.9256971891724845e-05, | |
| "loss": 0.0253, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.4471544715447155, | |
| "grad_norm": 0.3574526906013489, | |
| "learning_rate": 4.833347221880685e-05, | |
| "loss": 0.0308, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.4634146341463414, | |
| "grad_norm": 0.24301108717918396, | |
| "learning_rate": 4.7415944884388904e-05, | |
| "loss": 0.0295, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.4796747967479673, | |
| "grad_norm": 0.28461310267448425, | |
| "learning_rate": 4.6504495949088154e-05, | |
| "loss": 0.024, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.4959349593495936, | |
| "grad_norm": 0.2199145406484604, | |
| "learning_rate": 4.559923077089564e-05, | |
| "loss": 0.0188, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.5121951219512195, | |
| "grad_norm": 0.24626000225543976, | |
| "learning_rate": 4.470025399299728e-05, | |
| "loss": 0.0295, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.5284552845528454, | |
| "grad_norm": 0.1908273845911026, | |
| "learning_rate": 4.3807669531678054e-05, | |
| "loss": 0.0172, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.5447154471544717, | |
| "grad_norm": 0.36190247535705566, | |
| "learning_rate": 4.292158056430966e-05, | |
| "loss": 0.0156, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.5609756097560976, | |
| "grad_norm": 0.33335763216018677, | |
| "learning_rate": 4.204208951742393e-05, | |
| "loss": 0.0187, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.5772357723577235, | |
| "grad_norm": 0.2627616226673126, | |
| "learning_rate": 4.116929805487307e-05, | |
| "loss": 0.022, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.59349593495935, | |
| "grad_norm": 0.33589500188827515, | |
| "learning_rate": 4.0303307066077653e-05, | |
| "loss": 0.0251, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.6097560975609757, | |
| "grad_norm": 0.2956457734107971, | |
| "learning_rate": 3.944421665436462e-05, | |
| "loss": 0.0214, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.6260162601626016, | |
| "grad_norm": 0.1858706772327423, | |
| "learning_rate": 3.859212612539597e-05, | |
| "loss": 0.0269, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.642276422764228, | |
| "grad_norm": 0.21992811560630798, | |
| "learning_rate": 3.7747133975689385e-05, | |
| "loss": 0.0231, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.658536585365854, | |
| "grad_norm": 0.2564171850681305, | |
| "learning_rate": 3.690933788123296e-05, | |
| "loss": 0.0224, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.6747967479674797, | |
| "grad_norm": 0.2475014477968216, | |
| "learning_rate": 3.607883468619414e-05, | |
| "loss": 0.0279, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.6910569105691056, | |
| "grad_norm": 0.3992771506309509, | |
| "learning_rate": 3.525572039172528e-05, | |
| "loss": 0.0294, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.7073170731707314, | |
| "grad_norm": 0.2070523500442505, | |
| "learning_rate": 3.4440090144866536e-05, | |
| "loss": 0.0255, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.7235772357723578, | |
| "grad_norm": 0.3437440097332001, | |
| "learning_rate": 3.363203822754728e-05, | |
| "loss": 0.0238, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.7398373983739837, | |
| "grad_norm": 0.2936142385005951, | |
| "learning_rate": 3.283165804568778e-05, | |
| "loss": 0.0222, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.7560975609756095, | |
| "grad_norm": 0.41532281041145325, | |
| "learning_rate": 3.203904211840213e-05, | |
| "loss": 0.0194, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.772357723577236, | |
| "grad_norm": 0.20870675146579742, | |
| "learning_rate": 3.1254282067303355e-05, | |
| "loss": 0.0226, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.7886178861788617, | |
| "grad_norm": 0.22978569567203522, | |
| "learning_rate": 3.0477468605912784e-05, | |
| "loss": 0.0245, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.8048780487804876, | |
| "grad_norm": 0.15536998212337494, | |
| "learning_rate": 2.970869152917387e-05, | |
| "loss": 0.0183, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.821138211382114, | |
| "grad_norm": 0.2645808458328247, | |
| "learning_rate": 2.894803970307257e-05, | |
| "loss": 0.019, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.83739837398374, | |
| "grad_norm": 0.17497260868549347, | |
| "learning_rate": 2.819560105436504e-05, | |
| "loss": 0.019, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.8536585365853657, | |
| "grad_norm": 0.3548007905483246, | |
| "learning_rate": 2.7451462560413633e-05, | |
| "loss": 0.0233, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.869918699186992, | |
| "grad_norm": 0.1929531693458557, | |
| "learning_rate": 2.6715710239132973e-05, | |
| "loss": 0.0196, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.886178861788618, | |
| "grad_norm": 0.2864411175251007, | |
| "learning_rate": 2.5988429139046865e-05, | |
| "loss": 0.025, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.902439024390244, | |
| "grad_norm": 0.11372525990009308, | |
| "learning_rate": 2.5269703329456996e-05, | |
| "loss": 0.0163, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.91869918699187, | |
| "grad_norm": 0.23025590181350708, | |
| "learning_rate": 2.4559615890725284e-05, | |
| "loss": 0.0199, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.934959349593496, | |
| "grad_norm": 0.14693354070186615, | |
| "learning_rate": 2.3858248904670056e-05, | |
| "loss": 0.0168, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.951219512195122, | |
| "grad_norm": 0.18080760538578033, | |
| "learning_rate": 2.316568344507799e-05, | |
| "loss": 0.0145, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.9674796747967482, | |
| "grad_norm": 0.2424541711807251, | |
| "learning_rate": 2.248199956833258e-05, | |
| "loss": 0.0181, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.983739837398374, | |
| "grad_norm": 0.24186311662197113, | |
| "learning_rate": 2.1807276304159873e-05, | |
| "loss": 0.0235, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5643988847732544, | |
| "learning_rate": 2.1141591646493275e-05, | |
| "loss": 0.0222, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.016260162601626, | |
| "grad_norm": 0.21844126284122467, | |
| "learning_rate": 2.0485022544457933e-05, | |
| "loss": 0.0151, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.032520325203252, | |
| "grad_norm": 0.23713727295398712, | |
| "learning_rate": 1.983764489347574e-05, | |
| "loss": 0.0224, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.048780487804878, | |
| "grad_norm": 0.16830649971961975, | |
| "learning_rate": 1.9199533526492463e-05, | |
| "loss": 0.0243, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.065040650406504, | |
| "grad_norm": 0.31435462832450867, | |
| "learning_rate": 1.857076220532734e-05, | |
| "loss": 0.0229, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.08130081300813, | |
| "grad_norm": 0.4781794548034668, | |
| "learning_rate": 1.7951403612146744e-05, | |
| "loss": 0.0186, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.097560975609756, | |
| "grad_norm": 0.21629083156585693, | |
| "learning_rate": 1.7341529341062623e-05, | |
| "loss": 0.0168, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.1138211382113825, | |
| "grad_norm": 0.14135605096817017, | |
| "learning_rate": 1.6741209889856513e-05, | |
| "loss": 0.0178, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.130081300813008, | |
| "grad_norm": 0.1877690851688385, | |
| "learning_rate": 1.6150514651830483e-05, | |
| "loss": 0.0252, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.146341463414634, | |
| "grad_norm": 0.5082934498786926, | |
| "learning_rate": 1.5569511907785783e-05, | |
| "loss": 0.0164, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.16260162601626, | |
| "grad_norm": 0.20046326518058777, | |
| "learning_rate": 1.4998268818129802e-05, | |
| "loss": 0.0209, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.178861788617886, | |
| "grad_norm": 0.15702232718467712, | |
| "learning_rate": 1.443685141511294e-05, | |
| "loss": 0.0263, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.195121951219512, | |
| "grad_norm": 0.2763630449771881, | |
| "learning_rate": 1.3885324595195493e-05, | |
| "loss": 0.0218, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.211382113821138, | |
| "grad_norm": 0.1242925375699997, | |
| "learning_rate": 1.3343752111546138e-05, | |
| "loss": 0.0139, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.227642276422764, | |
| "grad_norm": 0.23633089661598206, | |
| "learning_rate": 1.2812196566672441e-05, | |
| "loss": 0.0192, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.2439024390243905, | |
| "grad_norm": 0.1781703680753708, | |
| "learning_rate": 1.2290719405184291e-05, | |
| "loss": 0.0182, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.260162601626016, | |
| "grad_norm": 0.20754830539226532, | |
| "learning_rate": 1.1779380906691361e-05, | |
| "loss": 0.0162, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.276422764227642, | |
| "grad_norm": 0.3521351218223572, | |
| "learning_rate": 1.1278240178835197e-05, | |
| "loss": 0.0233, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.2926829268292686, | |
| "grad_norm": 0.2133350968360901, | |
| "learning_rate": 1.0787355150456546e-05, | |
| "loss": 0.019, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.308943089430894, | |
| "grad_norm": 0.22535590827465057, | |
| "learning_rate": 1.0306782564899352e-05, | |
| "loss": 0.0174, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.32520325203252, | |
| "grad_norm": 0.361722856760025, | |
| "learning_rate": 9.83657797345141e-06, | |
| "loss": 0.016, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.341463414634147, | |
| "grad_norm": 0.19668163359165192, | |
| "learning_rate": 9.376795728923016e-06, | |
| "loss": 0.0239, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.357723577235772, | |
| "grad_norm": 0.1902504563331604, | |
| "learning_rate": 8.927488979364184e-06, | |
| "loss": 0.0148, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.373983739837398, | |
| "grad_norm": 0.344314306974411, | |
| "learning_rate": 8.488709661920946e-06, | |
| "loss": 0.0132, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.390243902439025, | |
| "grad_norm": 0.12004625797271729, | |
| "learning_rate": 8.060508496831743e-06, | |
| "loss": 0.0161, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.40650406504065, | |
| "grad_norm": 0.22964505851268768, | |
| "learning_rate": 7.642934981564609e-06, | |
| "loss": 0.014, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.4227642276422765, | |
| "grad_norm": 0.27576199173927307, | |
| "learning_rate": 7.23603738509534e-06, | |
| "loss": 0.0229, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.439024390243903, | |
| "grad_norm": 0.12927354872226715, | |
| "learning_rate": 6.839862742328107e-06, | |
| "loss": 0.0104, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.455284552845528, | |
| "grad_norm": 0.25259020924568176, | |
| "learning_rate": 6.454456848658363e-06, | |
| "loss": 0.0112, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.471544715447155, | |
| "grad_norm": 0.22319383919239044, | |
| "learning_rate": 6.079864254679191e-06, | |
| "loss": 0.0149, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.487804878048781, | |
| "grad_norm": 0.19844867289066315, | |
| "learning_rate": 5.716128261031628e-06, | |
| "loss": 0.0168, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.504065040650406, | |
| "grad_norm": 0.23324230313301086, | |
| "learning_rate": 5.363290913399232e-06, | |
| "loss": 0.014, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.520325203252033, | |
| "grad_norm": 0.22129973769187927, | |
| "learning_rate": 5.021392997647933e-06, | |
| "loss": 0.0155, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.536585365853659, | |
| "grad_norm": 0.4064948260784149, | |
| "learning_rate": 4.690474035111448e-06, | |
| "loss": 0.0202, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.5528455284552845, | |
| "grad_norm": 0.22724312543869019, | |
| "learning_rate": 4.370572278022788e-06, | |
| "loss": 0.0164, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.569105691056911, | |
| "grad_norm": 0.22973546385765076, | |
| "learning_rate": 4.061724705092574e-06, | |
| "loss": 0.0111, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.585365853658536, | |
| "grad_norm": 0.1258576661348343, | |
| "learning_rate": 3.76396701723456e-06, | |
| "loss": 0.0112, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.6016260162601625, | |
| "grad_norm": 0.2755131423473358, | |
| "learning_rate": 3.4773336334387064e-06, | |
| "loss": 0.0223, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.617886178861789, | |
| "grad_norm": 0.21121273934841156, | |
| "learning_rate": 3.2018576867926643e-06, | |
| "loss": 0.0156, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.634146341463414, | |
| "grad_norm": 0.0909205749630928, | |
| "learning_rate": 2.937571020651775e-06, | |
| "loss": 0.0162, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.650406504065041, | |
| "grad_norm": 0.19245898723602295, | |
| "learning_rate": 2.6845041849581033e-06, | |
| "loss": 0.0188, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.24272821843624115, | |
| "learning_rate": 2.442686432709096e-06, | |
| "loss": 0.0191, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.682926829268292, | |
| "grad_norm": 0.1934613585472107, | |
| "learning_rate": 2.212145716576142e-06, | |
| "loss": 0.0158, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.699186991869919, | |
| "grad_norm": 0.19412636756896973, | |
| "learning_rate": 1.9929086856733447e-06, | |
| "loss": 0.0164, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.715447154471545, | |
| "grad_norm": 0.26893916726112366, | |
| "learning_rate": 1.7850006824771004e-06, | |
| "loss": 0.0194, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.7317073170731705, | |
| "grad_norm": 0.20123203098773956, | |
| "learning_rate": 1.5884457398966046e-06, | |
| "loss": 0.0188, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.747967479674797, | |
| "grad_norm": 0.13909615576267242, | |
| "learning_rate": 1.4032665784958877e-06, | |
| "loss": 0.0162, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.764227642276423, | |
| "grad_norm": 0.17143380641937256, | |
| "learning_rate": 1.229484603867348e-06, | |
| "loss": 0.0165, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.780487804878049, | |
| "grad_norm": 0.11365949362516403, | |
| "learning_rate": 1.0671199041574542e-06, | |
| "loss": 0.0221, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.796747967479675, | |
| "grad_norm": 0.22491714358329773, | |
| "learning_rate": 9.161912477446688e-07, | |
| "loss": 0.0252, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.8130081300813, | |
| "grad_norm": 0.3459233343601227, | |
| "learning_rate": 7.76716081069917e-07, | |
| "loss": 0.0167, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.829268292682927, | |
| "grad_norm": 0.23336616158485413, | |
| "learning_rate": 6.487105266199333e-07, | |
| "loss": 0.0156, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.845528455284553, | |
| "grad_norm": 0.13937050104141235, | |
| "learning_rate": 5.321893810635525e-07, | |
| "loss": 0.0148, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.861788617886178, | |
| "grad_norm": 0.34142452478408813, | |
| "learning_rate": 4.2716611354127747e-07, | |
| "loss": 0.0156, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.878048780487805, | |
| "grad_norm": 0.1981387436389923, | |
| "learning_rate": 3.3365286410842465e-07, | |
| "loss": 0.029, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.894308943089431, | |
| "grad_norm": 0.12347165495157242, | |
| "learning_rate": 2.516604423317248e-07, | |
| "loss": 0.016, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.9105691056910565, | |
| "grad_norm": 0.17062854766845703, | |
| "learning_rate": 1.81198326039822e-07, | |
| "loss": 0.0165, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.926829268292683, | |
| "grad_norm": 0.16616424918174744, | |
| "learning_rate": 1.2227466022770584e-07, | |
| "loss": 0.0167, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.943089430894309, | |
| "grad_norm": 0.22849665582180023, | |
| "learning_rate": 7.489625611519779e-08, | |
| "loss": 0.0173, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.959349593495935, | |
| "grad_norm": 0.3534289598464966, | |
| "learning_rate": 3.906859035960331e-08, | |
| "loss": 0.0238, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.975609756097561, | |
| "grad_norm": 0.19245095551013947, | |
| "learning_rate": 1.4795804422651494e-08, | |
| "loss": 0.0134, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.991869918699187, | |
| "grad_norm": 0.15183484554290771, | |
| "learning_rate": 2.080704091766972e-09, | |
| "loss": 0.0118, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 5.001626016260163, | |
| "step": 3076, | |
| "total_flos": 1.078733099791872e+17, | |
| "train_loss": 0.05316672712002526, | |
| "train_runtime": 1399.9912, | |
| "train_samples_per_second": 35.155, | |
| "train_steps_per_second": 2.197 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3076, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "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": 1.078733099791872e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |