{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 3220, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00777302759424796, "grad_norm": 54.0, "learning_rate": 1.8461538461538464e-07, "loss": 2.9617641448974608, "step": 5 }, { "epoch": 0.01554605518849592, "grad_norm": 58.5, "learning_rate": 4.1538461538461544e-07, "loss": 2.940630340576172, "step": 10 }, { "epoch": 0.023319082782743878, "grad_norm": 53.0, "learning_rate": 6.461538461538462e-07, "loss": 2.918345832824707, "step": 15 }, { "epoch": 0.03109211037699184, "grad_norm": 75.0, "learning_rate": 8.76923076923077e-07, "loss": 2.9583749771118164, "step": 20 }, { "epoch": 0.0388651379712398, "grad_norm": 70.5, "learning_rate": 1.1076923076923077e-06, "loss": 3.0301372528076174, "step": 25 }, { "epoch": 0.046638165565487756, "grad_norm": 53.25, "learning_rate": 1.3384615384615385e-06, "loss": 2.978739929199219, "step": 30 }, { "epoch": 0.054411193159735714, "grad_norm": 33.5, "learning_rate": 1.5692307692307693e-06, "loss": 2.9058298110961913, "step": 35 }, { "epoch": 0.06218422075398368, "grad_norm": 42.0, "learning_rate": 1.8e-06, "loss": 2.968424987792969, "step": 40 }, { "epoch": 0.06995724834823164, "grad_norm": 32.5, "learning_rate": 2.030769230769231e-06, "loss": 2.8322284698486326, "step": 45 }, { "epoch": 0.0777302759424796, "grad_norm": 25.25, "learning_rate": 2.2615384615384617e-06, "loss": 2.83587532043457, "step": 50 }, { "epoch": 0.08550330353672755, "grad_norm": 19.375, "learning_rate": 2.4923076923076926e-06, "loss": 2.8118186950683595, "step": 55 }, { "epoch": 0.09327633113097551, "grad_norm": 17.0, "learning_rate": 2.723076923076923e-06, "loss": 2.7470529556274412, "step": 60 }, { "epoch": 0.10104935872522347, "grad_norm": 16.625, "learning_rate": 2.953846153846154e-06, "loss": 2.617660713195801, "step": 65 }, { "epoch": 0.10882238631947143, "grad_norm": 10.25, "learning_rate": 2.999988101788348e-06, "loss": 2.542320442199707, "step": 70 }, { "epoch": 0.11659541391371939, "grad_norm": 8.5, "learning_rate": 2.9999397656270157e-06, "loss": 2.391935348510742, "step": 75 }, { "epoch": 0.12436844150796736, "grad_norm": 7.21875, "learning_rate": 2.9998542490750054e-06, "loss": 2.460367965698242, "step": 80 }, { "epoch": 0.1321414691022153, "grad_norm": 6.40625, "learning_rate": 2.999731554252096e-06, "loss": 2.4480358123779298, "step": 85 }, { "epoch": 0.13991449669646328, "grad_norm": 9.25, "learning_rate": 2.999571684199636e-06, "loss": 2.3973709106445313, "step": 90 }, { "epoch": 0.14768752429071122, "grad_norm": 5.375, "learning_rate": 2.999374642880473e-06, "loss": 2.421527290344238, "step": 95 }, { "epoch": 0.1554605518849592, "grad_norm": 4.875, "learning_rate": 2.9991404351788515e-06, "loss": 2.3498933792114256, "step": 100 }, { "epoch": 0.16323357947920716, "grad_norm": 5.9375, "learning_rate": 2.998869066900293e-06, "loss": 2.2970048904418947, "step": 105 }, { "epoch": 0.1710066070734551, "grad_norm": 5.03125, "learning_rate": 2.9985605447714547e-06, "loss": 2.242618179321289, "step": 110 }, { "epoch": 0.17877963466770308, "grad_norm": 4.46875, "learning_rate": 2.9982148764399588e-06, "loss": 2.275969314575195, "step": 115 }, { "epoch": 0.18655266226195102, "grad_norm": 4.375, "learning_rate": 2.9978320704742046e-06, "loss": 2.223567008972168, "step": 120 }, { "epoch": 0.194325689856199, "grad_norm": 4.5, "learning_rate": 2.997412136363157e-06, "loss": 2.1891298294067383, "step": 125 }, { "epoch": 0.20209871745044694, "grad_norm": 5.84375, "learning_rate": 2.9969550845161107e-06, "loss": 2.2012889862060545, "step": 130 }, { "epoch": 0.2098717450446949, "grad_norm": 3.875, "learning_rate": 2.996460926262431e-06, "loss": 2.16024169921875, "step": 135 }, { "epoch": 0.21764477263894286, "grad_norm": 4.25, "learning_rate": 2.995929673851274e-06, "loss": 2.123866653442383, "step": 140 }, { "epoch": 0.22541780023319083, "grad_norm": 3.953125, "learning_rate": 2.9953613404512844e-06, "loss": 2.1333532333374023, "step": 145 }, { "epoch": 0.23319082782743877, "grad_norm": 4.28125, "learning_rate": 2.9947559401502653e-06, "loss": 2.078639602661133, "step": 150 }, { "epoch": 0.24096385542168675, "grad_norm": 3.765625, "learning_rate": 2.994113487954833e-06, "loss": 2.126356315612793, "step": 155 }, { "epoch": 0.24873688301593472, "grad_norm": 3.5625, "learning_rate": 2.9934339997900427e-06, "loss": 2.087117385864258, "step": 160 }, { "epoch": 0.2565099106101827, "grad_norm": 5.21875, "learning_rate": 2.992717492498995e-06, "loss": 2.0291786193847656, "step": 165 }, { "epoch": 0.2642829382044306, "grad_norm": 5.53125, "learning_rate": 2.991963983842416e-06, "loss": 2.0695690155029296, "step": 170 }, { "epoch": 0.2720559657986786, "grad_norm": 3.390625, "learning_rate": 2.99117349249822e-06, "loss": 2.032662773132324, "step": 175 }, { "epoch": 0.27982899339292655, "grad_norm": 3.65625, "learning_rate": 2.9903460380610464e-06, "loss": 2.060500907897949, "step": 180 }, { "epoch": 0.2876020209871745, "grad_norm": 3.4375, "learning_rate": 2.9894816410417703e-06, "loss": 2.0228250503540037, "step": 185 }, { "epoch": 0.29537504858142244, "grad_norm": 3.46875, "learning_rate": 2.988580322866999e-06, "loss": 1.982876968383789, "step": 190 }, { "epoch": 0.3031480761756704, "grad_norm": 3.234375, "learning_rate": 2.987642105878537e-06, "loss": 1.9798212051391602, "step": 195 }, { "epoch": 0.3109211037699184, "grad_norm": 3.140625, "learning_rate": 2.9866670133328338e-06, "loss": 2.0125585556030274, "step": 200 }, { "epoch": 0.31869413136416636, "grad_norm": 3.5625, "learning_rate": 2.985655069400409e-06, "loss": 1.9702707290649415, "step": 205 }, { "epoch": 0.3264671589584143, "grad_norm": 3.0625, "learning_rate": 2.98460629916525e-06, "loss": 1.8913560867309571, "step": 210 }, { "epoch": 0.33424018655266224, "grad_norm": 3.5625, "learning_rate": 2.983520728624192e-06, "loss": 1.97393741607666, "step": 215 }, { "epoch": 0.3420132141469102, "grad_norm": 5.28125, "learning_rate": 2.9823983846862734e-06, "loss": 1.9876958847045898, "step": 220 }, { "epoch": 0.3497862417411582, "grad_norm": 3.375, "learning_rate": 2.9812392951720687e-06, "loss": 1.9531164169311523, "step": 225 }, { "epoch": 0.35755926933540616, "grad_norm": 3.609375, "learning_rate": 2.980043488812999e-06, "loss": 1.962308120727539, "step": 230 }, { "epoch": 0.3653322969296541, "grad_norm": 3.40625, "learning_rate": 2.9788109952506196e-06, "loss": 1.9883455276489257, "step": 235 }, { "epoch": 0.37310532452390205, "grad_norm": 2.796875, "learning_rate": 2.977541845035886e-06, "loss": 1.9803726196289062, "step": 240 }, { "epoch": 0.38087835211815, "grad_norm": 3.046875, "learning_rate": 2.976236069628395e-06, "loss": 1.9581138610839843, "step": 245 }, { "epoch": 0.388651379712398, "grad_norm": 3.0625, "learning_rate": 2.974893701395606e-06, "loss": 1.9567842483520508, "step": 250 }, { "epoch": 0.39642440730664597, "grad_norm": 2.734375, "learning_rate": 2.9735147736120386e-06, "loss": 1.9849048614501954, "step": 255 }, { "epoch": 0.4041974349008939, "grad_norm": 2.75, "learning_rate": 2.972099320458448e-06, "loss": 1.9369998931884767, "step": 260 }, { "epoch": 0.41197046249514185, "grad_norm": 2.890625, "learning_rate": 2.970647377020978e-06, "loss": 1.9370773315429688, "step": 265 }, { "epoch": 0.4197434900893898, "grad_norm": 5.96875, "learning_rate": 2.969158979290289e-06, "loss": 1.8989524841308594, "step": 270 }, { "epoch": 0.4275165176836378, "grad_norm": 3.046875, "learning_rate": 2.967634164160669e-06, "loss": 1.9389614105224608, "step": 275 }, { "epoch": 0.4352895452778857, "grad_norm": 3.71875, "learning_rate": 2.9660729694291174e-06, "loss": 1.9260137557983399, "step": 280 }, { "epoch": 0.4430625728721337, "grad_norm": 2.890625, "learning_rate": 2.9644754337944084e-06, "loss": 1.8605312347412108, "step": 285 }, { "epoch": 0.45083560046638166, "grad_norm": 3.140625, "learning_rate": 2.962841596856131e-06, "loss": 1.9190540313720703, "step": 290 }, { "epoch": 0.45860862806062963, "grad_norm": 4.0625, "learning_rate": 2.961171499113709e-06, "loss": 1.898582649230957, "step": 295 }, { "epoch": 0.46638165565487755, "grad_norm": 4.25, "learning_rate": 2.9594651819653952e-06, "loss": 1.906097984313965, "step": 300 }, { "epoch": 0.4741546832491255, "grad_norm": 3.703125, "learning_rate": 2.957722687707248e-06, "loss": 1.9028240203857423, "step": 305 }, { "epoch": 0.4819277108433735, "grad_norm": 3.03125, "learning_rate": 2.9559440595320783e-06, "loss": 1.8781393051147461, "step": 310 }, { "epoch": 0.48970073843762146, "grad_norm": 2.78125, "learning_rate": 2.9541293415283853e-06, "loss": 1.9285717010498047, "step": 315 }, { "epoch": 0.49747376603186944, "grad_norm": 4.34375, "learning_rate": 2.9522785786792574e-06, "loss": 1.9032007217407227, "step": 320 }, { "epoch": 0.5052467936261174, "grad_norm": 2.953125, "learning_rate": 2.950391816861261e-06, "loss": 1.9081892013549804, "step": 325 }, { "epoch": 0.5130198212203654, "grad_norm": 2.703125, "learning_rate": 2.948469102843301e-06, "loss": 1.912176513671875, "step": 330 }, { "epoch": 0.5207928488146133, "grad_norm": 3.359375, "learning_rate": 2.946510484285465e-06, "loss": 1.9082002639770508, "step": 335 }, { "epoch": 0.5285658764088612, "grad_norm": 4.125, "learning_rate": 2.9445160097378367e-06, "loss": 1.8770530700683594, "step": 340 }, { "epoch": 0.5363389040031092, "grad_norm": 2.953125, "learning_rate": 2.9424857286392974e-06, "loss": 1.8937904357910156, "step": 345 }, { "epoch": 0.5441119315973572, "grad_norm": 2.40625, "learning_rate": 2.9404196913162967e-06, "loss": 1.8735397338867188, "step": 350 }, { "epoch": 0.5518849591916052, "grad_norm": 3.171875, "learning_rate": 2.9383179489816085e-06, "loss": 1.878750991821289, "step": 355 }, { "epoch": 0.5596579867858531, "grad_norm": 2.5, "learning_rate": 2.9361805537330583e-06, "loss": 1.8543106079101563, "step": 360 }, { "epoch": 0.567431014380101, "grad_norm": 6.0, "learning_rate": 2.934007558552235e-06, "loss": 1.851426124572754, "step": 365 }, { "epoch": 0.575204041974349, "grad_norm": 3.96875, "learning_rate": 2.9317990173031736e-06, "loss": 1.8661741256713866, "step": 370 }, { "epoch": 0.582977069568597, "grad_norm": 4.0625, "learning_rate": 2.9295549847310242e-06, "loss": 1.8610189437866211, "step": 375 }, { "epoch": 0.5907500971628449, "grad_norm": 3.046875, "learning_rate": 2.9272755164606915e-06, "loss": 1.8564264297485351, "step": 380 }, { "epoch": 0.5985231247570929, "grad_norm": 3.0625, "learning_rate": 2.924960668995459e-06, "loss": 1.8446855545043945, "step": 385 }, { "epoch": 0.6062961523513408, "grad_norm": 2.953125, "learning_rate": 2.9226104997155864e-06, "loss": 1.89840087890625, "step": 390 }, { "epoch": 0.6140691799455888, "grad_norm": 2.4375, "learning_rate": 2.920225066876887e-06, "loss": 1.8710636138916015, "step": 395 }, { "epoch": 0.6218422075398368, "grad_norm": 2.703125, "learning_rate": 2.9178044296092857e-06, "loss": 1.8294385910034179, "step": 400 }, { "epoch": 0.6296152351340847, "grad_norm": 3.484375, "learning_rate": 2.9153486479153515e-06, "loss": 1.8434131622314454, "step": 405 }, { "epoch": 0.6373882627283327, "grad_norm": 2.59375, "learning_rate": 2.9128577826688105e-06, "loss": 1.856954574584961, "step": 410 }, { "epoch": 0.6451612903225806, "grad_norm": 2.390625, "learning_rate": 2.9103318956130372e-06, "loss": 1.8874540328979492, "step": 415 }, { "epoch": 0.6529343179168287, "grad_norm": 2.875, "learning_rate": 2.9077710493595255e-06, "loss": 1.8662496566772462, "step": 420 }, { "epoch": 0.6607073455110766, "grad_norm": 2.578125, "learning_rate": 2.9051753073863318e-06, "loss": 1.8711732864379882, "step": 425 }, { "epoch": 0.6684803731053245, "grad_norm": 2.5625, "learning_rate": 2.9025447340365084e-06, "loss": 1.87451229095459, "step": 430 }, { "epoch": 0.6762534006995725, "grad_norm": 3.34375, "learning_rate": 2.8998793945165033e-06, "loss": 1.844712257385254, "step": 435 }, { "epoch": 0.6840264282938204, "grad_norm": 3.09375, "learning_rate": 2.8971793548945462e-06, "loss": 1.8232318878173828, "step": 440 }, { "epoch": 0.6917994558880683, "grad_norm": 3.8125, "learning_rate": 2.89444468209901e-06, "loss": 1.7989261627197266, "step": 445 }, { "epoch": 0.6995724834823164, "grad_norm": 2.6875, "learning_rate": 2.891675443916751e-06, "loss": 1.812358283996582, "step": 450 }, { "epoch": 0.7073455110765643, "grad_norm": 3.40625, "learning_rate": 2.8888717089914324e-06, "loss": 1.8319356918334961, "step": 455 }, { "epoch": 0.7151185386708123, "grad_norm": 4.15625, "learning_rate": 2.886033546821817e-06, "loss": 1.8614593505859376, "step": 460 }, { "epoch": 0.7228915662650602, "grad_norm": 4.4375, "learning_rate": 2.883161027760049e-06, "loss": 1.8424514770507812, "step": 465 }, { "epoch": 0.7306645938593082, "grad_norm": 2.4375, "learning_rate": 2.880254223009909e-06, "loss": 1.8635786056518555, "step": 470 }, { "epoch": 0.7384376214535562, "grad_norm": 3.265625, "learning_rate": 2.877313204625047e-06, "loss": 1.8209510803222657, "step": 475 }, { "epoch": 0.7462106490478041, "grad_norm": 2.5, "learning_rate": 2.874338045507199e-06, "loss": 1.8248987197875977, "step": 480 }, { "epoch": 0.7539836766420521, "grad_norm": 2.875, "learning_rate": 2.871328819404379e-06, "loss": 1.8343856811523438, "step": 485 }, { "epoch": 0.7617567042363, "grad_norm": 6.3125, "learning_rate": 2.86828560090905e-06, "loss": 1.8513946533203125, "step": 490 }, { "epoch": 0.769529731830548, "grad_norm": 6.3125, "learning_rate": 2.8652084654562765e-06, "loss": 1.7723899841308595, "step": 495 }, { "epoch": 0.777302759424796, "grad_norm": 2.65625, "learning_rate": 2.862097489321854e-06, "loss": 1.8408012390136719, "step": 500 }, { "epoch": 0.7850757870190439, "grad_norm": 2.96875, "learning_rate": 2.8589527496204174e-06, "loss": 1.8310670852661133, "step": 505 }, { "epoch": 0.7928488146132919, "grad_norm": 3.421875, "learning_rate": 2.8557743243035303e-06, "loss": 1.8204288482666016, "step": 510 }, { "epoch": 0.8006218422075398, "grad_norm": 3.046875, "learning_rate": 2.8525622921577544e-06, "loss": 1.8522125244140626, "step": 515 }, { "epoch": 0.8083948698017878, "grad_norm": 2.921875, "learning_rate": 2.849316732802692e-06, "loss": 1.8497058868408203, "step": 520 }, { "epoch": 0.8161678973960358, "grad_norm": 4.59375, "learning_rate": 2.8460377266890183e-06, "loss": 1.8126192092895508, "step": 525 }, { "epoch": 0.8239409249902837, "grad_norm": 2.671875, "learning_rate": 2.8427253550964826e-06, "loss": 1.8248754501342774, "step": 530 }, { "epoch": 0.8317139525845316, "grad_norm": 3.5625, "learning_rate": 2.8393797001318944e-06, "loss": 1.806284523010254, "step": 535 }, { "epoch": 0.8394869801787797, "grad_norm": 3.15625, "learning_rate": 2.8360008447270903e-06, "loss": 1.8294240951538085, "step": 540 }, { "epoch": 0.8472600077730276, "grad_norm": 2.578125, "learning_rate": 2.832588872636876e-06, "loss": 1.8396331787109375, "step": 545 }, { "epoch": 0.8550330353672756, "grad_norm": 3.984375, "learning_rate": 2.8291438684369513e-06, "loss": 1.7990753173828125, "step": 550 }, { "epoch": 0.8628060629615235, "grad_norm": 3.328125, "learning_rate": 2.8256659175218136e-06, "loss": 1.82193603515625, "step": 555 }, { "epoch": 0.8705790905557714, "grad_norm": 2.859375, "learning_rate": 2.8221551061026407e-06, "loss": 1.7653759002685547, "step": 560 }, { "epoch": 0.8783521181500195, "grad_norm": 6.90625, "learning_rate": 2.8186115212051537e-06, "loss": 1.7184221267700195, "step": 565 }, { "epoch": 0.8861251457442674, "grad_norm": 3.109375, "learning_rate": 2.8150352506674606e-06, "loss": 1.7966522216796874, "step": 570 }, { "epoch": 0.8938981733385154, "grad_norm": 3.125, "learning_rate": 2.811426383137878e-06, "loss": 1.823615264892578, "step": 575 }, { "epoch": 0.9016712009327633, "grad_norm": 2.765625, "learning_rate": 2.807785008072735e-06, "loss": 1.8239147186279296, "step": 580 }, { "epoch": 0.9094442285270112, "grad_norm": 2.796875, "learning_rate": 2.804111215734155e-06, "loss": 1.8108486175537108, "step": 585 }, { "epoch": 0.9172172561212593, "grad_norm": 2.984375, "learning_rate": 2.8004050971878165e-06, "loss": 1.7909955978393555, "step": 590 }, { "epoch": 0.9249902837155072, "grad_norm": 3.78125, "learning_rate": 2.7966667443007e-06, "loss": 1.7968484878540039, "step": 595 }, { "epoch": 0.9327633113097551, "grad_norm": 3.203125, "learning_rate": 2.792896249738806e-06, "loss": 1.831535530090332, "step": 600 }, { "epoch": 0.9405363389040031, "grad_norm": 2.65625, "learning_rate": 2.789093706964863e-06, "loss": 1.7990205764770508, "step": 605 }, { "epoch": 0.948309366498251, "grad_norm": 2.765625, "learning_rate": 2.7852592102360057e-06, "loss": 1.7934032440185548, "step": 610 }, { "epoch": 0.9560823940924991, "grad_norm": 2.796875, "learning_rate": 2.781392854601442e-06, "loss": 1.8229087829589843, "step": 615 }, { "epoch": 0.963855421686747, "grad_norm": 2.875, "learning_rate": 2.7774947359000955e-06, "loss": 1.7996929168701172, "step": 620 }, { "epoch": 0.9716284492809949, "grad_norm": 2.984375, "learning_rate": 2.773564950758231e-06, "loss": 1.8018980026245117, "step": 625 }, { "epoch": 0.9794014768752429, "grad_norm": 2.625, "learning_rate": 2.769603596587058e-06, "loss": 1.804305076599121, "step": 630 }, { "epoch": 0.9871745044694908, "grad_norm": 2.578125, "learning_rate": 2.7656107715803163e-06, "loss": 1.7911588668823242, "step": 635 }, { "epoch": 0.9949475320637389, "grad_norm": 2.5, "learning_rate": 2.7615865747118437e-06, "loss": 1.8311017990112304, "step": 640 }, { "epoch": 1.0015546055188496, "grad_norm": 2.421875, "learning_rate": 2.757531105733119e-06, "loss": 1.821329689025879, "step": 645 }, { "epoch": 1.0093276331130976, "grad_norm": 2.53125, "learning_rate": 2.7534444651707936e-06, "loss": 1.8120763778686524, "step": 650 }, { "epoch": 1.0171006607073456, "grad_norm": 2.484375, "learning_rate": 2.7493267543241976e-06, "loss": 1.8227361679077148, "step": 655 }, { "epoch": 1.0248736883015934, "grad_norm": 2.390625, "learning_rate": 2.7451780752628277e-06, "loss": 1.789077377319336, "step": 660 }, { "epoch": 1.0326467158958414, "grad_norm": 2.359375, "learning_rate": 2.740998530823819e-06, "loss": 1.7681766510009767, "step": 665 }, { "epoch": 1.0404197434900895, "grad_norm": 2.265625, "learning_rate": 2.7367882246093956e-06, "loss": 1.8091068267822266, "step": 670 }, { "epoch": 1.0481927710843373, "grad_norm": 3.0, "learning_rate": 2.732547260984301e-06, "loss": 1.7840948104858398, "step": 675 }, { "epoch": 1.0559657986785853, "grad_norm": 4.28125, "learning_rate": 2.7282757450732135e-06, "loss": 1.8019865036010743, "step": 680 }, { "epoch": 1.0637388262728333, "grad_norm": 3.140625, "learning_rate": 2.723973782758139e-06, "loss": 1.8016468048095704, "step": 685 }, { "epoch": 1.0715118538670811, "grad_norm": 3.53125, "learning_rate": 2.7196414806757844e-06, "loss": 1.8223690032958983, "step": 690 }, { "epoch": 1.0792848814613292, "grad_norm": 2.875, "learning_rate": 2.71527894621492e-06, "loss": 1.7492977142333985, "step": 695 }, { "epoch": 1.0870579090555772, "grad_norm": 3.859375, "learning_rate": 2.7108862875137103e-06, "loss": 1.7750293731689453, "step": 700 }, { "epoch": 1.094830936649825, "grad_norm": 2.484375, "learning_rate": 2.706463613457039e-06, "loss": 1.7934614181518556, "step": 705 }, { "epoch": 1.102603964244073, "grad_norm": 3.03125, "learning_rate": 2.702011033673809e-06, "loss": 1.8070440292358398, "step": 710 }, { "epoch": 1.110376991838321, "grad_norm": 2.90625, "learning_rate": 2.697528658534222e-06, "loss": 1.790481948852539, "step": 715 }, { "epoch": 1.118150019432569, "grad_norm": 2.6875, "learning_rate": 2.6930165991470454e-06, "loss": 1.7587804794311523, "step": 720 }, { "epoch": 1.125923047026817, "grad_norm": 3.203125, "learning_rate": 2.688474967356858e-06, "loss": 1.755472755432129, "step": 725 }, { "epoch": 1.133696074621065, "grad_norm": 2.484375, "learning_rate": 2.683903875741277e-06, "loss": 1.727252197265625, "step": 730 }, { "epoch": 1.141469102215313, "grad_norm": 2.640625, "learning_rate": 2.679303437608167e-06, "loss": 1.784395980834961, "step": 735 }, { "epoch": 1.1492421298095608, "grad_norm": 3.375, "learning_rate": 2.674673766992832e-06, "loss": 1.788182830810547, "step": 740 }, { "epoch": 1.1570151574038088, "grad_norm": 2.328125, "learning_rate": 2.670014978655188e-06, "loss": 1.7917722702026366, "step": 745 }, { "epoch": 1.1647881849980568, "grad_norm": 3.03125, "learning_rate": 2.66532718807692e-06, "loss": 1.8185247421264648, "step": 750 }, { "epoch": 1.1725612125923046, "grad_norm": 2.96875, "learning_rate": 2.6606105114586168e-06, "loss": 1.794622039794922, "step": 755 }, { "epoch": 1.1803342401865526, "grad_norm": 2.453125, "learning_rate": 2.655865065716893e-06, "loss": 1.8034521102905274, "step": 760 }, { "epoch": 1.1881072677808007, "grad_norm": 2.78125, "learning_rate": 2.651090968481491e-06, "loss": 1.7976842880249024, "step": 765 }, { "epoch": 1.1958802953750487, "grad_norm": 2.546875, "learning_rate": 2.6462883380923616e-06, "loss": 1.8003448486328124, "step": 770 }, { "epoch": 1.2036533229692965, "grad_norm": 2.328125, "learning_rate": 2.641457293596736e-06, "loss": 1.7950130462646485, "step": 775 }, { "epoch": 1.2114263505635445, "grad_norm": 2.671875, "learning_rate": 2.6365979547461694e-06, "loss": 1.7629627227783202, "step": 780 }, { "epoch": 1.2191993781577926, "grad_norm": 3.4375, "learning_rate": 2.6317104419935783e-06, "loss": 1.7977428436279297, "step": 785 }, { "epoch": 1.2269724057520404, "grad_norm": 2.59375, "learning_rate": 2.6267948764902487e-06, "loss": 1.8033893585205079, "step": 790 }, { "epoch": 1.2347454333462884, "grad_norm": 3.8125, "learning_rate": 2.621851380082838e-06, "loss": 1.8023849487304688, "step": 795 }, { "epoch": 1.2425184609405364, "grad_norm": 2.65625, "learning_rate": 2.6168800753103526e-06, "loss": 1.7480819702148438, "step": 800 }, { "epoch": 1.2502914885347842, "grad_norm": 3.421875, "learning_rate": 2.611881085401109e-06, "loss": 1.7807357788085938, "step": 805 }, { "epoch": 1.2580645161290323, "grad_norm": 2.796875, "learning_rate": 2.606854534269682e-06, "loss": 1.7672304153442382, "step": 810 }, { "epoch": 1.2658375437232803, "grad_norm": 2.8125, "learning_rate": 2.6018005465138317e-06, "loss": 1.801127815246582, "step": 815 }, { "epoch": 1.273610571317528, "grad_norm": 4.8125, "learning_rate": 2.5967192474114146e-06, "loss": 1.7710298538208007, "step": 820 }, { "epoch": 1.2813835989117761, "grad_norm": 3.109375, "learning_rate": 2.5916107629172797e-06, "loss": 1.794015121459961, "step": 825 }, { "epoch": 1.2891566265060241, "grad_norm": 2.65625, "learning_rate": 2.586475219660144e-06, "loss": 1.8152124404907226, "step": 830 }, { "epoch": 1.2969296541002722, "grad_norm": 2.515625, "learning_rate": 2.5813127449394568e-06, "loss": 1.7653358459472657, "step": 835 }, { "epoch": 1.30470268169452, "grad_norm": 5.0, "learning_rate": 2.57612346672224e-06, "loss": 1.729343795776367, "step": 840 }, { "epoch": 1.312475709288768, "grad_norm": 3.15625, "learning_rate": 2.570907513639922e-06, "loss": 1.7817445755004884, "step": 845 }, { "epoch": 1.3202487368830158, "grad_norm": 3.953125, "learning_rate": 2.565665014985141e-06, "loss": 1.7719129562377929, "step": 850 }, { "epoch": 1.3280217644772638, "grad_norm": 2.625, "learning_rate": 2.5603961007085493e-06, "loss": 1.7770267486572267, "step": 855 }, { "epoch": 1.3357947920715119, "grad_norm": 2.65625, "learning_rate": 2.555100901415586e-06, "loss": 1.787220573425293, "step": 860 }, { "epoch": 1.3435678196657599, "grad_norm": 3.515625, "learning_rate": 2.5497795483632387e-06, "loss": 1.7110469818115235, "step": 865 }, { "epoch": 1.3513408472600077, "grad_norm": 2.875, "learning_rate": 2.5444321734567963e-06, "loss": 1.7569700241088868, "step": 870 }, { "epoch": 1.3591138748542557, "grad_norm": 3.3125, "learning_rate": 2.5390589092465733e-06, "loss": 1.7665533065795898, "step": 875 }, { "epoch": 1.3668869024485037, "grad_norm": 3.21875, "learning_rate": 2.5336598889246264e-06, "loss": 1.8013175964355468, "step": 880 }, { "epoch": 1.3746599300427516, "grad_norm": 3.03125, "learning_rate": 2.5282352463214528e-06, "loss": 1.7704401016235352, "step": 885 }, { "epoch": 1.3824329576369996, "grad_norm": 2.59375, "learning_rate": 2.522785115902673e-06, "loss": 1.7569721221923829, "step": 890 }, { "epoch": 1.3902059852312476, "grad_norm": 3.625, "learning_rate": 2.5173096327656977e-06, "loss": 1.796013641357422, "step": 895 }, { "epoch": 1.3979790128254956, "grad_norm": 3.5, "learning_rate": 2.5118089326363784e-06, "loss": 1.7931339263916015, "step": 900 }, { "epoch": 1.4057520404197434, "grad_norm": 4.65625, "learning_rate": 2.506283151865644e-06, "loss": 1.7504447937011718, "step": 905 }, { "epoch": 1.4135250680139915, "grad_norm": 2.84375, "learning_rate": 2.5007324274261194e-06, "loss": 1.7627891540527343, "step": 910 }, { "epoch": 1.4212980956082393, "grad_norm": 3.6875, "learning_rate": 2.4951568969087327e-06, "loss": 1.7439870834350586, "step": 915 }, { "epoch": 1.4290711232024873, "grad_norm": 2.859375, "learning_rate": 2.489556698519301e-06, "loss": 1.780399703979492, "step": 920 }, { "epoch": 1.4368441507967353, "grad_norm": 4.28125, "learning_rate": 2.4839319710751095e-06, "loss": 1.7574060440063477, "step": 925 }, { "epoch": 1.4446171783909834, "grad_norm": 2.421875, "learning_rate": 2.478282854001466e-06, "loss": 1.8056983947753906, "step": 930 }, { "epoch": 1.4523902059852314, "grad_norm": 3.71875, "learning_rate": 2.472609487328246e-06, "loss": 1.7682207107543946, "step": 935 }, { "epoch": 1.4601632335794792, "grad_norm": 3.859375, "learning_rate": 2.466912011686423e-06, "loss": 1.711320686340332, "step": 940 }, { "epoch": 1.4679362611737272, "grad_norm": 2.5625, "learning_rate": 2.461190568304582e-06, "loss": 1.776519775390625, "step": 945 }, { "epoch": 1.475709288767975, "grad_norm": 2.28125, "learning_rate": 2.4554452990054184e-06, "loss": 1.7973337173461914, "step": 950 }, { "epoch": 1.483482316362223, "grad_norm": 2.65625, "learning_rate": 2.4496763462022223e-06, "loss": 1.7991920471191407, "step": 955 }, { "epoch": 1.491255343956471, "grad_norm": 2.671875, "learning_rate": 2.4438838528953484e-06, "loss": 1.7844413757324218, "step": 960 }, { "epoch": 1.499028371550719, "grad_norm": 2.53125, "learning_rate": 2.438067962668672e-06, "loss": 1.778104591369629, "step": 965 }, { "epoch": 1.506801399144967, "grad_norm": 2.625, "learning_rate": 2.43222881968603e-06, "loss": 1.746272087097168, "step": 970 }, { "epoch": 1.514574426739215, "grad_norm": 2.578125, "learning_rate": 2.426366568687645e-06, "loss": 1.779219627380371, "step": 975 }, { "epoch": 1.5223474543334627, "grad_norm": 2.40625, "learning_rate": 2.4204813549865413e-06, "loss": 1.8096736907958983, "step": 980 }, { "epoch": 1.5301204819277108, "grad_norm": 2.65625, "learning_rate": 2.41457332446494e-06, "loss": 1.797464942932129, "step": 985 }, { "epoch": 1.5378935095219588, "grad_norm": 3.015625, "learning_rate": 2.408642623570644e-06, "loss": 1.7678077697753907, "step": 990 }, { "epoch": 1.5456665371162068, "grad_norm": 2.953125, "learning_rate": 2.402689399313407e-06, "loss": 1.8119983673095703, "step": 995 }, { "epoch": 1.5534395647104549, "grad_norm": 2.34375, "learning_rate": 2.396713799261291e-06, "loss": 1.7814701080322266, "step": 1000 }, { "epoch": 1.5612125923047027, "grad_norm": 3.515625, "learning_rate": 2.390715971537006e-06, "loss": 1.752749252319336, "step": 1005 }, { "epoch": 1.5689856198989507, "grad_norm": 2.546875, "learning_rate": 2.384696064814242e-06, "loss": 1.795747947692871, "step": 1010 }, { "epoch": 1.5767586474931985, "grad_norm": 4.1875, "learning_rate": 2.3786542283139805e-06, "loss": 1.7686239242553712, "step": 1015 }, { "epoch": 1.5845316750874465, "grad_norm": 2.515625, "learning_rate": 2.372590611800796e-06, "loss": 1.7375780105590821, "step": 1020 }, { "epoch": 1.5923047026816946, "grad_norm": 2.796875, "learning_rate": 2.3665053655791444e-06, "loss": 1.789957618713379, "step": 1025 }, { "epoch": 1.6000777302759426, "grad_norm": 3.25, "learning_rate": 2.3603986404896376e-06, "loss": 1.7653923034667969, "step": 1030 }, { "epoch": 1.6078507578701906, "grad_norm": 2.984375, "learning_rate": 2.3542705879053043e-06, "loss": 1.729112434387207, "step": 1035 }, { "epoch": 1.6156237854644384, "grad_norm": 3.953125, "learning_rate": 2.348121359727837e-06, "loss": 1.7447639465332032, "step": 1040 }, { "epoch": 1.6233968130586862, "grad_norm": 2.609375, "learning_rate": 2.3419511083838278e-06, "loss": 1.7655605316162108, "step": 1045 }, { "epoch": 1.6311698406529342, "grad_norm": 2.796875, "learning_rate": 2.335759986820987e-06, "loss": 1.7414113998413085, "step": 1050 }, { "epoch": 1.6389428682471823, "grad_norm": 3.296875, "learning_rate": 2.329548148504358e-06, "loss": 1.7887720108032226, "step": 1055 }, { "epoch": 1.6467158958414303, "grad_norm": 2.90625, "learning_rate": 2.3233157474125075e-06, "loss": 1.7684999465942384, "step": 1060 }, { "epoch": 1.6544889234356783, "grad_norm": 2.1875, "learning_rate": 2.317062938033709e-06, "loss": 1.7593379974365235, "step": 1065 }, { "epoch": 1.6622619510299261, "grad_norm": 2.859375, "learning_rate": 2.3107898753621195e-06, "loss": 1.7033002853393555, "step": 1070 }, { "epoch": 1.6700349786241742, "grad_norm": 2.359375, "learning_rate": 2.3044967148939284e-06, "loss": 1.7695625305175782, "step": 1075 }, { "epoch": 1.677808006218422, "grad_norm": 3.046875, "learning_rate": 2.2981836126235104e-06, "loss": 1.774606704711914, "step": 1080 }, { "epoch": 1.68558103381267, "grad_norm": 3.0, "learning_rate": 2.291850725039555e-06, "loss": 1.772963523864746, "step": 1085 }, { "epoch": 1.693354061406918, "grad_norm": 2.703125, "learning_rate": 2.2854982091211894e-06, "loss": 1.734603500366211, "step": 1090 }, { "epoch": 1.701127089001166, "grad_norm": 2.6875, "learning_rate": 2.279126222334086e-06, "loss": 1.7291036605834962, "step": 1095 }, { "epoch": 1.708900116595414, "grad_norm": 3.390625, "learning_rate": 2.2727349226265593e-06, "loss": 1.7703891754150392, "step": 1100 }, { "epoch": 1.7166731441896619, "grad_norm": 2.75, "learning_rate": 2.2663244684256506e-06, "loss": 1.7763927459716797, "step": 1105 }, { "epoch": 1.7244461717839097, "grad_norm": 3.21875, "learning_rate": 2.2598950186332025e-06, "loss": 1.7746389389038086, "step": 1110 }, { "epoch": 1.7322191993781577, "grad_norm": 2.421875, "learning_rate": 2.2534467326219183e-06, "loss": 1.7058826446533204, "step": 1115 }, { "epoch": 1.7399922269724057, "grad_norm": 3.140625, "learning_rate": 2.2469797702314114e-06, "loss": 1.7668472290039063, "step": 1120 }, { "epoch": 1.7477652545666538, "grad_norm": 2.4375, "learning_rate": 2.2404942917642458e-06, "loss": 1.7704736709594726, "step": 1125 }, { "epoch": 1.7555382821609018, "grad_norm": 5.46875, "learning_rate": 2.233990457981957e-06, "loss": 1.7767803192138671, "step": 1130 }, { "epoch": 1.7633113097551496, "grad_norm": 3.03125, "learning_rate": 2.2274684301010746e-06, "loss": 1.7955919265747071, "step": 1135 }, { "epoch": 1.7710843373493976, "grad_norm": 2.59375, "learning_rate": 2.2209283697891193e-06, "loss": 1.7463964462280273, "step": 1140 }, { "epoch": 1.7788573649436454, "grad_norm": 3.0, "learning_rate": 2.2143704391605995e-06, "loss": 1.794013786315918, "step": 1145 }, { "epoch": 1.7866303925378935, "grad_norm": 2.515625, "learning_rate": 2.2077948007729917e-06, "loss": 1.7816959381103517, "step": 1150 }, { "epoch": 1.7944034201321415, "grad_norm": 2.46875, "learning_rate": 2.2012016176227107e-06, "loss": 1.7438255310058595, "step": 1155 }, { "epoch": 1.8021764477263895, "grad_norm": 3.5, "learning_rate": 2.1945910531410693e-06, "loss": 1.781534767150879, "step": 1160 }, { "epoch": 1.8099494753206375, "grad_norm": 3.59375, "learning_rate": 2.187963271190228e-06, "loss": 1.7872081756591798, "step": 1165 }, { "epoch": 1.8177225029148854, "grad_norm": 2.421875, "learning_rate": 2.1813184360591315e-06, "loss": 1.7323631286621093, "step": 1170 }, { "epoch": 1.8254955305091332, "grad_norm": 2.5, "learning_rate": 2.1746567124594385e-06, "loss": 1.775542640686035, "step": 1175 }, { "epoch": 1.8332685581033812, "grad_norm": 3.015625, "learning_rate": 2.167978265521438e-06, "loss": 1.7475069046020508, "step": 1180 }, { "epoch": 1.8410415856976292, "grad_norm": 2.890625, "learning_rate": 2.161283260789954e-06, "loss": 1.7649881362915039, "step": 1185 }, { "epoch": 1.8488146132918772, "grad_norm": 3.875, "learning_rate": 2.154571864220246e-06, "loss": 1.765119743347168, "step": 1190 }, { "epoch": 1.8565876408861253, "grad_norm": 2.921875, "learning_rate": 2.147844242173891e-06, "loss": 1.7782596588134765, "step": 1195 }, { "epoch": 1.864360668480373, "grad_norm": 3.078125, "learning_rate": 2.1411005614146637e-06, "loss": 1.7638042449951172, "step": 1200 }, { "epoch": 1.872133696074621, "grad_norm": 3.59375, "learning_rate": 2.1343409891044e-06, "loss": 1.7743385314941407, "step": 1205 }, { "epoch": 1.879906723668869, "grad_norm": 2.5, "learning_rate": 2.127565692798855e-06, "loss": 1.7695886611938476, "step": 1210 }, { "epoch": 1.887679751263117, "grad_norm": 5.375, "learning_rate": 2.120774840443548e-06, "loss": 1.7191287994384765, "step": 1215 }, { "epoch": 1.895452778857365, "grad_norm": 2.578125, "learning_rate": 2.1139686003696003e-06, "loss": 1.7808595657348634, "step": 1220 }, { "epoch": 1.903225806451613, "grad_norm": 2.546875, "learning_rate": 2.1071471412895645e-06, "loss": 1.754263687133789, "step": 1225 }, { "epoch": 1.910998834045861, "grad_norm": 2.53125, "learning_rate": 2.10031063229324e-06, "loss": 1.7451448440551758, "step": 1230 }, { "epoch": 1.9187718616401088, "grad_norm": 2.84375, "learning_rate": 2.0934592428434806e-06, "loss": 1.757664108276367, "step": 1235 }, { "epoch": 1.9265448892343566, "grad_norm": 3.578125, "learning_rate": 2.086593142771998e-06, "loss": 1.7408138275146485, "step": 1240 }, { "epoch": 1.9343179168286047, "grad_norm": 2.796875, "learning_rate": 2.079712502275148e-06, "loss": 1.7485319137573243, "step": 1245 }, { "epoch": 1.9420909444228527, "grad_norm": 2.65625, "learning_rate": 2.072817491909715e-06, "loss": 1.7801921844482422, "step": 1250 }, { "epoch": 1.9498639720171007, "grad_norm": 2.390625, "learning_rate": 2.065908282588681e-06, "loss": 1.774850082397461, "step": 1255 }, { "epoch": 1.9576369996113487, "grad_norm": 4.96875, "learning_rate": 2.058985045576992e-06, "loss": 1.7353193283081054, "step": 1260 }, { "epoch": 1.9654100272055965, "grad_norm": 2.84375, "learning_rate": 2.052047952487309e-06, "loss": 1.770503616333008, "step": 1265 }, { "epoch": 1.9731830547998446, "grad_norm": 2.875, "learning_rate": 2.045097175275759e-06, "loss": 1.770104217529297, "step": 1270 }, { "epoch": 1.9809560823940924, "grad_norm": 2.40625, "learning_rate": 2.038132886237669e-06, "loss": 1.7520822525024413, "step": 1275 }, { "epoch": 1.9887291099883404, "grad_norm": 3.890625, "learning_rate": 2.0311552580032965e-06, "loss": 1.7534919738769532, "step": 1280 }, { "epoch": 1.9965021375825884, "grad_norm": 2.375, "learning_rate": 2.024164463533548e-06, "loss": 1.7719182968139648, "step": 1285 }, { "epoch": 2.003109211037699, "grad_norm": 3.234375, "learning_rate": 2.0171606761156962e-06, "loss": 1.644389533996582, "step": 1290 }, { "epoch": 2.010882238631947, "grad_norm": 2.734375, "learning_rate": 2.0101440693590803e-06, "loss": 1.7732845306396485, "step": 1295 }, { "epoch": 2.018655266226195, "grad_norm": 2.8125, "learning_rate": 2.003114817190805e-06, "loss": 1.7711856842041016, "step": 1300 }, { "epoch": 2.026428293820443, "grad_norm": 3.625, "learning_rate": 1.996073093851428e-06, "loss": 1.7617452621459961, "step": 1305 }, { "epoch": 2.0342013214146912, "grad_norm": 3.953125, "learning_rate": 1.9890190738906433e-06, "loss": 1.7819772720336915, "step": 1310 }, { "epoch": 2.041974349008939, "grad_norm": 3.21875, "learning_rate": 1.9819529321629494e-06, "loss": 1.7369638442993165, "step": 1315 }, { "epoch": 2.049747376603187, "grad_norm": 2.625, "learning_rate": 1.97487484382332e-06, "loss": 1.7876413345336915, "step": 1320 }, { "epoch": 2.057520404197435, "grad_norm": 2.421875, "learning_rate": 1.967784984322861e-06, "loss": 1.7732719421386718, "step": 1325 }, { "epoch": 2.065293431791683, "grad_norm": 3.109375, "learning_rate": 1.9606835294044603e-06, "loss": 1.7772335052490233, "step": 1330 }, { "epoch": 2.073066459385931, "grad_norm": 2.328125, "learning_rate": 1.953570655098432e-06, "loss": 1.7934331893920898, "step": 1335 }, { "epoch": 2.080839486980179, "grad_norm": 2.34375, "learning_rate": 1.946446537718153e-06, "loss": 1.777911376953125, "step": 1340 }, { "epoch": 2.0886125145744265, "grad_norm": 2.4375, "learning_rate": 1.9393113538556924e-06, "loss": 1.7777915954589845, "step": 1345 }, { "epoch": 2.0963855421686746, "grad_norm": 2.796875, "learning_rate": 1.932165280377436e-06, "loss": 1.788250732421875, "step": 1350 }, { "epoch": 2.1041585697629226, "grad_norm": 2.859375, "learning_rate": 1.9250084944196988e-06, "loss": 1.7718017578125, "step": 1355 }, { "epoch": 2.1119315973571706, "grad_norm": 2.703125, "learning_rate": 1.9178411733843364e-06, "loss": 1.7869396209716797, "step": 1360 }, { "epoch": 2.1197046249514186, "grad_norm": 2.40625, "learning_rate": 1.9106634949343473e-06, "loss": 1.8019922256469727, "step": 1365 }, { "epoch": 2.1274776525456667, "grad_norm": 2.296875, "learning_rate": 1.9034756369894692e-06, "loss": 1.7722969055175781, "step": 1370 }, { "epoch": 2.1352506801399147, "grad_norm": 3.5, "learning_rate": 1.8962777777217669e-06, "loss": 1.7499950408935547, "step": 1375 }, { "epoch": 2.1430237077341623, "grad_norm": 3.28125, "learning_rate": 1.8890700955512189e-06, "loss": 1.7738359451293946, "step": 1380 }, { "epoch": 2.1507967353284103, "grad_norm": 3.171875, "learning_rate": 1.8818527691412925e-06, "loss": 1.7729887008666991, "step": 1385 }, { "epoch": 2.1585697629226583, "grad_norm": 4.21875, "learning_rate": 1.8746259773945148e-06, "loss": 1.7873857498168946, "step": 1390 }, { "epoch": 2.1663427905169064, "grad_norm": 2.6875, "learning_rate": 1.8673898994480398e-06, "loss": 1.765378189086914, "step": 1395 }, { "epoch": 2.1741158181111544, "grad_norm": 2.34375, "learning_rate": 1.8601447146692066e-06, "loss": 1.7550296783447266, "step": 1400 }, { "epoch": 2.1818888457054024, "grad_norm": 2.8125, "learning_rate": 1.8528906026510948e-06, "loss": 1.7230934143066405, "step": 1405 }, { "epoch": 2.18966187329965, "grad_norm": 2.71875, "learning_rate": 1.8456277432080706e-06, "loss": 1.7671232223510742, "step": 1410 }, { "epoch": 2.197434900893898, "grad_norm": 2.265625, "learning_rate": 1.8383563163713308e-06, "loss": 1.7675060272216796, "step": 1415 }, { "epoch": 2.205207928488146, "grad_norm": 2.734375, "learning_rate": 1.831076502384441e-06, "loss": 1.7740877151489258, "step": 1420 }, { "epoch": 2.212980956082394, "grad_norm": 2.5625, "learning_rate": 1.8237884816988655e-06, "loss": 1.7807220458984374, "step": 1425 }, { "epoch": 2.220753983676642, "grad_norm": 2.765625, "learning_rate": 1.8164924349694965e-06, "loss": 1.7493669509887695, "step": 1430 }, { "epoch": 2.22852701127089, "grad_norm": 3.4375, "learning_rate": 1.8091885430501755e-06, "loss": 1.7585426330566407, "step": 1435 }, { "epoch": 2.236300038865138, "grad_norm": 2.71875, "learning_rate": 1.8018769869892085e-06, "loss": 1.7532352447509765, "step": 1440 }, { "epoch": 2.2440730664593858, "grad_norm": 3.03125, "learning_rate": 1.7945579480248812e-06, "loss": 1.721951675415039, "step": 1445 }, { "epoch": 2.251846094053634, "grad_norm": 2.546875, "learning_rate": 1.7872316075809636e-06, "loss": 1.7741312026977538, "step": 1450 }, { "epoch": 2.259619121647882, "grad_norm": 2.1875, "learning_rate": 1.7798981472622156e-06, "loss": 1.7756109237670898, "step": 1455 }, { "epoch": 2.26739214924213, "grad_norm": 2.734375, "learning_rate": 1.772557748849883e-06, "loss": 1.7617694854736328, "step": 1460 }, { "epoch": 2.275165176836378, "grad_norm": 3.09375, "learning_rate": 1.765210594297193e-06, "loss": 1.7698179244995118, "step": 1465 }, { "epoch": 2.282938204430626, "grad_norm": 2.90625, "learning_rate": 1.7578568657248436e-06, "loss": 1.7777706146240235, "step": 1470 }, { "epoch": 2.290711232024874, "grad_norm": 2.46875, "learning_rate": 1.7504967454164882e-06, "loss": 1.7665998458862304, "step": 1475 }, { "epoch": 2.2984842596191215, "grad_norm": 2.765625, "learning_rate": 1.7431304158142182e-06, "loss": 1.715095329284668, "step": 1480 }, { "epoch": 2.3062572872133695, "grad_norm": 2.34375, "learning_rate": 1.7357580595140425e-06, "loss": 1.787546157836914, "step": 1485 }, { "epoch": 2.3140303148076176, "grad_norm": 2.71875, "learning_rate": 1.7283798592613564e-06, "loss": 1.7780780792236328, "step": 1490 }, { "epoch": 2.3218033424018656, "grad_norm": 2.46875, "learning_rate": 1.720995997946416e-06, "loss": 1.7788579940795899, "step": 1495 }, { "epoch": 2.3295763699961136, "grad_norm": 3.0625, "learning_rate": 1.7136066585998048e-06, "loss": 1.7214733123779298, "step": 1500 }, { "epoch": 2.337349397590361, "grad_norm": 3.34375, "learning_rate": 1.7062120243878925e-06, "loss": 1.7493986129760741, "step": 1505 }, { "epoch": 2.3451224251846092, "grad_norm": 2.84375, "learning_rate": 1.6988122786082993e-06, "loss": 1.7621702194213866, "step": 1510 }, { "epoch": 2.3528954527788573, "grad_norm": 3.609375, "learning_rate": 1.6914076046853503e-06, "loss": 1.7600994110107422, "step": 1515 }, { "epoch": 2.3606684803731053, "grad_norm": 2.40625, "learning_rate": 1.6839981861655287e-06, "loss": 1.7418390274047852, "step": 1520 }, { "epoch": 2.3684415079673533, "grad_norm": 2.734375, "learning_rate": 1.6765842067129274e-06, "loss": 1.776521873474121, "step": 1525 }, { "epoch": 2.3762145355616013, "grad_norm": 2.34375, "learning_rate": 1.6691658501046935e-06, "loss": 1.7667064666748047, "step": 1530 }, { "epoch": 2.3839875631558494, "grad_norm": 2.546875, "learning_rate": 1.661743300226476e-06, "loss": 1.7320545196533204, "step": 1535 }, { "epoch": 2.3917605907500974, "grad_norm": 2.859375, "learning_rate": 1.6543167410678667e-06, "loss": 1.784928321838379, "step": 1540 }, { "epoch": 2.399533618344345, "grad_norm": 2.796875, "learning_rate": 1.6468863567178379e-06, "loss": 1.7708812713623048, "step": 1545 }, { "epoch": 2.407306645938593, "grad_norm": 2.4375, "learning_rate": 1.6394523313601812e-06, "loss": 1.7657438278198243, "step": 1550 }, { "epoch": 2.415079673532841, "grad_norm": 3.046875, "learning_rate": 1.6320148492689412e-06, "loss": 1.712628173828125, "step": 1555 }, { "epoch": 2.422852701127089, "grad_norm": 2.890625, "learning_rate": 1.6245740948038475e-06, "loss": 1.7688016891479492, "step": 1560 }, { "epoch": 2.430625728721337, "grad_norm": 3.21875, "learning_rate": 1.6171302524057453e-06, "loss": 1.7326457977294922, "step": 1565 }, { "epoch": 2.438398756315585, "grad_norm": 4.0, "learning_rate": 1.609683506592023e-06, "loss": 1.7959550857543944, "step": 1570 }, { "epoch": 2.4461717839098327, "grad_norm": 2.890625, "learning_rate": 1.602234041952039e-06, "loss": 1.771004867553711, "step": 1575 }, { "epoch": 2.4539448115040807, "grad_norm": 2.78125, "learning_rate": 1.5947820431425452e-06, "loss": 1.7740625381469726, "step": 1580 }, { "epoch": 2.4617178390983288, "grad_norm": 2.671875, "learning_rate": 1.587327694883111e-06, "loss": 1.7554828643798828, "step": 1585 }, { "epoch": 2.469490866692577, "grad_norm": 2.796875, "learning_rate": 1.5798711819515433e-06, "loss": 1.7785354614257813, "step": 1590 }, { "epoch": 2.477263894286825, "grad_norm": 3.265625, "learning_rate": 1.5724126891793064e-06, "loss": 1.75738525390625, "step": 1595 }, { "epoch": 2.485036921881073, "grad_norm": 2.859375, "learning_rate": 1.5649524014469419e-06, "loss": 1.7681261062622071, "step": 1600 }, { "epoch": 2.492809949475321, "grad_norm": 2.859375, "learning_rate": 1.5574905036794833e-06, "loss": 1.7711904525756836, "step": 1605 }, { "epoch": 2.5005829770695684, "grad_norm": 2.71875, "learning_rate": 1.5500271808418748e-06, "loss": 1.7633676528930664, "step": 1610 }, { "epoch": 2.5083560046638165, "grad_norm": 2.90625, "learning_rate": 1.542562617934384e-06, "loss": 1.7807418823242187, "step": 1615 }, { "epoch": 2.5161290322580645, "grad_norm": 2.40625, "learning_rate": 1.535096999988018e-06, "loss": 1.7224298477172852, "step": 1620 }, { "epoch": 2.5239020598523125, "grad_norm": 2.703125, "learning_rate": 1.5276305120599364e-06, "loss": 1.7611215591430665, "step": 1625 }, { "epoch": 2.5316750874465606, "grad_norm": 2.4375, "learning_rate": 1.520163339228863e-06, "loss": 1.7787666320800781, "step": 1630 }, { "epoch": 2.539448115040808, "grad_norm": 2.890625, "learning_rate": 1.5126956665904987e-06, "loss": 1.7301008224487304, "step": 1635 }, { "epoch": 2.547221142635056, "grad_norm": 3.0625, "learning_rate": 1.5052276792529354e-06, "loss": 1.7142169952392579, "step": 1640 }, { "epoch": 2.554994170229304, "grad_norm": 3.140625, "learning_rate": 1.4977595623320637e-06, "loss": 1.7511466979980468, "step": 1645 }, { "epoch": 2.5627671978235522, "grad_norm": 2.703125, "learning_rate": 1.4902915009469875e-06, "loss": 1.7742849349975587, "step": 1650 }, { "epoch": 2.5705402254178003, "grad_norm": 2.8125, "learning_rate": 1.482823680215433e-06, "loss": 1.7731920242309571, "step": 1655 }, { "epoch": 2.5783132530120483, "grad_norm": 2.765625, "learning_rate": 1.4753562852491628e-06, "loss": 1.7701074600219726, "step": 1660 }, { "epoch": 2.5860862806062963, "grad_norm": 3.0, "learning_rate": 1.4678895011493835e-06, "loss": 1.7559581756591798, "step": 1665 }, { "epoch": 2.5938593082005443, "grad_norm": 3.171875, "learning_rate": 1.4604235130021621e-06, "loss": 1.7824094772338868, "step": 1670 }, { "epoch": 2.601632335794792, "grad_norm": 2.734375, "learning_rate": 1.4529585058738333e-06, "loss": 1.7687917709350587, "step": 1675 }, { "epoch": 2.60940536338904, "grad_norm": 3.203125, "learning_rate": 1.4454946648064153e-06, "loss": 1.7860563278198243, "step": 1680 }, { "epoch": 2.617178390983288, "grad_norm": 2.671875, "learning_rate": 1.4380321748130227e-06, "loss": 1.74415283203125, "step": 1685 }, { "epoch": 2.624951418577536, "grad_norm": 2.734375, "learning_rate": 1.4305712208732788e-06, "loss": 1.759819793701172, "step": 1690 }, { "epoch": 2.632724446171784, "grad_norm": 3.453125, "learning_rate": 1.423111987928732e-06, "loss": 1.7638904571533203, "step": 1695 }, { "epoch": 2.6404974737660316, "grad_norm": 3.078125, "learning_rate": 1.41565466087827e-06, "loss": 1.7653776168823243, "step": 1700 }, { "epoch": 2.6482705013602796, "grad_norm": 4.125, "learning_rate": 1.4081994245735376e-06, "loss": 1.7490694046020507, "step": 1705 }, { "epoch": 2.6560435289545277, "grad_norm": 6.9375, "learning_rate": 1.4007464638143558e-06, "loss": 1.7174070358276368, "step": 1710 }, { "epoch": 2.6638165565487757, "grad_norm": 2.1875, "learning_rate": 1.393295963344137e-06, "loss": 1.7607751846313477, "step": 1715 }, { "epoch": 2.6715895841430237, "grad_norm": 2.90625, "learning_rate": 1.3858481078453108e-06, "loss": 1.7637535095214845, "step": 1720 }, { "epoch": 2.6793626117372717, "grad_norm": 5.34375, "learning_rate": 1.3784030819347412e-06, "loss": 1.7166439056396485, "step": 1725 }, { "epoch": 2.6871356393315198, "grad_norm": 2.453125, "learning_rate": 1.3709610701591534e-06, "loss": 1.7579578399658202, "step": 1730 }, { "epoch": 2.694908666925768, "grad_norm": 3.828125, "learning_rate": 1.36352225699056e-06, "loss": 1.7800247192382812, "step": 1735 }, { "epoch": 2.7026816945200154, "grad_norm": 2.578125, "learning_rate": 1.3560868268216835e-06, "loss": 1.7647558212280274, "step": 1740 }, { "epoch": 2.7104547221142634, "grad_norm": 2.828125, "learning_rate": 1.3486549639613922e-06, "loss": 1.7599218368530274, "step": 1745 }, { "epoch": 2.7182277497085114, "grad_norm": 2.6875, "learning_rate": 1.3412268526301264e-06, "loss": 1.7933555603027345, "step": 1750 }, { "epoch": 2.7260007773027595, "grad_norm": 4.75, "learning_rate": 1.3338026769553333e-06, "loss": 1.7205957412719726, "step": 1755 }, { "epoch": 2.7337738048970075, "grad_norm": 3.203125, "learning_rate": 1.3263826209669057e-06, "loss": 1.7575620651245116, "step": 1760 }, { "epoch": 2.741546832491255, "grad_norm": 2.65625, "learning_rate": 1.3189668685926148e-06, "loss": 1.7774202346801757, "step": 1765 }, { "epoch": 2.749319860085503, "grad_norm": 2.828125, "learning_rate": 1.3115556036535573e-06, "loss": 1.7546514511108398, "step": 1770 }, { "epoch": 2.757092887679751, "grad_norm": 4.46875, "learning_rate": 1.3041490098595935e-06, "loss": 1.7550155639648437, "step": 1775 }, { "epoch": 2.764865915273999, "grad_norm": 2.6875, "learning_rate": 1.296747270804796e-06, "loss": 1.7426984786987305, "step": 1780 }, { "epoch": 2.772638942868247, "grad_norm": 2.5625, "learning_rate": 1.2893505699629001e-06, "loss": 1.7772262573242188, "step": 1785 }, { "epoch": 2.780411970462495, "grad_norm": 3.125, "learning_rate": 1.281959090682752e-06, "loss": 1.7860633850097656, "step": 1790 }, { "epoch": 2.7881849980567432, "grad_norm": 4.0625, "learning_rate": 1.274573016183768e-06, "loss": 1.7018733978271485, "step": 1795 }, { "epoch": 2.7959580256509913, "grad_norm": 2.296875, "learning_rate": 1.2671925295513897e-06, "loss": 1.7680021286010743, "step": 1800 }, { "epoch": 2.803731053245239, "grad_norm": 2.828125, "learning_rate": 1.259817813732548e-06, "loss": 1.7612527847290038, "step": 1805 }, { "epoch": 2.811504080839487, "grad_norm": 2.6875, "learning_rate": 1.252449051531128e-06, "loss": 1.7553150177001953, "step": 1810 }, { "epoch": 2.819277108433735, "grad_norm": 3.265625, "learning_rate": 1.2450864256034349e-06, "loss": 1.7797388076782226, "step": 1815 }, { "epoch": 2.827050136027983, "grad_norm": 2.765625, "learning_rate": 1.2377301184536706e-06, "loss": 1.7565105438232422, "step": 1820 }, { "epoch": 2.834823163622231, "grad_norm": 3.0, "learning_rate": 1.230380312429406e-06, "loss": 1.7580532073974608, "step": 1825 }, { "epoch": 2.8425961912164786, "grad_norm": 2.9375, "learning_rate": 1.223037189717064e-06, "loss": 1.753823471069336, "step": 1830 }, { "epoch": 2.8503692188107266, "grad_norm": 3.171875, "learning_rate": 1.2157009323374012e-06, "loss": 1.764747428894043, "step": 1835 }, { "epoch": 2.8581422464049746, "grad_norm": 2.953125, "learning_rate": 1.2083717221409963e-06, "loss": 1.7954195022583008, "step": 1840 }, { "epoch": 2.8659152739992226, "grad_norm": 2.984375, "learning_rate": 1.201049740803744e-06, "loss": 1.7779060363769532, "step": 1845 }, { "epoch": 2.8736883015934707, "grad_norm": 2.578125, "learning_rate": 1.193735169822351e-06, "loss": 1.7677398681640626, "step": 1850 }, { "epoch": 2.8814613291877187, "grad_norm": 2.796875, "learning_rate": 1.1864281905098354e-06, "loss": 1.7579959869384765, "step": 1855 }, { "epoch": 2.8892343567819667, "grad_norm": 2.34375, "learning_rate": 1.1791289839910343e-06, "loss": 1.7535253524780274, "step": 1860 }, { "epoch": 2.8970073843762147, "grad_norm": 2.375, "learning_rate": 1.1718377311981124e-06, "loss": 1.726751708984375, "step": 1865 }, { "epoch": 2.9047804119704628, "grad_norm": 3.1875, "learning_rate": 1.1645546128660805e-06, "loss": 1.7149169921875, "step": 1870 }, { "epoch": 2.9125534395647104, "grad_norm": 3.53125, "learning_rate": 1.1572798095283102e-06, "loss": 1.7750076293945312, "step": 1875 }, { "epoch": 2.9203264671589584, "grad_norm": 2.34375, "learning_rate": 1.1500135015120636e-06, "loss": 1.7679439544677735, "step": 1880 }, { "epoch": 2.9280994947532064, "grad_norm": 2.453125, "learning_rate": 1.1427558689340206e-06, "loss": 1.7440618515014648, "step": 1885 }, { "epoch": 2.9358725223474544, "grad_norm": 2.3125, "learning_rate": 1.1355070916958149e-06, "loss": 1.770624542236328, "step": 1890 }, { "epoch": 2.943645549941702, "grad_norm": 2.59375, "learning_rate": 1.1282673494795747e-06, "loss": 1.7704587936401368, "step": 1895 }, { "epoch": 2.95141857753595, "grad_norm": 3.875, "learning_rate": 1.1210368217434694e-06, "loss": 1.7376117706298828, "step": 1900 }, { "epoch": 2.959191605130198, "grad_norm": 2.34375, "learning_rate": 1.1138156877172608e-06, "loss": 1.7749898910522461, "step": 1905 }, { "epoch": 2.966964632724446, "grad_norm": 2.828125, "learning_rate": 1.1066041263978588e-06, "loss": 1.7454212188720704, "step": 1910 }, { "epoch": 2.974737660318694, "grad_norm": 5.3125, "learning_rate": 1.099402316544886e-06, "loss": 1.7421390533447265, "step": 1915 }, { "epoch": 2.982510687912942, "grad_norm": 3.984375, "learning_rate": 1.0922104366762485e-06, "loss": 1.7417736053466797, "step": 1920 }, { "epoch": 2.99028371550719, "grad_norm": 2.46875, "learning_rate": 1.0850286650637059e-06, "loss": 1.7803096771240234, "step": 1925 }, { "epoch": 2.998056743101438, "grad_norm": 2.65625, "learning_rate": 1.0778571797284569e-06, "loss": 1.706426239013672, "step": 1930 }, { "epoch": 3.004663816556549, "grad_norm": 2.453125, "learning_rate": 1.0706961584367247e-06, "loss": 1.7661842346191405, "step": 1935 }, { "epoch": 3.012436844150797, "grad_norm": 2.640625, "learning_rate": 1.063545778695349e-06, "loss": 1.7753393173217773, "step": 1940 }, { "epoch": 3.0202098717450445, "grad_norm": 3.296875, "learning_rate": 1.0564062177473899e-06, "loss": 1.7629749298095703, "step": 1945 }, { "epoch": 3.0279828993392925, "grad_norm": 2.34375, "learning_rate": 1.04927765256773e-06, "loss": 1.7747407913208009, "step": 1950 }, { "epoch": 3.0357559269335406, "grad_norm": 2.296875, "learning_rate": 1.0421602598586914e-06, "loss": 1.7691226959228517, "step": 1955 }, { "epoch": 3.0435289545277886, "grad_norm": 2.703125, "learning_rate": 1.035054216045653e-06, "loss": 1.7747447967529297, "step": 1960 }, { "epoch": 3.0513019821220366, "grad_norm": 2.765625, "learning_rate": 1.0279596972726766e-06, "loss": 1.761591339111328, "step": 1965 }, { "epoch": 3.0590750097162847, "grad_norm": 2.28125, "learning_rate": 1.0208768793981459e-06, "loss": 1.756491470336914, "step": 1970 }, { "epoch": 3.0668480373105322, "grad_norm": 2.328125, "learning_rate": 1.0138059379904e-06, "loss": 1.75128116607666, "step": 1975 }, { "epoch": 3.0746210649047803, "grad_norm": 2.578125, "learning_rate": 1.0067470483233877e-06, "loss": 1.7842300415039063, "step": 1980 }, { "epoch": 3.0823940924990283, "grad_norm": 2.296875, "learning_rate": 9.997003853723182e-07, "loss": 1.779800033569336, "step": 1985 }, { "epoch": 3.0901671200932763, "grad_norm": 2.78125, "learning_rate": 9.926661238093271e-07, "loss": 1.7459890365600585, "step": 1990 }, { "epoch": 3.0979401476875243, "grad_norm": 2.421875, "learning_rate": 9.856444379991451e-07, "loss": 1.7735584259033204, "step": 1995 }, { "epoch": 3.1057131752817724, "grad_norm": 3.515625, "learning_rate": 9.786355019947756e-07, "loss": 1.7522598266601563, "step": 2000 }, { "epoch": 3.1134862028760204, "grad_norm": 3.234375, "learning_rate": 9.716394895331811e-07, "loss": 1.7496126174926758, "step": 2005 }, { "epoch": 3.121259230470268, "grad_norm": 2.46875, "learning_rate": 9.64656574030976e-07, "loss": 1.7627361297607422, "step": 2010 }, { "epoch": 3.129032258064516, "grad_norm": 3.484375, "learning_rate": 9.576869285801299e-07, "loss": 1.7688138961791993, "step": 2015 }, { "epoch": 3.136805285658764, "grad_norm": 2.703125, "learning_rate": 9.507307259436733e-07, "loss": 1.7860504150390626, "step": 2020 }, { "epoch": 3.144578313253012, "grad_norm": 2.71875, "learning_rate": 9.437881385514176e-07, "loss": 1.7684850692749023, "step": 2025 }, { "epoch": 3.15235134084726, "grad_norm": 2.78125, "learning_rate": 9.368593384956822e-07, "loss": 1.7632024765014649, "step": 2030 }, { "epoch": 3.160124368441508, "grad_norm": 2.34375, "learning_rate": 9.299444975270249e-07, "loss": 1.7403291702270507, "step": 2035 }, { "epoch": 3.167897396035756, "grad_norm": 2.546875, "learning_rate": 9.230437870499888e-07, "loss": 1.758439826965332, "step": 2040 }, { "epoch": 3.1756704236300037, "grad_norm": 2.515625, "learning_rate": 9.161573781188502e-07, "loss": 1.7483726501464845, "step": 2045 }, { "epoch": 3.1834434512242518, "grad_norm": 3.15625, "learning_rate": 9.092854414333794e-07, "loss": 1.764262580871582, "step": 2050 }, { "epoch": 3.1912164788185, "grad_norm": 2.359375, "learning_rate": 9.02428147334611e-07, "loss": 1.7782501220703124, "step": 2055 }, { "epoch": 3.198989506412748, "grad_norm": 3.640625, "learning_rate": 8.955856658006192e-07, "loss": 1.724038314819336, "step": 2060 }, { "epoch": 3.206762534006996, "grad_norm": 2.296875, "learning_rate": 8.88758166442307e-07, "loss": 1.7864198684692383, "step": 2065 }, { "epoch": 3.214535561601244, "grad_norm": 3.03125, "learning_rate": 8.819458184991987e-07, "loss": 1.7696599960327148, "step": 2070 }, { "epoch": 3.2223085891954915, "grad_norm": 2.75, "learning_rate": 8.751487908352473e-07, "loss": 1.7572134017944336, "step": 2075 }, { "epoch": 3.2300816167897395, "grad_norm": 2.515625, "learning_rate": 8.683672519346478e-07, "loss": 1.754510498046875, "step": 2080 }, { "epoch": 3.2378546443839875, "grad_norm": 2.609375, "learning_rate": 8.616013698976607e-07, "loss": 1.8111570358276368, "step": 2085 }, { "epoch": 3.2456276719782355, "grad_norm": 3.328125, "learning_rate": 8.548513124364465e-07, "loss": 1.7244117736816407, "step": 2090 }, { "epoch": 3.2534006995724836, "grad_norm": 2.828125, "learning_rate": 8.48117246870906e-07, "loss": 1.7631061553955079, "step": 2095 }, { "epoch": 3.2611737271667316, "grad_norm": 3.40625, "learning_rate": 8.413993401245337e-07, "loss": 1.774520492553711, "step": 2100 }, { "epoch": 3.2689467547609796, "grad_norm": 2.5625, "learning_rate": 8.346977587202832e-07, "loss": 1.7586915969848633, "step": 2105 }, { "epoch": 3.276719782355227, "grad_norm": 2.453125, "learning_rate": 8.280126687764333e-07, "loss": 1.7768667221069336, "step": 2110 }, { "epoch": 3.2844928099494752, "grad_norm": 3.1875, "learning_rate": 8.213442360024769e-07, "loss": 1.7587329864501953, "step": 2115 }, { "epoch": 3.2922658375437233, "grad_norm": 2.921875, "learning_rate": 8.146926256950084e-07, "loss": 1.777347946166992, "step": 2120 }, { "epoch": 3.3000388651379713, "grad_norm": 2.9375, "learning_rate": 8.080580027336279e-07, "loss": 1.766446304321289, "step": 2125 }, { "epoch": 3.3078118927322193, "grad_norm": 2.453125, "learning_rate": 8.014405315768565e-07, "loss": 1.7943761825561524, "step": 2130 }, { "epoch": 3.3155849203264673, "grad_norm": 3.203125, "learning_rate": 7.948403762580551e-07, "loss": 1.7770727157592774, "step": 2135 }, { "epoch": 3.323357947920715, "grad_norm": 4.40625, "learning_rate": 7.882577003813635e-07, "loss": 1.7405311584472656, "step": 2140 }, { "epoch": 3.331130975514963, "grad_norm": 2.921875, "learning_rate": 7.816926671176409e-07, "loss": 1.7624347686767579, "step": 2145 }, { "epoch": 3.338904003109211, "grad_norm": 2.5625, "learning_rate": 7.751454392004223e-07, "loss": 1.7643087387084961, "step": 2150 }, { "epoch": 3.346677030703459, "grad_norm": 2.421875, "learning_rate": 7.686161789218868e-07, "loss": 1.7300819396972655, "step": 2155 }, { "epoch": 3.354450058297707, "grad_norm": 2.71875, "learning_rate": 7.621050481288318e-07, "loss": 1.763786506652832, "step": 2160 }, { "epoch": 3.362223085891955, "grad_norm": 5.75, "learning_rate": 7.556122082186634e-07, "loss": 1.7216815948486328, "step": 2165 }, { "epoch": 3.369996113486203, "grad_norm": 2.765625, "learning_rate": 7.491378201353939e-07, "loss": 1.753557014465332, "step": 2170 }, { "epoch": 3.3777691410804507, "grad_norm": 3.09375, "learning_rate": 7.426820443656533e-07, "loss": 1.762174415588379, "step": 2175 }, { "epoch": 3.3855421686746987, "grad_norm": 2.4375, "learning_rate": 7.36245040934711e-07, "loss": 1.7352489471435546, "step": 2180 }, { "epoch": 3.3933151962689467, "grad_norm": 2.609375, "learning_rate": 7.298269694025087e-07, "loss": 1.7582456588745117, "step": 2185 }, { "epoch": 3.4010882238631948, "grad_norm": 2.234375, "learning_rate": 7.234279888597073e-07, "loss": 1.7559436798095702, "step": 2190 }, { "epoch": 3.408861251457443, "grad_norm": 2.515625, "learning_rate": 7.170482579237399e-07, "loss": 1.7775749206542968, "step": 2195 }, { "epoch": 3.416634279051691, "grad_norm": 2.875, "learning_rate": 7.10687934734883e-07, "loss": 1.7569839477539062, "step": 2200 }, { "epoch": 3.4244073066459384, "grad_norm": 3.890625, "learning_rate": 7.043471769523351e-07, "loss": 1.6683605194091797, "step": 2205 }, { "epoch": 3.4321803342401864, "grad_norm": 2.5625, "learning_rate": 6.980261417503076e-07, "loss": 1.6849105834960938, "step": 2210 }, { "epoch": 3.4399533618344345, "grad_norm": 2.515625, "learning_rate": 6.917249858141326e-07, "loss": 1.7555824279785157, "step": 2215 }, { "epoch": 3.4477263894286825, "grad_norm": 3.1875, "learning_rate": 6.854438653363741e-07, "loss": 1.7801778793334961, "step": 2220 }, { "epoch": 3.4554994170229305, "grad_norm": 2.671875, "learning_rate": 6.791829360129608e-07, "loss": 1.7139925003051757, "step": 2225 }, { "epoch": 3.4632724446171785, "grad_norm": 2.75, "learning_rate": 6.729423530393231e-07, "loss": 1.781975555419922, "step": 2230 }, { "epoch": 3.4710454722114266, "grad_norm": 3.328125, "learning_rate": 6.667222711065476e-07, "loss": 1.7374834060668944, "step": 2235 }, { "epoch": 3.478818499805674, "grad_norm": 4.0625, "learning_rate": 6.605228443975444e-07, "loss": 1.7408964157104492, "step": 2240 }, { "epoch": 3.486591527399922, "grad_norm": 2.578125, "learning_rate": 6.543442265832211e-07, "loss": 1.777541732788086, "step": 2245 }, { "epoch": 3.49436455499417, "grad_norm": 2.671875, "learning_rate": 6.481865708186784e-07, "loss": 1.7833541870117187, "step": 2250 }, { "epoch": 3.5021375825884182, "grad_norm": 3.0, "learning_rate": 6.420500297394098e-07, "loss": 1.7197647094726562, "step": 2255 }, { "epoch": 3.5099106101826663, "grad_norm": 2.5625, "learning_rate": 6.359347554575194e-07, "loss": 1.795170783996582, "step": 2260 }, { "epoch": 3.517683637776914, "grad_norm": 3.3125, "learning_rate": 6.298408995579535e-07, "loss": 1.7494256973266602, "step": 2265 }, { "epoch": 3.525456665371162, "grad_norm": 2.734375, "learning_rate": 6.237686130947391e-07, "loss": 1.753997802734375, "step": 2270 }, { "epoch": 3.53322969296541, "grad_norm": 3.25, "learning_rate": 6.177180465872422e-07, "loss": 1.741074562072754, "step": 2275 }, { "epoch": 3.541002720559658, "grad_norm": 2.765625, "learning_rate": 6.116893500164374e-07, "loss": 1.7701303482055664, "step": 2280 }, { "epoch": 3.548775748153906, "grad_norm": 2.765625, "learning_rate": 6.056826728211877e-07, "loss": 1.7552713394165038, "step": 2285 }, { "epoch": 3.556548775748154, "grad_norm": 6.375, "learning_rate": 5.996981638945417e-07, "loss": 1.741415023803711, "step": 2290 }, { "epoch": 3.564321803342402, "grad_norm": 3.21875, "learning_rate": 5.937359715800429e-07, "loss": 1.7522384643554687, "step": 2295 }, { "epoch": 3.57209483093665, "grad_norm": 2.625, "learning_rate": 5.877962436680529e-07, "loss": 1.7885507583618163, "step": 2300 }, { "epoch": 3.579867858530898, "grad_norm": 2.9375, "learning_rate": 5.818791273920869e-07, "loss": 1.7706375122070312, "step": 2305 }, { "epoch": 3.5876408861251456, "grad_norm": 2.546875, "learning_rate": 5.759847694251635e-07, "loss": 1.7742149353027343, "step": 2310 }, { "epoch": 3.5954139137193937, "grad_norm": 2.421875, "learning_rate": 5.701133158761728e-07, "loss": 1.7870452880859375, "step": 2315 }, { "epoch": 3.6031869413136417, "grad_norm": 2.546875, "learning_rate": 5.64264912286249e-07, "loss": 1.778057861328125, "step": 2320 }, { "epoch": 3.6109599689078897, "grad_norm": 3.671875, "learning_rate": 5.584397036251682e-07, "loss": 1.731595802307129, "step": 2325 }, { "epoch": 3.6187329965021373, "grad_norm": 4.03125, "learning_rate": 5.526378342877504e-07, "loss": 1.7263044357299804, "step": 2330 }, { "epoch": 3.6265060240963853, "grad_norm": 3.359375, "learning_rate": 5.468594480902838e-07, "loss": 1.748951530456543, "step": 2335 }, { "epoch": 3.6342790516906334, "grad_norm": 3.328125, "learning_rate": 5.411046882669577e-07, "loss": 1.801791000366211, "step": 2340 }, { "epoch": 3.6420520792848814, "grad_norm": 2.484375, "learning_rate": 5.353736974663117e-07, "loss": 1.7621234893798827, "step": 2345 }, { "epoch": 3.6498251068791294, "grad_norm": 2.84375, "learning_rate": 5.29666617747703e-07, "loss": 1.7628608703613282, "step": 2350 }, { "epoch": 3.6575981344733774, "grad_norm": 3.515625, "learning_rate": 5.239835905777802e-07, "loss": 1.7559473037719726, "step": 2355 }, { "epoch": 3.6653711620676255, "grad_norm": 2.9375, "learning_rate": 5.183247568269814e-07, "loss": 1.7627080917358398, "step": 2360 }, { "epoch": 3.6731441896618735, "grad_norm": 2.921875, "learning_rate": 5.126902567660389e-07, "loss": 1.767656135559082, "step": 2365 }, { "epoch": 3.6809172172561215, "grad_norm": 3.015625, "learning_rate": 5.070802300625025e-07, "loss": 1.7655416488647462, "step": 2370 }, { "epoch": 3.688690244850369, "grad_norm": 3.4375, "learning_rate": 5.01494815777281e-07, "loss": 1.7869661331176758, "step": 2375 }, { "epoch": 3.696463272444617, "grad_norm": 3.4375, "learning_rate": 4.9593415236119e-07, "loss": 1.7594249725341797, "step": 2380 }, { "epoch": 3.704236300038865, "grad_norm": 2.53125, "learning_rate": 4.903983776515234e-07, "loss": 1.7484663009643555, "step": 2385 }, { "epoch": 3.712009327633113, "grad_norm": 2.859375, "learning_rate": 4.848876288686359e-07, "loss": 1.755828285217285, "step": 2390 }, { "epoch": 3.719782355227361, "grad_norm": 3.046875, "learning_rate": 4.794020426125406e-07, "loss": 1.7893360137939454, "step": 2395 }, { "epoch": 3.727555382821609, "grad_norm": 3.65625, "learning_rate": 4.739417548595257e-07, "loss": 1.7429431915283202, "step": 2400 }, { "epoch": 3.735328410415857, "grad_norm": 2.921875, "learning_rate": 4.685069009587798e-07, "loss": 1.7324928283691405, "step": 2405 }, { "epoch": 3.743101438010105, "grad_norm": 3.4375, "learning_rate": 4.630976156290413e-07, "loss": 1.7435697555541991, "step": 2410 }, { "epoch": 3.750874465604353, "grad_norm": 2.765625, "learning_rate": 4.577140329552552e-07, "loss": 1.795145606994629, "step": 2415 }, { "epoch": 3.758647493198601, "grad_norm": 3.1875, "learning_rate": 4.5235628638525116e-07, "loss": 1.7927955627441405, "step": 2420 }, { "epoch": 3.766420520792849, "grad_norm": 3.203125, "learning_rate": 4.4702450872643693e-07, "loss": 1.721831703186035, "step": 2425 }, { "epoch": 3.774193548387097, "grad_norm": 2.5, "learning_rate": 4.4171883214250293e-07, "loss": 1.75341739654541, "step": 2430 }, { "epoch": 3.781966575981345, "grad_norm": 2.21875, "learning_rate": 4.3643938815015037e-07, "loss": 1.7757869720458985, "step": 2435 }, { "epoch": 3.7897396035755926, "grad_norm": 2.859375, "learning_rate": 4.311863076158272e-07, "loss": 1.7683055877685547, "step": 2440 }, { "epoch": 3.7975126311698406, "grad_norm": 2.71875, "learning_rate": 4.25959720752486e-07, "loss": 1.8101987838745117, "step": 2445 }, { "epoch": 3.8052856587640886, "grad_norm": 2.171875, "learning_rate": 4.2075975711635817e-07, "loss": 1.7586313247680665, "step": 2450 }, { "epoch": 3.8130586863583367, "grad_norm": 3.3125, "learning_rate": 4.155865456037383e-07, "loss": 1.7471582412719726, "step": 2455 }, { "epoch": 3.8208317139525847, "grad_norm": 2.59375, "learning_rate": 4.1044021444779323e-07, "loss": 1.7756813049316407, "step": 2460 }, { "epoch": 3.8286047415468323, "grad_norm": 3.125, "learning_rate": 4.053208912153806e-07, "loss": 1.7852455139160157, "step": 2465 }, { "epoch": 3.8363777691410803, "grad_norm": 2.71875, "learning_rate": 4.0022870280388756e-07, "loss": 1.7516096115112305, "step": 2470 }, { "epoch": 3.8441507967353283, "grad_norm": 4.1875, "learning_rate": 3.951637754380867e-07, "loss": 1.7447877883911134, "step": 2475 }, { "epoch": 3.8519238243295764, "grad_norm": 3.890625, "learning_rate": 3.9012623466700447e-07, "loss": 1.7418466567993165, "step": 2480 }, { "epoch": 3.8596968519238244, "grad_norm": 2.65625, "learning_rate": 3.8511620536081147e-07, "loss": 1.7564357757568358, "step": 2485 }, { "epoch": 3.8674698795180724, "grad_norm": 2.28125, "learning_rate": 3.801338117077256e-07, "loss": 1.7164739608764648, "step": 2490 }, { "epoch": 3.8752429071123204, "grad_norm": 2.875, "learning_rate": 3.751791772109343e-07, "loss": 1.7621259689331055, "step": 2495 }, { "epoch": 3.8830159347065685, "grad_norm": 2.59375, "learning_rate": 3.702524246855344e-07, "loss": 1.779111671447754, "step": 2500 }, { "epoch": 3.890788962300816, "grad_norm": 2.71875, "learning_rate": 3.6535367625548496e-07, "loss": 1.776869773864746, "step": 2505 }, { "epoch": 3.898561989895064, "grad_norm": 2.546875, "learning_rate": 3.604830533505832e-07, "loss": 1.7571853637695312, "step": 2510 }, { "epoch": 3.906335017489312, "grad_norm": 2.390625, "learning_rate": 3.556406767034513e-07, "loss": 1.753143882751465, "step": 2515 }, { "epoch": 3.91410804508356, "grad_norm": 2.71875, "learning_rate": 3.5082666634654744e-07, "loss": 1.7345037460327148, "step": 2520 }, { "epoch": 3.921881072677808, "grad_norm": 2.859375, "learning_rate": 3.4604114160918663e-07, "loss": 1.7644924163818358, "step": 2525 }, { "epoch": 3.9296541002720558, "grad_norm": 2.265625, "learning_rate": 3.4128422111458514e-07, "loss": 1.7600223541259765, "step": 2530 }, { "epoch": 3.9374271278663038, "grad_norm": 2.84375, "learning_rate": 3.365560227769199e-07, "loss": 1.7658714294433593, "step": 2535 }, { "epoch": 3.945200155460552, "grad_norm": 3.359375, "learning_rate": 3.318566637984047e-07, "loss": 1.7390407562255858, "step": 2540 }, { "epoch": 3.9529731830548, "grad_norm": 2.578125, "learning_rate": 3.2718626066638596e-07, "loss": 1.7679237365722655, "step": 2545 }, { "epoch": 3.960746210649048, "grad_norm": 2.34375, "learning_rate": 3.225449291504545e-07, "loss": 1.778669548034668, "step": 2550 }, { "epoch": 3.968519238243296, "grad_norm": 3.484375, "learning_rate": 3.179327842995759e-07, "loss": 1.7730709075927735, "step": 2555 }, { "epoch": 3.976292265837544, "grad_norm": 2.84375, "learning_rate": 3.1334994043924e-07, "loss": 1.7566652297973633, "step": 2560 }, { "epoch": 3.984065293431792, "grad_norm": 3.046875, "learning_rate": 3.087965111686245e-07, "loss": 1.7090982437133788, "step": 2565 }, { "epoch": 3.9918383210260395, "grad_norm": 4.09375, "learning_rate": 3.0427260935778227e-07, "loss": 1.77779541015625, "step": 2570 }, { "epoch": 3.9996113486202876, "grad_norm": 3.171875, "learning_rate": 2.997783471448401e-07, "loss": 1.751749038696289, "step": 2575 }, { "epoch": 4.006218422075398, "grad_norm": 2.34375, "learning_rate": 2.9531383593322173e-07, "loss": 1.7432666778564454, "step": 2580 }, { "epoch": 4.013991449669646, "grad_norm": 2.328125, "learning_rate": 2.9087918638888514e-07, "loss": 1.7630302429199218, "step": 2585 }, { "epoch": 4.021764477263894, "grad_norm": 2.5625, "learning_rate": 2.86474508437579e-07, "loss": 1.780599594116211, "step": 2590 }, { "epoch": 4.029537504858142, "grad_norm": 2.375, "learning_rate": 2.8209991126211986e-07, "loss": 1.7590412139892577, "step": 2595 }, { "epoch": 4.03731053245239, "grad_norm": 3.25, "learning_rate": 2.777555032996832e-07, "loss": 1.7594148635864257, "step": 2600 }, { "epoch": 4.045083560046638, "grad_norm": 2.8125, "learning_rate": 2.7344139223911646e-07, "loss": 1.7629459381103516, "step": 2605 }, { "epoch": 4.052856587640886, "grad_norm": 2.375, "learning_rate": 2.6915768501827127e-07, "loss": 1.7560413360595704, "step": 2610 }, { "epoch": 4.060629615235134, "grad_norm": 3.328125, "learning_rate": 2.6490448782134934e-07, "loss": 1.7653579711914062, "step": 2615 }, { "epoch": 4.0684026428293825, "grad_norm": 2.46875, "learning_rate": 2.6068190607627395e-07, "loss": 1.7767929077148437, "step": 2620 }, { "epoch": 4.07617567042363, "grad_norm": 4.34375, "learning_rate": 2.564900444520737e-07, "loss": 1.7532543182373046, "step": 2625 }, { "epoch": 4.083948698017878, "grad_norm": 2.609375, "learning_rate": 2.523290068562896e-07, "loss": 1.7864208221435547, "step": 2630 }, { "epoch": 4.091721725612126, "grad_norm": 4.09375, "learning_rate": 2.4819889643239987e-07, "loss": 1.7288122177124023, "step": 2635 }, { "epoch": 4.099494753206374, "grad_norm": 2.75, "learning_rate": 2.4409981555726053e-07, "loss": 1.7514692306518556, "step": 2640 }, { "epoch": 4.107267780800622, "grad_norm": 2.5, "learning_rate": 2.400318658385717e-07, "loss": 1.7463350296020508, "step": 2645 }, { "epoch": 4.11504080839487, "grad_norm": 3.4375, "learning_rate": 2.3599514811235534e-07, "loss": 1.7647014617919923, "step": 2650 }, { "epoch": 4.122813835989118, "grad_norm": 2.84375, "learning_rate": 2.3198976244045733e-07, "loss": 1.7601783752441407, "step": 2655 }, { "epoch": 4.130586863583366, "grad_norm": 2.78125, "learning_rate": 2.280158081080681e-07, "loss": 1.7404991149902345, "step": 2660 }, { "epoch": 4.138359891177614, "grad_norm": 3.59375, "learning_rate": 2.240733836212588e-07, "loss": 1.7805627822875976, "step": 2665 }, { "epoch": 4.146132918771862, "grad_norm": 2.546875, "learning_rate": 2.2016258670454286e-07, "loss": 1.7652338027954102, "step": 2670 }, { "epoch": 4.15390594636611, "grad_norm": 5.25, "learning_rate": 2.1628351429845078e-07, "loss": 1.768269920349121, "step": 2675 }, { "epoch": 4.161678973960358, "grad_norm": 2.96875, "learning_rate": 2.12436262557129e-07, "loss": 1.778640365600586, "step": 2680 }, { "epoch": 4.169452001554606, "grad_norm": 2.828125, "learning_rate": 2.0862092684595496e-07, "loss": 1.7216337203979493, "step": 2685 }, { "epoch": 4.177225029148853, "grad_norm": 2.90625, "learning_rate": 2.048376017391756e-07, "loss": 1.7343669891357423, "step": 2690 }, { "epoch": 4.184998056743101, "grad_norm": 2.390625, "learning_rate": 2.0108638101755972e-07, "loss": 1.7394224166870118, "step": 2695 }, { "epoch": 4.192771084337349, "grad_norm": 2.65625, "learning_rate": 1.973673576660759e-07, "loss": 1.7478919982910157, "step": 2700 }, { "epoch": 4.200544111931597, "grad_norm": 3.234375, "learning_rate": 1.9368062387158725e-07, "loss": 1.769978141784668, "step": 2705 }, { "epoch": 4.208317139525845, "grad_norm": 3.328125, "learning_rate": 1.9002627102056518e-07, "loss": 1.7497880935668946, "step": 2710 }, { "epoch": 4.216090167120093, "grad_norm": 2.484375, "learning_rate": 1.8640438969682466e-07, "loss": 1.7618110656738282, "step": 2715 }, { "epoch": 4.223863194714341, "grad_norm": 2.296875, "learning_rate": 1.828150696792798e-07, "loss": 1.7837394714355468, "step": 2720 }, { "epoch": 4.231636222308589, "grad_norm": 2.796875, "learning_rate": 1.7925839993971637e-07, "loss": 1.77918701171875, "step": 2725 }, { "epoch": 4.239409249902837, "grad_norm": 2.96875, "learning_rate": 1.7573446864058906e-07, "loss": 1.7677013397216796, "step": 2730 }, { "epoch": 4.247182277497085, "grad_norm": 3.90625, "learning_rate": 1.7224336313283352e-07, "loss": 1.7647846221923829, "step": 2735 }, { "epoch": 4.254955305091333, "grad_norm": 3.609375, "learning_rate": 1.6878516995370207e-07, "loss": 1.7514135360717773, "step": 2740 }, { "epoch": 4.262728332685581, "grad_norm": 2.515625, "learning_rate": 1.6535997482462006e-07, "loss": 1.7736982345581054, "step": 2745 }, { "epoch": 4.270501360279829, "grad_norm": 2.78125, "learning_rate": 1.6196786264905823e-07, "loss": 1.7826234817504882, "step": 2750 }, { "epoch": 4.2782743878740765, "grad_norm": 2.4375, "learning_rate": 1.5860891751043095e-07, "loss": 1.7481828689575196, "step": 2755 }, { "epoch": 4.286047415468325, "grad_norm": 3.4375, "learning_rate": 1.552832226700096e-07, "loss": 1.774028205871582, "step": 2760 }, { "epoch": 4.293820443062573, "grad_norm": 2.515625, "learning_rate": 1.5199086056486006e-07, "loss": 1.7415176391601563, "step": 2765 }, { "epoch": 4.301593470656821, "grad_norm": 2.890625, "learning_rate": 1.4873191280579955e-07, "loss": 1.7713994979858398, "step": 2770 }, { "epoch": 4.309366498251069, "grad_norm": 2.796875, "learning_rate": 1.4550646017537243e-07, "loss": 1.7674455642700195, "step": 2775 }, { "epoch": 4.317139525845317, "grad_norm": 3.125, "learning_rate": 1.42314582625848e-07, "loss": 1.7799386978149414, "step": 2780 }, { "epoch": 4.324912553439565, "grad_norm": 2.40625, "learning_rate": 1.391563592772403e-07, "loss": 1.7861083984375, "step": 2785 }, { "epoch": 4.332685581033813, "grad_norm": 3.234375, "learning_rate": 1.360318684153448e-07, "loss": 1.7898414611816407, "step": 2790 }, { "epoch": 4.340458608628061, "grad_norm": 2.828125, "learning_rate": 1.3294118748979878e-07, "loss": 1.7847719192504883, "step": 2795 }, { "epoch": 4.348231636222309, "grad_norm": 3.078125, "learning_rate": 1.298843931121611e-07, "loss": 1.7442861557006837, "step": 2800 }, { "epoch": 4.356004663816557, "grad_norm": 2.515625, "learning_rate": 1.268615610540148e-07, "loss": 1.7581628799438476, "step": 2805 }, { "epoch": 4.363777691410805, "grad_norm": 2.328125, "learning_rate": 1.2387276624508598e-07, "loss": 1.7922929763793944, "step": 2810 }, { "epoch": 4.371550719005053, "grad_norm": 3.0625, "learning_rate": 1.2091808277138894e-07, "loss": 1.7822298049926757, "step": 2815 }, { "epoch": 4.3793237465993, "grad_norm": 2.5625, "learning_rate": 1.1799758387338889e-07, "loss": 1.777001953125, "step": 2820 }, { "epoch": 4.387096774193548, "grad_norm": 2.5, "learning_rate": 1.1511134194418582e-07, "loss": 1.771342658996582, "step": 2825 }, { "epoch": 4.394869801787796, "grad_norm": 2.421875, "learning_rate": 1.1225942852772137e-07, "loss": 1.7197996139526368, "step": 2830 }, { "epoch": 4.402642829382044, "grad_norm": 2.375, "learning_rate": 1.0944191431700429e-07, "loss": 1.763011360168457, "step": 2835 }, { "epoch": 4.410415856976292, "grad_norm": 2.515625, "learning_rate": 1.0665886915235795e-07, "loss": 1.737653923034668, "step": 2840 }, { "epoch": 4.41818888457054, "grad_norm": 3.796875, "learning_rate": 1.0391036201969101e-07, "loss": 1.7512134552001952, "step": 2845 }, { "epoch": 4.425961912164788, "grad_norm": 2.5, "learning_rate": 1.011964610487846e-07, "loss": 1.77310791015625, "step": 2850 }, { "epoch": 4.433734939759036, "grad_norm": 2.578125, "learning_rate": 9.851723351160641e-08, "loss": 1.7227460861206054, "step": 2855 }, { "epoch": 4.441507967353284, "grad_norm": 3.28125, "learning_rate": 9.587274582064026e-08, "loss": 1.7299663543701171, "step": 2860 }, { "epoch": 4.449280994947532, "grad_norm": 2.53125, "learning_rate": 9.326306352724268e-08, "loss": 1.7719409942626954, "step": 2865 }, { "epoch": 4.45705402254178, "grad_norm": 3.140625, "learning_rate": 9.068825132001596e-08, "loss": 1.7672760009765625, "step": 2870 }, { "epoch": 4.464827050136028, "grad_norm": 2.75, "learning_rate": 8.814837302320516e-08, "loss": 1.7817211151123047, "step": 2875 }, { "epoch": 4.472600077730276, "grad_norm": 2.828125, "learning_rate": 8.564349159511736e-08, "loss": 1.690296745300293, "step": 2880 }, { "epoch": 4.4803731053245235, "grad_norm": 3.34375, "learning_rate": 8.317366912655905e-08, "loss": 1.7582836151123047, "step": 2885 }, { "epoch": 4.4881461329187715, "grad_norm": 3.46875, "learning_rate": 8.073896683929788e-08, "loss": 1.7542804718017577, "step": 2890 }, { "epoch": 4.4959191605130195, "grad_norm": 2.515625, "learning_rate": 7.833944508454588e-08, "loss": 1.7722579956054687, "step": 2895 }, { "epoch": 4.503692188107268, "grad_norm": 3.546875, "learning_rate": 7.597516334146198e-08, "loss": 1.748293685913086, "step": 2900 }, { "epoch": 4.511465215701516, "grad_norm": 3.03125, "learning_rate": 7.36461802156792e-08, "loss": 1.7780683517456055, "step": 2905 }, { "epoch": 4.519238243295764, "grad_norm": 3.25, "learning_rate": 7.135255343785029e-08, "loss": 1.782621192932129, "step": 2910 }, { "epoch": 4.527011270890012, "grad_norm": 3.65625, "learning_rate": 6.909433986221875e-08, "loss": 1.7398847579956054, "step": 2915 }, { "epoch": 4.53478429848426, "grad_norm": 3.234375, "learning_rate": 6.687159546520722e-08, "loss": 1.7721050262451172, "step": 2920 }, { "epoch": 4.542557326078508, "grad_norm": 2.8125, "learning_rate": 6.468437534403165e-08, "loss": 1.778801155090332, "step": 2925 }, { "epoch": 4.550330353672756, "grad_norm": 2.828125, "learning_rate": 6.253273371533508e-08, "loss": 1.7390150070190429, "step": 2930 }, { "epoch": 4.558103381267004, "grad_norm": 2.75, "learning_rate": 6.041672391384311e-08, "loss": 1.7798133850097657, "step": 2935 }, { "epoch": 4.565876408861252, "grad_norm": 2.78125, "learning_rate": 5.8336398391043264e-08, "loss": 1.7365947723388673, "step": 2940 }, { "epoch": 4.573649436455499, "grad_norm": 2.625, "learning_rate": 5.6291808713883064e-08, "loss": 1.7829244613647461, "step": 2945 }, { "epoch": 4.581422464049748, "grad_norm": 2.765625, "learning_rate": 5.4283005563492806e-08, "loss": 1.746117401123047, "step": 2950 }, { "epoch": 4.589195491643995, "grad_norm": 2.171875, "learning_rate": 5.23100387339297e-08, "loss": 1.7521102905273438, "step": 2955 }, { "epoch": 4.596968519238243, "grad_norm": 3.28125, "learning_rate": 5.03729571309422e-08, "loss": 1.7810224533081054, "step": 2960 }, { "epoch": 4.604741546832491, "grad_norm": 2.765625, "learning_rate": 4.847180877075946e-08, "loss": 1.7393953323364257, "step": 2965 }, { "epoch": 4.612514574426739, "grad_norm": 4.0625, "learning_rate": 4.6606640778899476e-08, "loss": 1.7397171020507813, "step": 2970 }, { "epoch": 4.620287602020987, "grad_norm": 2.90625, "learning_rate": 4.477749938900183e-08, "loss": 1.7733592987060547, "step": 2975 }, { "epoch": 4.628060629615235, "grad_norm": 4.71875, "learning_rate": 4.298442994168161e-08, "loss": 1.7416141510009766, "step": 2980 }, { "epoch": 4.635833657209483, "grad_norm": 4.0625, "learning_rate": 4.1227476883405504e-08, "loss": 1.745661735534668, "step": 2985 }, { "epoch": 4.643606684803731, "grad_norm": 2.515625, "learning_rate": 3.950668376538963e-08, "loss": 1.7503889083862305, "step": 2990 }, { "epoch": 4.651379712397979, "grad_norm": 2.546875, "learning_rate": 3.782209324252012e-08, "loss": 1.7682682037353517, "step": 2995 }, { "epoch": 4.659152739992227, "grad_norm": 2.78125, "learning_rate": 3.6173747072296756e-08, "loss": 1.7788461685180663, "step": 3000 }, { "epoch": 4.666925767586475, "grad_norm": 3.703125, "learning_rate": 3.4561686113796674e-08, "loss": 1.7764659881591798, "step": 3005 }, { "epoch": 4.674698795180722, "grad_norm": 3.09375, "learning_rate": 3.298595032666163e-08, "loss": 1.7665237426757812, "step": 3010 }, { "epoch": 4.682471822774971, "grad_norm": 4.09375, "learning_rate": 3.1446578770108646e-08, "loss": 1.7437681198120116, "step": 3015 }, { "epoch": 4.6902448503692185, "grad_norm": 3.828125, "learning_rate": 2.99436096019603e-08, "loss": 1.713943099975586, "step": 3020 }, { "epoch": 4.6980178779634665, "grad_norm": 2.71875, "learning_rate": 2.847708007769978e-08, "loss": 1.7692811965942383, "step": 3025 }, { "epoch": 4.7057909055577145, "grad_norm": 3.4375, "learning_rate": 2.7047026549547494e-08, "loss": 1.763226318359375, "step": 3030 }, { "epoch": 4.7135639331519625, "grad_norm": 3.203125, "learning_rate": 2.565348446555943e-08, "loss": 1.7892520904541016, "step": 3035 }, { "epoch": 4.721336960746211, "grad_norm": 3.359375, "learning_rate": 2.429648836874887e-08, "loss": 1.769230270385742, "step": 3040 }, { "epoch": 4.729109988340459, "grad_norm": 3.484375, "learning_rate": 2.2976071896229577e-08, "loss": 1.7419490814208984, "step": 3045 }, { "epoch": 4.736883015934707, "grad_norm": 2.421875, "learning_rate": 2.1692267778383123e-08, "loss": 1.7890913009643554, "step": 3050 }, { "epoch": 4.744656043528955, "grad_norm": 2.984375, "learning_rate": 2.044510783804654e-08, "loss": 1.7666656494140625, "step": 3055 }, { "epoch": 4.752429071123203, "grad_norm": 2.703125, "learning_rate": 1.9234622989723282e-08, "loss": 1.776426124572754, "step": 3060 }, { "epoch": 4.760202098717451, "grad_norm": 2.65625, "learning_rate": 1.8060843238818513e-08, "loss": 1.7808853149414063, "step": 3065 }, { "epoch": 4.767975126311699, "grad_norm": 2.421875, "learning_rate": 1.6923797680893362e-08, "loss": 1.7926462173461915, "step": 3070 }, { "epoch": 4.775748153905946, "grad_norm": 3.65625, "learning_rate": 1.582351450094516e-08, "loss": 1.7616849899291993, "step": 3075 }, { "epoch": 4.783521181500195, "grad_norm": 3.6875, "learning_rate": 1.4760020972708343e-08, "loss": 1.764903450012207, "step": 3080 }, { "epoch": 4.791294209094442, "grad_norm": 2.859375, "learning_rate": 1.3733343457977665e-08, "loss": 1.769833755493164, "step": 3085 }, { "epoch": 4.79906723668869, "grad_norm": 3.078125, "learning_rate": 1.2743507405955867e-08, "loss": 1.7617315292358398, "step": 3090 }, { "epoch": 4.806840264282938, "grad_norm": 2.8125, "learning_rate": 1.1790537352622533e-08, "loss": 1.788639450073242, "step": 3095 }, { "epoch": 4.814613291877186, "grad_norm": 2.84375, "learning_rate": 1.0874456920125397e-08, "loss": 1.7513429641723632, "step": 3100 }, { "epoch": 4.822386319471434, "grad_norm": 2.796875, "learning_rate": 9.995288816195158e-09, "loss": 1.7295230865478515, "step": 3105 }, { "epoch": 4.830159347065682, "grad_norm": 3.390625, "learning_rate": 9.153054833582752e-09, "loss": 1.7681365966796876, "step": 3110 }, { "epoch": 4.83793237465993, "grad_norm": 3.25, "learning_rate": 8.347775849518791e-09, "loss": 1.744300651550293, "step": 3115 }, { "epoch": 4.845705402254178, "grad_norm": 2.65625, "learning_rate": 7.579471825196305e-09, "loss": 1.7566238403320313, "step": 3120 }, { "epoch": 4.853478429848426, "grad_norm": 2.96875, "learning_rate": 6.848161805275976e-09, "loss": 1.7817750930786134, "step": 3125 }, { "epoch": 4.861251457442674, "grad_norm": 3.21875, "learning_rate": 6.153863917414015e-09, "loss": 1.755167579650879, "step": 3130 }, { "epoch": 4.869024485036922, "grad_norm": 2.625, "learning_rate": 5.496595371812352e-09, "loss": 1.7471593856811523, "step": 3135 }, { "epoch": 4.87679751263117, "grad_norm": 2.640625, "learning_rate": 4.876372460793143e-09, "loss": 1.767136001586914, "step": 3140 }, { "epoch": 4.884570540225418, "grad_norm": 2.859375, "learning_rate": 4.293210558393435e-09, "loss": 1.7452085494995118, "step": 3145 }, { "epoch": 4.892343567819665, "grad_norm": 2.984375, "learning_rate": 3.7471241199856235e-09, "loss": 1.7779275894165039, "step": 3150 }, { "epoch": 4.900116595413913, "grad_norm": 2.625, "learning_rate": 3.2381266819180854e-09, "loss": 1.7648542404174805, "step": 3155 }, { "epoch": 4.9078896230081615, "grad_norm": 3.0625, "learning_rate": 2.7662308611794417e-09, "loss": 1.7670942306518556, "step": 3160 }, { "epoch": 4.9156626506024095, "grad_norm": 2.859375, "learning_rate": 2.3314483550869736e-09, "loss": 1.7594139099121093, "step": 3165 }, { "epoch": 4.9234356781966575, "grad_norm": 2.421875, "learning_rate": 1.933789940995856e-09, "loss": 1.7814929962158204, "step": 3170 }, { "epoch": 4.9312087057909055, "grad_norm": 2.953125, "learning_rate": 1.5732654760320376e-09, "loss": 1.7994659423828125, "step": 3175 }, { "epoch": 4.938981733385154, "grad_norm": 6.0625, "learning_rate": 1.24988389684777e-09, "loss": 1.7598079681396483, "step": 3180 }, { "epoch": 4.946754760979402, "grad_norm": 3.5, "learning_rate": 9.636532194006176e-10, "loss": 1.7375944137573243, "step": 3185 }, { "epoch": 4.95452778857365, "grad_norm": 3.625, "learning_rate": 7.145805387547832e-10, "loss": 1.7086051940917968, "step": 3190 }, { "epoch": 4.962300816167898, "grad_norm": 3.140625, "learning_rate": 5.026720289042497e-10, "loss": 1.7740161895751954, "step": 3195 }, { "epoch": 4.970073843762146, "grad_norm": 2.484375, "learning_rate": 3.2793294262073446e-10, "loss": 1.751610565185547, "step": 3200 }, { "epoch": 4.977846871356394, "grad_norm": 2.546875, "learning_rate": 1.903676113232944e-10, "loss": 1.759403610229492, "step": 3205 }, { "epoch": 4.985619898950642, "grad_norm": 2.75, "learning_rate": 8.997944497074473e-11, "loss": 1.7566078186035157, "step": 3210 }, { "epoch": 4.993392926544889, "grad_norm": 3.4375, "learning_rate": 2.6770931977060375e-11, "loss": 1.7355569839477538, "step": 3215 }, { "epoch": 5.0, "grad_norm": 8.3125, "learning_rate": 7.43639149924924e-13, "loss": 1.7696298599243163, "step": 3220 }, { "epoch": 5.0, "step": 3220, "total_flos": 1.5488981277124608e+18, "train_loss": 1.8179572771795047, "train_runtime": 26174.2071, "train_samples_per_second": 0.492, "train_steps_per_second": 0.123 } ], "logging_steps": 5, "max_steps": 3220, "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": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.5488981277124608e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }