| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.8720626631853787, | |
| "eval_steps": 250, | |
| "global_step": 4400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0032637075718015664, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 3.7499999999999997e-06, | |
| "loss": 0.9583, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006527415143603133, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 7.499999999999999e-06, | |
| "loss": 0.9966, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0097911227154047, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 1.1249999999999999e-05, | |
| "loss": 0.8656, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013054830287206266, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 1.4999999999999999e-05, | |
| "loss": 0.9513, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016318537859007835, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.875e-05, | |
| "loss": 0.9421, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0195822454308094, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 2.2499999999999998e-05, | |
| "loss": 0.8782, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.022845953002610966, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 2.6249999999999998e-05, | |
| "loss": 0.9951, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02610966057441253, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 0.7883, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0293733681462141, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 3.375e-05, | |
| "loss": 0.7147, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03263707571801567, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.7809, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03590078328981723, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 4.125e-05, | |
| "loss": 0.6191, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0391644908616188, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 0.8668, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04242819843342036, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 4.875e-05, | |
| "loss": 0.7034, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04569190600522193, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5.2499999999999995e-05, | |
| "loss": 0.6424, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0489556135770235, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 5.625e-05, | |
| "loss": 0.6463, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05221932114882506, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 0.6884, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05548302872062663, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 6.374999999999999e-05, | |
| "loss": 0.7838, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0587467362924282, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 6.75e-05, | |
| "loss": 0.48, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06201044386422976, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 7.125e-05, | |
| "loss": 0.5492, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06527415143603134, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 7.5e-05, | |
| "loss": 0.4949, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0685378590078329, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.499977113046922e-05, | |
| "loss": 0.5546, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07180156657963446, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 7.499908452467055e-05, | |
| "loss": 0.7387, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07506527415143603, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 7.499794019098496e-05, | |
| "loss": 0.5367, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0783289817232376, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 7.499633814338062e-05, | |
| "loss": 0.4813, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08159268929503917, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 7.499427840141272e-05, | |
| "loss": 0.5815, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08485639686684072, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.499176099022322e-05, | |
| "loss": 0.4612, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0881201044386423, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 7.498878594054062e-05, | |
| "loss": 0.3535, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09138381201044386, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 7.498535328867947e-05, | |
| "loss": 0.3999, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09464751958224543, | |
| "grad_norm": 0.125, | |
| "learning_rate": 7.498146307654004e-05, | |
| "loss": 0.4473, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.097911227154047, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 7.497711535160767e-05, | |
| "loss": 0.5172, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10117493472584857, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 7.497231016695234e-05, | |
| "loss": 0.5024, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10443864229765012, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 7.496704758122795e-05, | |
| "loss": 0.5156, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1077023498694517, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 7.496132765867157e-05, | |
| "loss": 0.5329, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.11096605744125326, | |
| "grad_norm": 0.09130859375, | |
| "learning_rate": 7.495515046910274e-05, | |
| "loss": 0.4888, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11422976501305483, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 7.494851608792255e-05, | |
| "loss": 0.6306, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1174934725848564, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 7.494142459611273e-05, | |
| "loss": 0.4552, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12075718015665797, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 7.49338760802347e-05, | |
| "loss": 0.5988, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12402088772845953, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 7.492587063242849e-05, | |
| "loss": 0.4453, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1272845953002611, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 7.491740835041156e-05, | |
| "loss": 0.499, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.13054830287206268, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 7.490848933747772e-05, | |
| "loss": 0.4354, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13381201044386423, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 7.489911370249579e-05, | |
| "loss": 0.5372, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1370757180156658, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 7.488928155990828e-05, | |
| "loss": 0.6053, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.14033942558746737, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 7.487899302973e-05, | |
| "loss": 0.4278, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.14360313315926893, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 7.486824823754662e-05, | |
| "loss": 0.5667, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1468668407310705, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.485704731451311e-05, | |
| "loss": 0.8726, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.15013054830287206, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 7.484539039735212e-05, | |
| "loss": 0.429, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15339425587467362, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 7.483327762835236e-05, | |
| "loss": 0.401, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1566579634464752, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 7.482070915536683e-05, | |
| "loss": 0.4806, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15992167101827676, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 7.480768513181102e-05, | |
| "loss": 0.62, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.16318537859007834, | |
| "grad_norm": 0.0625, | |
| "learning_rate": 7.479420571666103e-05, | |
| "loss": 0.571, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16318537859007834, | |
| "eval_loss": 0.47697925567626953, | |
| "eval_runtime": 97.2401, | |
| "eval_samples_per_second": 1.707, | |
| "eval_steps_per_second": 0.216, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1664490861618799, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 7.47802710744517e-05, | |
| "loss": 0.3673, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.16971279373368145, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.476588137527446e-05, | |
| "loss": 0.4231, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.17297650130548303, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.475103679477539e-05, | |
| "loss": 0.5686, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.1762402088772846, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.473573751415299e-05, | |
| "loss": 0.3383, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.17950391644908617, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 7.471998372015603e-05, | |
| "loss": 0.4023, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.18276762402088773, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 7.470377560508124e-05, | |
| "loss": 0.4471, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1860313315926893, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 7.468711336677091e-05, | |
| "loss": 0.4911, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.18929503916449086, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 7.466999720861061e-05, | |
| "loss": 0.4399, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.19255874673629242, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 7.465242733952656e-05, | |
| "loss": 0.3883, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.195822454308094, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 7.463440397398319e-05, | |
| "loss": 0.5332, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19908616187989556, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 7.461592733198044e-05, | |
| "loss": 0.3769, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.20234986945169714, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 7.459699763905112e-05, | |
| "loss": 0.6075, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2056135770234987, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 7.45776151262582e-05, | |
| "loss": 0.3442, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.20887728459530025, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 7.455778003019187e-05, | |
| "loss": 0.3987, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.21214099216710183, | |
| "grad_norm": 0.0498046875, | |
| "learning_rate": 7.453749259296675e-05, | |
| "loss": 0.377, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2154046997389034, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 7.451675306221892e-05, | |
| "loss": 0.4471, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.21866840731070497, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 7.449556169110285e-05, | |
| "loss": 0.5509, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.22193211488250653, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 7.447391873828839e-05, | |
| "loss": 0.5118, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.22519582245430808, | |
| "grad_norm": 0.0546875, | |
| "learning_rate": 7.44518244679575e-05, | |
| "loss": 0.4309, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.22845953002610966, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 7.442927914980116e-05, | |
| "loss": 0.3991, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.23172323759791122, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 7.440628305901597e-05, | |
| "loss": 0.5483, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.2349869451697128, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 7.438283647630083e-05, | |
| "loss": 0.4699, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.23825065274151436, | |
| "grad_norm": 0.046630859375, | |
| "learning_rate": 7.435893968785353e-05, | |
| "loss": 0.4702, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.24151436031331594, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 7.433459298536721e-05, | |
| "loss": 0.4805, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2447780678851175, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 7.430979666602687e-05, | |
| "loss": 0.4445, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.24804177545691905, | |
| "grad_norm": 0.06787109375, | |
| "learning_rate": 7.428455103250569e-05, | |
| "loss": 0.5559, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.25130548302872063, | |
| "grad_norm": 0.07666015625, | |
| "learning_rate": 7.425885639296131e-05, | |
| "loss": 0.5209, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.2545691906005222, | |
| "grad_norm": 0.03955078125, | |
| "learning_rate": 7.423271306103217e-05, | |
| "loss": 0.3874, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.25783289817232374, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 7.420612135583353e-05, | |
| "loss": 0.4156, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.26109660574412535, | |
| "grad_norm": 0.0625, | |
| "learning_rate": 7.417908160195377e-05, | |
| "loss": 0.36, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2643603133159269, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 7.415159412945024e-05, | |
| "loss": 0.4679, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.26762402088772846, | |
| "grad_norm": 0.09130859375, | |
| "learning_rate": 7.412365927384536e-05, | |
| "loss": 0.4482, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.27088772845953, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 7.409527737612241e-05, | |
| "loss": 0.4426, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.2741514360313316, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 7.406644878272151e-05, | |
| "loss": 0.454, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2774151436031332, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 7.403717384553527e-05, | |
| "loss": 0.4397, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.28067885117493474, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 7.400745292190453e-05, | |
| "loss": 0.5127, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2839425587467363, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 7.397728637461407e-05, | |
| "loss": 0.4335, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.28720626631853785, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 7.394667457188805e-05, | |
| "loss": 0.6522, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.2904699738903394, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 7.391561788738562e-05, | |
| "loss": 0.5314, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.293733681462141, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 7.388411670019631e-05, | |
| "loss": 0.4786, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.29699738903394257, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 7.385217139483543e-05, | |
| "loss": 0.4069, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3002610966057441, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 7.381978236123936e-05, | |
| "loss": 0.4885, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3035248041775457, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 7.378694999476078e-05, | |
| "loss": 0.4406, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.30678851174934724, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 7.375367469616387e-05, | |
| "loss": 0.8256, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.31005221932114885, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 7.37199568716194e-05, | |
| "loss": 0.4636, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.3133159268929504, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 7.36857969326998e-05, | |
| "loss": 0.6288, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.31657963446475196, | |
| "grad_norm": 0.05810546875, | |
| "learning_rate": 7.365119529637406e-05, | |
| "loss": 0.3789, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.3198433420365535, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 7.361615238500275e-05, | |
| "loss": 0.6206, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.32310704960835507, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 7.358066862633276e-05, | |
| "loss": 0.3703, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3263707571801567, | |
| "grad_norm": 0.057373046875, | |
| "learning_rate": 7.35447444534922e-05, | |
| "loss": 0.401, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3263707571801567, | |
| "eval_loss": 0.46315205097198486, | |
| "eval_runtime": 97.1749, | |
| "eval_samples_per_second": 1.708, | |
| "eval_steps_per_second": 0.216, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32963446475195823, | |
| "grad_norm": 0.04833984375, | |
| "learning_rate": 7.350838030498496e-05, | |
| "loss": 0.4603, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.3328981723237598, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 7.347157662468547e-05, | |
| "loss": 0.4309, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.33616187989556134, | |
| "grad_norm": 0.057373046875, | |
| "learning_rate": 7.343433386183327e-05, | |
| "loss": 0.3169, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.3394255874673629, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 7.339665247102748e-05, | |
| "loss": 0.5101, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3426892950391645, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 7.33585329122213e-05, | |
| "loss": 0.459, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.34595300261096606, | |
| "grad_norm": 0.056640625, | |
| "learning_rate": 7.331997565071634e-05, | |
| "loss": 0.3522, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3492167101827676, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 7.328098115715699e-05, | |
| "loss": 0.4109, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.3524804177545692, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 7.324154990752467e-05, | |
| "loss": 0.4155, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.35574412532637073, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 7.3201682383132e-05, | |
| "loss": 0.4897, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.35900783289817234, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 7.316137907061691e-05, | |
| "loss": 0.4048, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3622715404699739, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 7.312064046193679e-05, | |
| "loss": 0.5676, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.36553524804177545, | |
| "grad_norm": 0.060546875, | |
| "learning_rate": 7.307946705436233e-05, | |
| "loss": 0.3708, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.368798955613577, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 7.30378593504716e-05, | |
| "loss": 0.4704, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.3720626631853786, | |
| "grad_norm": 0.052490234375, | |
| "learning_rate": 7.299581785814385e-05, | |
| "loss": 0.5599, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.37532637075718017, | |
| "grad_norm": 0.048828125, | |
| "learning_rate": 7.295334309055327e-05, | |
| "loss": 0.3238, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3785900783289817, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 7.291043556616283e-05, | |
| "loss": 0.7988, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3818537859007833, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 7.286709580871783e-05, | |
| "loss": 0.4842, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.38511749347258484, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 7.282332434723962e-05, | |
| "loss": 0.5732, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.38838120104438645, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 7.277912171601907e-05, | |
| "loss": 0.4025, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.391644908616188, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 7.273448845461008e-05, | |
| "loss": 0.3352, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.39490861618798956, | |
| "grad_norm": 0.045166015625, | |
| "learning_rate": 7.268942510782298e-05, | |
| "loss": 0.3121, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.3981723237597911, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 7.264393222571783e-05, | |
| "loss": 0.3868, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.40143603133159267, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 7.259801036359786e-05, | |
| "loss": 0.4367, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4046997389033943, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 7.255166008200252e-05, | |
| "loss": 0.472, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.40796344647519583, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 7.250488194670071e-05, | |
| "loss": 0.6129, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4112271540469974, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 7.245767652868393e-05, | |
| "loss": 0.4201, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.41449086161879894, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 7.241004440415919e-05, | |
| "loss": 0.43, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.4177545691906005, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.236198615454207e-05, | |
| "loss": 0.9362, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.4210182767624021, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 7.231350236644957e-05, | |
| "loss": 0.5425, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.42428198433420367, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 7.2264593631693e-05, | |
| "loss": 0.4919, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4275456919060052, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 7.221526054727071e-05, | |
| "loss": 0.475, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.4308093994778068, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 7.216550371536083e-05, | |
| "loss": 0.4382, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.43407310704960833, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 7.21153237433139e-05, | |
| "loss": 0.4489, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.43733681462140994, | |
| "grad_norm": 0.05859375, | |
| "learning_rate": 7.20647212436455e-05, | |
| "loss": 0.4689, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.4406005221932115, | |
| "grad_norm": 0.0546875, | |
| "learning_rate": 7.201369683402869e-05, | |
| "loss": 0.3707, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.44386422976501305, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 7.196225113728656e-05, | |
| "loss": 0.4597, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.4471279373368146, | |
| "grad_norm": 0.06787109375, | |
| "learning_rate": 7.191038478138458e-05, | |
| "loss": 0.5113, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.45039164490861616, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 7.18580983994229e-05, | |
| "loss": 0.3574, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.4536553524804178, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 7.180539262962875e-05, | |
| "loss": 0.853, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.45691906005221933, | |
| "grad_norm": 0.038818359375, | |
| "learning_rate": 7.175226811534853e-05, | |
| "loss": 0.6318, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4601827676240209, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 7.169872550503992e-05, | |
| "loss": 0.5808, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.46344647519582244, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 7.164476545226415e-05, | |
| "loss": 0.5037, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.466710182767624, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 7.159038861567783e-05, | |
| "loss": 0.5246, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.4699738903394256, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 7.153559565902503e-05, | |
| "loss": 0.5053, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.47323759791122716, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 7.148038725112913e-05, | |
| "loss": 0.472, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.4765013054830287, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 7.142476406588466e-05, | |
| "loss": 0.5306, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.47976501305483027, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 7.136872678224905e-05, | |
| "loss": 0.5424, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.4830287206266319, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 7.131227608423442e-05, | |
| "loss": 0.3975, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.48629242819843344, | |
| "grad_norm": 0.059814453125, | |
| "learning_rate": 7.125541266089917e-05, | |
| "loss": 0.6061, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.489556135770235, | |
| "grad_norm": 0.05224609375, | |
| "learning_rate": 7.119813720633953e-05, | |
| "loss": 0.465, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.489556135770235, | |
| "eval_loss": 0.4532851278781891, | |
| "eval_runtime": 97.0469, | |
| "eval_samples_per_second": 1.711, | |
| "eval_steps_per_second": 0.216, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.49281984334203655, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.114045041968121e-05, | |
| "loss": 0.5156, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.4960835509138381, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 7.108235300507073e-05, | |
| "loss": 0.4453, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4993472584856397, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 7.102384567166694e-05, | |
| "loss": 0.6377, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.5026109660574413, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 7.096492913363228e-05, | |
| "loss": 0.463, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5058746736292428, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 7.090560411012412e-05, | |
| "loss": 0.4592, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.5091383812010444, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 7.084587132528592e-05, | |
| "loss": 0.422, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.512402088772846, | |
| "grad_norm": 0.1064453125, | |
| "learning_rate": 7.078573150823847e-05, | |
| "loss": 0.5943, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.5156657963446475, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 7.072518539307092e-05, | |
| "loss": 0.3763, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5189295039164491, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 7.066423371883183e-05, | |
| "loss": 0.5364, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.5221932114882507, | |
| "grad_norm": 0.0595703125, | |
| "learning_rate": 7.060287722952024e-05, | |
| "loss": 0.4057, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5254569190600522, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 7.054111667407642e-05, | |
| "loss": 0.3785, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.5287206266318538, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 7.04789528063729e-05, | |
| "loss": 0.5555, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5319843342036553, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 7.041638638520513e-05, | |
| "loss": 0.4257, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.5352480417754569, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 7.035341817428235e-05, | |
| "loss": 0.4849, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.5385117493472585, | |
| "grad_norm": 0.06103515625, | |
| "learning_rate": 7.029004894221812e-05, | |
| "loss": 0.3831, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.54177545691906, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 7.022627946252105e-05, | |
| "loss": 0.5633, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.5450391644908616, | |
| "grad_norm": 0.0927734375, | |
| "learning_rate": 7.016211051358534e-05, | |
| "loss": 0.4286, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.5483028720626631, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 7.009754287868123e-05, | |
| "loss": 0.4714, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5515665796344648, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 7.003257734594548e-05, | |
| "loss": 0.5155, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.5548302872062664, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 6.996721470837176e-05, | |
| "loss": 0.3779, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.5580939947780679, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 6.990145576380089e-05, | |
| "loss": 0.5331, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.5613577023498695, | |
| "grad_norm": 0.09130859375, | |
| "learning_rate": 6.983530131491127e-05, | |
| "loss": 0.536, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.564621409921671, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 6.976875216920886e-05, | |
| "loss": 0.4476, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.5678851174934726, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 6.970180913901751e-05, | |
| "loss": 0.495, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5711488250652742, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 6.963447304146893e-05, | |
| "loss": 0.6087, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.5744125326370757, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 6.956674469849279e-05, | |
| "loss": 0.3776, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5776762402088773, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 6.949862493680663e-05, | |
| "loss": 0.3869, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.5809399477806788, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 6.943011458790582e-05, | |
| "loss": 0.5347, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.5842036553524804, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 6.936121448805336e-05, | |
| "loss": 0.4809, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.587467362924282, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.929192547826971e-05, | |
| "loss": 0.4091, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5907310704960835, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 6.922224840432248e-05, | |
| "loss": 0.5521, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.5939947780678851, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 6.915218411671621e-05, | |
| "loss": 0.4078, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5972584856396866, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 6.908173347068183e-05, | |
| "loss": 0.5531, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.6005221932114883, | |
| "grad_norm": 0.052001953125, | |
| "learning_rate": 6.901089732616634e-05, | |
| "loss": 0.46, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6037859007832899, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 6.893967654782231e-05, | |
| "loss": 0.3918, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.6070496083550914, | |
| "grad_norm": 0.05615234375, | |
| "learning_rate": 6.886807200499723e-05, | |
| "loss": 0.4585, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.610313315926893, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 6.879608457172302e-05, | |
| "loss": 0.3973, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.6135770234986945, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 6.872371512670529e-05, | |
| "loss": 0.3886, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.6168407310704961, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 6.865096455331263e-05, | |
| "loss": 0.577, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.6201044386422977, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 6.857783373956578e-05, | |
| "loss": 0.4755, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.6233681462140992, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 6.850432357812692e-05, | |
| "loss": 0.5983, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.6266318537859008, | |
| "grad_norm": 0.03466796875, | |
| "learning_rate": 6.843043496628861e-05, | |
| "loss": 0.3643, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6298955613577023, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 6.835616880596298e-05, | |
| "loss": 0.3388, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.6331592689295039, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 6.82815260036706e-05, | |
| "loss": 0.4281, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.6364229765013055, | |
| "grad_norm": 0.05810546875, | |
| "learning_rate": 6.820650747052951e-05, | |
| "loss": 0.3822, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.639686684073107, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 6.81311141222441e-05, | |
| "loss": 0.493, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.6429503916449086, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 6.805534687909379e-05, | |
| "loss": 0.3711, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.6462140992167101, | |
| "grad_norm": 0.060546875, | |
| "learning_rate": 6.7979206665922e-05, | |
| "loss": 0.3482, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.6494778067885117, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 6.790269441212471e-05, | |
| "loss": 0.3724, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.6527415143603134, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 6.782581105163919e-05, | |
| "loss": 0.4655, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6527415143603134, | |
| "eval_loss": 0.44582512974739075, | |
| "eval_runtime": 97.0933, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6560052219321149, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 6.774855752293254e-05, | |
| "loss": 0.4789, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.6592689295039165, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 6.767093476899033e-05, | |
| "loss": 0.4196, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.662532637075718, | |
| "grad_norm": 0.052734375, | |
| "learning_rate": 6.759294373730496e-05, | |
| "loss": 0.3802, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.6657963446475196, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 6.751458537986426e-05, | |
| "loss": 0.4858, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.6690600522193212, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 6.74358606531397e-05, | |
| "loss": 0.4663, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.6723237597911227, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 6.735677051807481e-05, | |
| "loss": 0.5681, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.6755874673629243, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 6.727731594007343e-05, | |
| "loss": 0.5687, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.6788511749347258, | |
| "grad_norm": 0.06103515625, | |
| "learning_rate": 6.719749788898796e-05, | |
| "loss": 0.3671, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.6821148825065274, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 6.711731733910743e-05, | |
| "loss": 0.5654, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.685378590078329, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 6.703677526914571e-05, | |
| "loss": 0.4642, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6886422976501305, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 6.69558726622295e-05, | |
| "loss": 0.4356, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.6919060052219321, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 6.687461050588637e-05, | |
| "loss": 0.4206, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.6951697127937336, | |
| "grad_norm": 0.054931640625, | |
| "learning_rate": 6.679298979203266e-05, | |
| "loss": 0.4315, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.6984334203655352, | |
| "grad_norm": 0.056884765625, | |
| "learning_rate": 6.671101151696142e-05, | |
| "loss": 0.6175, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.7016971279373369, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 6.662867668133022e-05, | |
| "loss": 0.5087, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.7049608355091384, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 6.654598629014891e-05, | |
| "loss": 0.5919, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.70822454308094, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 6.646294135276744e-05, | |
| "loss": 0.4144, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.7114882506527415, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 6.637954288286343e-05, | |
| "loss": 0.514, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.7147519582245431, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 6.62957918984299e-05, | |
| "loss": 0.4498, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.7180156657963447, | |
| "grad_norm": 0.061767578125, | |
| "learning_rate": 6.621168942176275e-05, | |
| "loss": 0.3705, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7212793733681462, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.612723647944835e-05, | |
| "loss": 0.4258, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.7245430809399478, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 6.6042434102351e-05, | |
| "loss": 0.4407, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.7278067885117493, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 6.595728332560032e-05, | |
| "loss": 0.4246, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.7310704960835509, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 6.587178518857858e-05, | |
| "loss": 0.5558, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.7343342036553525, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 6.578594073490814e-05, | |
| "loss": 0.4846, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.737597911227154, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 6.569975101243855e-05, | |
| "loss": 0.3695, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.7408616187989556, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 6.561321707323392e-05, | |
| "loss": 0.499, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.7441253263707572, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 6.552633997355995e-05, | |
| "loss": 0.4023, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.7473890339425587, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 6.543912077387108e-05, | |
| "loss": 0.4374, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.7506527415143603, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 6.535156053879758e-05, | |
| "loss": 0.5137, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7539164490861618, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 6.52636603371325e-05, | |
| "loss": 0.4295, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.7571801566579635, | |
| "grad_norm": 0.05908203125, | |
| "learning_rate": 6.517542124181867e-05, | |
| "loss": 0.3519, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.7604438642297651, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 6.508684432993559e-05, | |
| "loss": 0.4373, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.7637075718015666, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 6.499793068268623e-05, | |
| "loss": 0.4145, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.7669712793733682, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 6.490868138538394e-05, | |
| "loss": 0.4133, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.7702349869451697, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 6.481909752743909e-05, | |
| "loss": 0.4096, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.7734986945169713, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 6.472918020234584e-05, | |
| "loss": 0.4889, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.7767624020887729, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 6.463893050766879e-05, | |
| "loss": 0.3739, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.7800261096605744, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 6.454834954502954e-05, | |
| "loss": 0.3955, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.783289817232376, | |
| "grad_norm": 0.05615234375, | |
| "learning_rate": 6.44574384200933e-05, | |
| "loss": 0.475, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7865535248041775, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 6.436619824255533e-05, | |
| "loss": 0.4624, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.7898172323759791, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 6.427463012612748e-05, | |
| "loss": 0.5408, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.7930809399477807, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 6.418273518852448e-05, | |
| "loss": 0.4784, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.7963446475195822, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 6.409051455145043e-05, | |
| "loss": 0.4466, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.7996083550913838, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 6.399796934058499e-05, | |
| "loss": 0.429, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.8028720626631853, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 6.39051006855697e-05, | |
| "loss": 0.4353, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.806135770234987, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 6.381190971999419e-05, | |
| "loss": 0.5023, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.8093994778067886, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 6.371839758138234e-05, | |
| "loss": 0.5856, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.8126631853785901, | |
| "grad_norm": 0.06396484375, | |
| "learning_rate": 6.362456541117837e-05, | |
| "loss": 0.4508, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.8159268929503917, | |
| "grad_norm": 0.05810546875, | |
| "learning_rate": 6.353041435473294e-05, | |
| "loss": 0.406, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8159268929503917, | |
| "eval_loss": 0.44355788826942444, | |
| "eval_runtime": 97.2181, | |
| "eval_samples_per_second": 1.708, | |
| "eval_steps_per_second": 0.216, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8191906005221932, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 6.343594556128913e-05, | |
| "loss": 0.4317, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.8224543080939948, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 6.334116018396848e-05, | |
| "loss": 0.391, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8257180156657964, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 6.324605937975685e-05, | |
| "loss": 0.5945, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.8289817232375979, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 6.31506443094903e-05, | |
| "loss": 0.4871, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.8322454308093995, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 6.305491613784096e-05, | |
| "loss": 0.6679, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.835509138381201, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 6.295887603330279e-05, | |
| "loss": 0.3991, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.8387728459530026, | |
| "grad_norm": 0.08837890625, | |
| "learning_rate": 6.286252516817732e-05, | |
| "loss": 0.3772, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.8420365535248042, | |
| "grad_norm": 0.04150390625, | |
| "learning_rate": 6.276586471855933e-05, | |
| "loss": 0.3838, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.8453002610966057, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 6.266889586432253e-05, | |
| "loss": 0.4486, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.8485639686684073, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 6.257161978910512e-05, | |
| "loss": 0.5229, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8518276762402088, | |
| "grad_norm": 0.061279296875, | |
| "learning_rate": 6.247403768029532e-05, | |
| "loss": 0.3971, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.8550913838120104, | |
| "grad_norm": 0.068359375, | |
| "learning_rate": 6.237615072901696e-05, | |
| "loss": 0.3869, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.858355091383812, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 6.227796013011487e-05, | |
| "loss": 0.4435, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.8616187989556136, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.217946708214033e-05, | |
| "loss": 0.5546, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.8648825065274152, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 6.20806727873364e-05, | |
| "loss": 0.4012, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.8681462140992167, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 6.19815784516233e-05, | |
| "loss": 0.4893, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.8714099216710183, | |
| "grad_norm": 0.07666015625, | |
| "learning_rate": 6.188218528458363e-05, | |
| "loss": 0.4179, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.8746736292428199, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 6.178249449944768e-05, | |
| "loss": 0.4676, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.8779373368146214, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 6.168250731307853e-05, | |
| "loss": 0.3883, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.881201044386423, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 6.158222494595731e-05, | |
| "loss": 0.5093, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.8844647519582245, | |
| "grad_norm": 0.061767578125, | |
| "learning_rate": 6.148164862216815e-05, | |
| "loss": 0.4054, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.8877284595300261, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 6.138077956938341e-05, | |
| "loss": 0.4264, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.8909921671018277, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 6.127961901884854e-05, | |
| "loss": 0.5763, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.8942558746736292, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 6.117816820536717e-05, | |
| "loss": 0.453, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.8975195822454308, | |
| "grad_norm": 0.06103515625, | |
| "learning_rate": 6.107642836728597e-05, | |
| "loss": 0.4166, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.9007832898172323, | |
| "grad_norm": 0.059814453125, | |
| "learning_rate": 6.097440074647955e-05, | |
| "loss": 0.3718, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9040469973890339, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 6.087208658833532e-05, | |
| "loss": 0.4553, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.9073107049608355, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 6.076948714173824e-05, | |
| "loss": 0.4116, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.910574412532637, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 6.066660365905563e-05, | |
| "loss": 0.4693, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.9138381201044387, | |
| "grad_norm": 0.052734375, | |
| "learning_rate": 6.056343739612187e-05, | |
| "loss": 0.4375, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9171018276762402, | |
| "grad_norm": 0.060302734375, | |
| "learning_rate": 6.0459989612223045e-05, | |
| "loss": 0.4129, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.9203655352480418, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 6.035626157008159e-05, | |
| "loss": 0.5467, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.9236292428198434, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 6.025225453584089e-05, | |
| "loss": 0.5297, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.9268929503916449, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 6.0147969779049794e-05, | |
| "loss": 0.3658, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.9301566579634465, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 6.0043408572647165e-05, | |
| "loss": 0.3903, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.933420365535248, | |
| "grad_norm": 0.06005859375, | |
| "learning_rate": 5.9938572192946285e-05, | |
| "loss": 0.4358, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.9366840731070496, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 5.983346191961932e-05, | |
| "loss": 0.5026, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.9399477806788512, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 5.9728079035681674e-05, | |
| "loss": 0.5702, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.9432114882506527, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 5.962242482747635e-05, | |
| "loss": 0.5427, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.9464751958224543, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 5.951650058465822e-05, | |
| "loss": 0.4387, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.9497389033942559, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 5.9410307600178336e-05, | |
| "loss": 0.5693, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.9530026109660574, | |
| "grad_norm": 0.1083984375, | |
| "learning_rate": 5.930384717026806e-05, | |
| "loss": 0.5604, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.956266318537859, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 5.919712059442334e-05, | |
| "loss": 0.4111, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.9595300261096605, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 5.909012917538876e-05, | |
| "loss": 0.5463, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.9627937336814621, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 5.8982874219141704e-05, | |
| "loss": 0.4476, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.9660574412532638, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 5.88753570348764e-05, | |
| "loss": 0.4974, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.9693211488250653, | |
| "grad_norm": 0.06396484375, | |
| "learning_rate": 5.8767578934987916e-05, | |
| "loss": 0.4921, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.9725848563968669, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 5.865954123505613e-05, | |
| "loss": 0.4618, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.9758485639686684, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 5.855124525382973e-05, | |
| "loss": 0.478, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.97911227154047, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 5.8442692313210084e-05, | |
| "loss": 0.4921, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.97911227154047, | |
| "eval_loss": 0.4450150430202484, | |
| "eval_runtime": 97.0546, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9823759791122716, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 5.8333883738235076e-05, | |
| "loss": 0.5148, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.9856396866840731, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 5.822482085706298e-05, | |
| "loss": 0.6192, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.9889033942558747, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 5.8115505000956215e-05, | |
| "loss": 0.4433, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.9921671018276762, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 5.8005937504265115e-05, | |
| "loss": 0.3957, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.9954308093994778, | |
| "grad_norm": 0.04150390625, | |
| "learning_rate": 5.789611970441162e-05, | |
| "loss": 0.3821, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.9986945169712794, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 5.7786052941872984e-05, | |
| "loss": 0.3636, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.001958224543081, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 5.767573856016539e-05, | |
| "loss": 0.458, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 1.0052219321148825, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 5.756517790582752e-05, | |
| "loss": 0.3841, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.0084856396866841, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 5.7454372328404215e-05, | |
| "loss": 0.4227, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.0117493472584855, | |
| "grad_norm": 0.1171875, | |
| "learning_rate": 5.734332318042986e-05, | |
| "loss": 0.4105, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.0150130548302871, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 5.7232031817412024e-05, | |
| "loss": 0.3316, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 1.0182767624020888, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 5.7120499597814814e-05, | |
| "loss": 0.5457, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.0215404699738904, | |
| "grad_norm": 0.059814453125, | |
| "learning_rate": 5.700872788304232e-05, | |
| "loss": 0.4402, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 1.024804177545692, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 5.6896718037422e-05, | |
| "loss": 0.4161, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.0280678851174936, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 5.6784471428188035e-05, | |
| "loss": 0.3828, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.031331592689295, | |
| "grad_norm": 0.06396484375, | |
| "learning_rate": 5.667198942546462e-05, | |
| "loss": 0.3445, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.0345953002610966, | |
| "grad_norm": 0.04296875, | |
| "learning_rate": 5.655927340224927e-05, | |
| "loss": 0.3913, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 1.0378590078328982, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 5.644632473439602e-05, | |
| "loss": 0.3671, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.0411227154046998, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 5.633314480059867e-05, | |
| "loss": 0.3709, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.0443864229765012, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 5.621973498237393e-05, | |
| "loss": 0.3355, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.0476501305483028, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 5.610609666404457e-05, | |
| "loss": 0.3738, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.0509138381201044, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 5.59922312327225e-05, | |
| "loss": 0.3342, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.054177545691906, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 5.587814007829189e-05, | |
| "loss": 0.4449, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 1.0574412532637076, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 5.576382459339214e-05, | |
| "loss": 0.5308, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.0607049608355092, | |
| "grad_norm": 0.048828125, | |
| "learning_rate": 5.5649286173400923e-05, | |
| "loss": 0.3753, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.0639686684073106, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 5.5534526216417155e-05, | |
| "loss": 0.422, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.0672323759791122, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 5.5419546123243896e-05, | |
| "loss": 0.4433, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.0704960835509139, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 5.530434729737127e-05, | |
| "loss": 0.3864, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.0737597911227155, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 5.5188931144959356e-05, | |
| "loss": 0.4461, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 1.0770234986945169, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 5.507329907482098e-05, | |
| "loss": 0.3522, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.0802872062663185, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 5.4957452498404546e-05, | |
| "loss": 0.4492, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 1.08355091383812, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 5.4841392829776824e-05, | |
| "loss": 0.4852, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.0868146214099217, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 5.472512148560563e-05, | |
| "loss": 0.3427, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.0900783289817233, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 5.46086398851426e-05, | |
| "loss": 0.4092, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.093342036553525, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 5.449194945020581e-05, | |
| "loss": 0.4882, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.0966057441253263, | |
| "grad_norm": 0.060302734375, | |
| "learning_rate": 5.437505160516248e-05, | |
| "loss": 0.3359, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.099869451697128, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 5.4257947776911545e-05, | |
| "loss": 0.3242, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 1.1031331592689295, | |
| "grad_norm": 0.0927734375, | |
| "learning_rate": 5.414063939486623e-05, | |
| "loss": 0.3366, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.1063968668407311, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 5.4023127890936645e-05, | |
| "loss": 0.3718, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.1096605744125327, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 5.3905414699512265e-05, | |
| "loss": 0.4508, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.1129242819843341, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 5.378750125744444e-05, | |
| "loss": 0.3937, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 1.1161879895561357, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 5.366938900402887e-05, | |
| "loss": 0.3259, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.1194516971279374, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 5.3551079380988005e-05, | |
| "loss": 0.4968, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 1.122715404699739, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 5.343257383245346e-05, | |
| "loss": 0.4245, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.1259791122715406, | |
| "grad_norm": 0.058349609375, | |
| "learning_rate": 5.3313873804948404e-05, | |
| "loss": 0.4106, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.129242819843342, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 5.319498074736988e-05, | |
| "loss": 0.4104, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.1325065274151436, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 5.3075896110971115e-05, | |
| "loss": 0.3383, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 1.1357702349869452, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 5.2956621349343856e-05, | |
| "loss": 0.4316, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.1390339425587468, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 5.2837157918400574e-05, | |
| "loss": 0.4306, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 1.1422976501305482, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 5.271750727635669e-05, | |
| "loss": 0.5231, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.1422976501305482, | |
| "eval_loss": 0.4393230080604553, | |
| "eval_runtime": 97.053, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.1455613577023498, | |
| "grad_norm": 0.05419921875, | |
| "learning_rate": 5.2597670883712805e-05, | |
| "loss": 0.5465, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.1488250652741514, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 5.247765020323686e-05, | |
| "loss": 0.4918, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.152088772845953, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 5.2357446699946307e-05, | |
| "loss": 0.3916, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 1.1553524804177546, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 5.223706184109015e-05, | |
| "loss": 0.345, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.1586161879895562, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 5.211649709613114e-05, | |
| "loss": 0.4214, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.1618798955613576, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 5.1995753936727766e-05, | |
| "loss": 0.3919, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.1651436031331592, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 5.187483383671628e-05, | |
| "loss": 0.3803, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.1684073107049608, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 5.175373827209279e-05, | |
| "loss": 0.38, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.1716710182767625, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 5.1632468720995166e-05, | |
| "loss": 0.5457, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 1.174934725848564, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 5.1511026663685e-05, | |
| "loss": 0.5518, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.1781984334203655, | |
| "grad_norm": 0.0625, | |
| "learning_rate": 5.1389413582529605e-05, | |
| "loss": 0.3359, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 1.181462140992167, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 5.1267630961983835e-05, | |
| "loss": 0.4177, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.1847258485639687, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 5.114568028857202e-05, | |
| "loss": 0.4588, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.1879895561357703, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 5.102356305086983e-05, | |
| "loss": 0.5688, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.191253263707572, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 5.090128073948603e-05, | |
| "loss": 0.3719, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.1945169712793733, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 5.0778834847044394e-05, | |
| "loss": 0.5584, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.197780678851175, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 5.065622686816537e-05, | |
| "loss": 0.4056, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 1.2010443864229765, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 5.053345829944795e-05, | |
| "loss": 0.42, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.2043080939947781, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 5.041053063945131e-05, | |
| "loss": 0.4956, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.2075718015665797, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 5.0287445388676564e-05, | |
| "loss": 0.4849, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.2108355091383811, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 5.0164204049548426e-05, | |
| "loss": 0.3587, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.2140992167101827, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 5.0040808126396905e-05, | |
| "loss": 0.5099, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.2173629242819843, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 4.9917259125438897e-05, | |
| "loss": 0.526, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 1.220626631853786, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 4.9793558554759854e-05, | |
| "loss": 0.3404, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.2238903394255876, | |
| "grad_norm": 0.05859375, | |
| "learning_rate": 4.966970792429532e-05, | |
| "loss": 0.4093, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.227154046997389, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.954570874581252e-05, | |
| "loss": 0.4384, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.2304177545691906, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 4.942156253289194e-05, | |
| "loss": 0.4096, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 1.2336814621409922, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 4.929727080090879e-05, | |
| "loss": 0.4918, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.2369451697127938, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.917283506701456e-05, | |
| "loss": 0.3785, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 1.2402088772845954, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 4.904825685011848e-05, | |
| "loss": 0.4161, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.2434725848563968, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 4.8923537670868995e-05, | |
| "loss": 0.3862, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.2467362924281984, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 4.879867905163515e-05, | |
| "loss": 0.6231, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 4.8673682516488076e-05, | |
| "loss": 0.3735, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 1.2532637075718016, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 4.854854959118237e-05, | |
| "loss": 0.4026, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.2565274151436032, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 4.842328180313742e-05, | |
| "loss": 0.4664, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.2597911227154048, | |
| "grad_norm": 0.08837890625, | |
| "learning_rate": 4.829788068141881e-05, | |
| "loss": 0.3852, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.2630548302872062, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 4.817234775671969e-05, | |
| "loss": 0.5633, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.2663185378590078, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 4.804668456134198e-05, | |
| "loss": 0.3544, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.2695822454308094, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 4.792089262917777e-05, | |
| "loss": 0.3355, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 1.2728459530026108, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 4.7794973495690564e-05, | |
| "loss": 0.3438, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.2761096605744124, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.7668928697896504e-05, | |
| "loss": 0.4627, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 1.279373368146214, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 4.7542759774345685e-05, | |
| "loss": 0.3699, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.2826370757180157, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 4.741646826510326e-05, | |
| "loss": 0.5245, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.2859007832898173, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 4.729005571173077e-05, | |
| "loss": 0.3119, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.2891644908616189, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 4.716352365726723e-05, | |
| "loss": 0.3892, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.2924281984334205, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 4.703687364621036e-05, | |
| "loss": 0.434, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.2956919060052219, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 4.691010722449768e-05, | |
| "loss": 0.4034, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 1.2989556135770235, | |
| "grad_norm": 0.08056640625, | |
| "learning_rate": 4.678322593948766e-05, | |
| "loss": 0.5026, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.302219321148825, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 4.6656231339940844e-05, | |
| "loss": 0.4592, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.3054830287206267, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 4.6529124976000944e-05, | |
| "loss": 0.3529, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.3054830287206267, | |
| "eval_loss": 0.4323955774307251, | |
| "eval_runtime": 96.9439, | |
| "eval_samples_per_second": 1.712, | |
| "eval_steps_per_second": 0.217, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.308746736292428, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 4.640190839917588e-05, | |
| "loss": 0.4401, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 1.3120104438642297, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 4.627458316231891e-05, | |
| "loss": 0.6135, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.3152741514360313, | |
| "grad_norm": 0.07666015625, | |
| "learning_rate": 4.6147150819609606e-05, | |
| "loss": 0.3826, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 1.318537859007833, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 4.6019612926534947e-05, | |
| "loss": 0.3976, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.3218015665796345, | |
| "grad_norm": 0.051025390625, | |
| "learning_rate": 4.5891971039870255e-05, | |
| "loss": 0.4904, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.3250652741514362, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 4.5764226717660275e-05, | |
| "loss": 0.3828, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.3283289817232375, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 4.56363815192001e-05, | |
| "loss": 0.4767, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 1.3315926892950392, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 4.5508437005016164e-05, | |
| "loss": 0.4446, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.3348563968668408, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 4.5380394736847195e-05, | |
| "loss": 0.3533, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 1.3381201044386424, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 4.525225627762511e-05, | |
| "loss": 0.3049, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.3413838120104438, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.5124023191455995e-05, | |
| "loss": 0.4373, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.3446475195822454, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 4.499569704360099e-05, | |
| "loss": 0.5643, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.347911227154047, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 4.486727940045717e-05, | |
| "loss": 0.4793, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 1.3511749347258486, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 4.473877182953845e-05, | |
| "loss": 0.3729, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.3544386422976502, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 4.461017589945642e-05, | |
| "loss": 0.3655, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.3577023498694518, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 4.4481493179901224e-05, | |
| "loss": 0.4118, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.3609660574412532, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 4.43527252416224e-05, | |
| "loss": 0.4969, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 1.3642297650130548, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.422387365640966e-05, | |
| "loss": 0.4363, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.3674934725848564, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 4.409493999707379e-05, | |
| "loss": 0.4515, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 1.370757180156658, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 4.3965925837427384e-05, | |
| "loss": 0.4461, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.3740208877284594, | |
| "grad_norm": 0.068359375, | |
| "learning_rate": 4.383683275226564e-05, | |
| "loss": 0.3828, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.377284595300261, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 4.370766231734717e-05, | |
| "loss": 0.4346, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.3805483028720626, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 4.357841610937474e-05, | |
| "loss": 0.3767, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 1.3838120104438643, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 4.3449095705976025e-05, | |
| "loss": 0.3613, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.3870757180156659, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 4.331970268568436e-05, | |
| "loss": 0.3936, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.3903394255874675, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 4.319023862791948e-05, | |
| "loss": 0.3828, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.3936031331592689, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 4.306070511296821e-05, | |
| "loss": 0.3694, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 1.3968668407310705, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 4.293110372196519e-05, | |
| "loss": 0.3963, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.400130548302872, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 4.280143603687363e-05, | |
| "loss": 0.4214, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 1.4033942558746737, | |
| "grad_norm": 0.057861328125, | |
| "learning_rate": 4.267170364046589e-05, | |
| "loss": 0.3812, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.406657963446475, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.254190811630426e-05, | |
| "loss": 0.4025, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 1.4099216710182767, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 4.241205104872158e-05, | |
| "loss": 0.511, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.4131853785900783, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 4.228213402280189e-05, | |
| "loss": 0.3358, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 1.41644908616188, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 4.215215862436114e-05, | |
| "loss": 0.3802, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.4197127937336815, | |
| "grad_norm": 0.060546875, | |
| "learning_rate": 4.202212643992779e-05, | |
| "loss": 0.3486, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.4229765013054831, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 4.189203905672342e-05, | |
| "loss": 0.3345, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.4262402088772845, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 4.1761898062643425e-05, | |
| "loss": 0.3261, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 1.4295039164490861, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 4.163170504623757e-05, | |
| "loss": 0.4276, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.4327676240208878, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 4.1501461596690634e-05, | |
| "loss": 0.3816, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 1.4360313315926894, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 4.1371169303803006e-05, | |
| "loss": 0.4142, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.4392950391644908, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 4.124082975797126e-05, | |
| "loss": 0.4228, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 1.4425587467362924, | |
| "grad_norm": 0.04833984375, | |
| "learning_rate": 4.111044455016877e-05, | |
| "loss": 0.4628, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.445822454308094, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 4.098001527192628e-05, | |
| "loss": 0.3949, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 1.4490861618798956, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 4.084954351531246e-05, | |
| "loss": 0.3102, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.4523498694516972, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 4.0719030872914496e-05, | |
| "loss": 0.3904, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.4556135770234988, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 4.0588478937818636e-05, | |
| "loss": 0.3932, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.4588772845953002, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 4.0457889303590765e-05, | |
| "loss": 0.3941, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 1.4621409921671018, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 4.032726356425693e-05, | |
| "loss": 0.5806, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.4654046997389034, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 4.019660331428387e-05, | |
| "loss": 0.327, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 1.468668407310705, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 4.006591014855959e-05, | |
| "loss": 0.3498, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.468668407310705, | |
| "eval_loss": 0.4334380328655243, | |
| "eval_runtime": 97.3635, | |
| "eval_samples_per_second": 1.705, | |
| "eval_steps_per_second": 0.216, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.4719321148825064, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 3.993518566237389e-05, | |
| "loss": 0.3929, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 1.475195822454308, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 3.9804431451398866e-05, | |
| "loss": 0.3313, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.4784595300261096, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 3.967364911166944e-05, | |
| "loss": 0.383, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 1.4817232375979112, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 3.954284023956389e-05, | |
| "loss": 0.4548, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.4849869451697129, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 3.941200643178436e-05, | |
| "loss": 0.3716, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.4882506527415145, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 3.928114928533737e-05, | |
| "loss": 0.335, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.4915143603133159, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 3.9150270397514325e-05, | |
| "loss": 0.3662, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 1.4947780678851175, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 3.9019371365871994e-05, | |
| "loss": 0.4185, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.498041775456919, | |
| "grad_norm": 0.05517578125, | |
| "learning_rate": 3.8888453788213045e-05, | |
| "loss": 0.3562, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 1.5013054830287205, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 3.875751926256653e-05, | |
| "loss": 0.3962, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.504569190600522, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.862656938716834e-05, | |
| "loss": 0.4031, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 1.5078328981723237, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 3.8495605760441784e-05, | |
| "loss": 0.3921, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.5110966057441253, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 3.8364629980977995e-05, | |
| "loss": 0.3896, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 1.514360313315927, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 3.823364364751644e-05, | |
| "loss": 0.3412, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.5176240208877285, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 3.8102648358925415e-05, | |
| "loss": 0.4458, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.5208877284595301, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 3.7971645714182555e-05, | |
| "loss": 0.3774, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.5241514360313317, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 3.784063731235525e-05, | |
| "loss": 0.4596, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 1.5274151436031331, | |
| "grad_norm": 0.060791015625, | |
| "learning_rate": 3.770962475258118e-05, | |
| "loss": 0.2522, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.5306788511749347, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.7578609634048774e-05, | |
| "loss": 0.4601, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 1.5339425587467361, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 3.7447593555977706e-05, | |
| "loss": 0.329, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.5372062663185377, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 3.731657811759933e-05, | |
| "loss": 0.396, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 1.5404699738903394, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 3.718556491813724e-05, | |
| "loss": 0.5214, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.543733681462141, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 3.705455555678766e-05, | |
| "loss": 0.4289, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 1.5469973890339426, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.6923551632699994e-05, | |
| "loss": 0.3971, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.5502610966057442, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 3.6792554744957235e-05, | |
| "loss": 0.4115, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.5535248041775458, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 3.666156649255656e-05, | |
| "loss": 0.3961, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.5567885117493474, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 3.653058847438964e-05, | |
| "loss": 0.3608, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 1.5600522193211488, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 3.639962228922332e-05, | |
| "loss": 0.4572, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.5633159268929504, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 3.626866953567995e-05, | |
| "loss": 0.3655, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 1.566579634464752, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 3.613773181221794e-05, | |
| "loss": 0.3551, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.5698433420365534, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.600681071711225e-05, | |
| "loss": 0.3538, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 1.573107049608355, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 3.587590784843486e-05, | |
| "loss": 0.416, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.5763707571801566, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 3.574502480403527e-05, | |
| "loss": 0.4449, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 1.5796344647519582, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 3.561416318152098e-05, | |
| "loss": 0.5846, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.5828981723237598, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 3.548332457823805e-05, | |
| "loss": 0.3468, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.5861618798955615, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 3.5352510591251506e-05, | |
| "loss": 0.3941, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.589425587467363, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 3.522172281732596e-05, | |
| "loss": 0.3585, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 1.5926892950391645, | |
| "grad_norm": 0.05908203125, | |
| "learning_rate": 3.509096285290598e-05, | |
| "loss": 0.3889, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.595953002610966, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 3.496023229409676e-05, | |
| "loss": 0.3794, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 1.5992167101827677, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 3.482953273664449e-05, | |
| "loss": 0.3316, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.602480417754569, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 3.469886577591702e-05, | |
| "loss": 0.4095, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 1.6057441253263707, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 3.4568233006884225e-05, | |
| "loss": 0.3882, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.6090078328981723, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 3.44376360240987e-05, | |
| "loss": 0.4681, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 1.612271540469974, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 3.4307076421676166e-05, | |
| "loss": 0.327, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.6155352480417755, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 3.417655579327611e-05, | |
| "loss": 0.3109, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.6187989556135771, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.4046075732082225e-05, | |
| "loss": 0.4118, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.6220626631853787, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 3.391563783078311e-05, | |
| "loss": 0.3506, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 1.6253263707571801, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.3785243681552654e-05, | |
| "loss": 0.3268, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.6285900783289817, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 3.365489487603078e-05, | |
| "loss": 0.4438, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 1.6318537859007833, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 3.352459300530387e-05, | |
| "loss": 0.55, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6318537859007833, | |
| "eval_loss": 0.428629994392395, | |
| "eval_runtime": 96.937, | |
| "eval_samples_per_second": 1.712, | |
| "eval_steps_per_second": 0.217, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6351174934725847, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 3.33943396598854e-05, | |
| "loss": 0.4103, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 1.6383812010443863, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 3.326413642969657e-05, | |
| "loss": 0.373, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.641644908616188, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 3.313398490404682e-05, | |
| "loss": 0.4306, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 1.6449086161879896, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 3.300388667161448e-05, | |
| "loss": 0.3307, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.6481723237597912, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 3.287384332042734e-05, | |
| "loss": 0.3758, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.6514360313315928, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 3.274385643784335e-05, | |
| "loss": 0.345, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.6546997389033944, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 3.2613927610531086e-05, | |
| "loss": 0.3448, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 1.6579634464751958, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 3.248405842445059e-05, | |
| "loss": 0.4311, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.6612271540469974, | |
| "grad_norm": 0.061767578125, | |
| "learning_rate": 3.235425046483379e-05, | |
| "loss": 0.3656, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 1.664490861618799, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.2224505316165364e-05, | |
| "loss": 0.3653, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.6677545691906004, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 3.209482456216321e-05, | |
| "loss": 0.3706, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 1.671018276762402, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 3.196520978575927e-05, | |
| "loss": 0.338, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.6742819843342036, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 3.183566256908008e-05, | |
| "loss": 0.3287, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 1.6775456919060052, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 3.170618449342757e-05, | |
| "loss": 0.4767, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.6808093994778068, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 3.157677713925965e-05, | |
| "loss": 0.5161, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.6840731070496084, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.144744208617105e-05, | |
| "loss": 0.4437, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.68733681462141, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 3.131818091287389e-05, | |
| "loss": 0.3929, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 1.6906005221932114, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 3.1188995197178525e-05, | |
| "loss": 0.4241, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.693864229765013, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 3.1059886515974266e-05, | |
| "loss": 0.3692, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 1.6971279373368147, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.093085644521006e-05, | |
| "loss": 0.4276, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.700391644908616, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 3.080190655987537e-05, | |
| "loss": 0.3514, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 1.7036553524804177, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 3.06730384339808e-05, | |
| "loss": 0.4211, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.7069190600522193, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 3.0544253640539064e-05, | |
| "loss": 0.3813, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 1.7101827676240209, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.0415553751545607e-05, | |
| "loss": 0.3778, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.7134464751958225, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 3.0286940337959565e-05, | |
| "loss": 0.656, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.716710182767624, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 3.0158414969684457e-05, | |
| "loss": 0.3079, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.7199738903394257, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 3.002997921554915e-05, | |
| "loss": 0.4795, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 1.723237597911227, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 2.9901634643288596e-05, | |
| "loss": 0.3406, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.7265013054830287, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 2.97733828195248e-05, | |
| "loss": 0.5483, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 1.7297650130548303, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.9645225309747576e-05, | |
| "loss": 0.5863, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.7330287206266317, | |
| "grad_norm": 0.057373046875, | |
| "learning_rate": 2.9517163678295583e-05, | |
| "loss": 0.5168, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 1.7362924281984333, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.9389199488337076e-05, | |
| "loss": 0.4032, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.739556135770235, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 2.9261334301850975e-05, | |
| "loss": 0.4681, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 1.7428198433420365, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 2.9133569679607677e-05, | |
| "loss": 0.3807, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.7460835509138382, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 2.9005907181150066e-05, | |
| "loss": 0.4555, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.7493472584856398, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 2.8878348364774473e-05, | |
| "loss": 0.4289, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.7526109660574414, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 2.8750894787511633e-05, | |
| "loss": 0.392, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 1.7558746736292428, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.862354800510773e-05, | |
| "loss": 0.4388, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.7591383812010444, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 2.849630957200531e-05, | |
| "loss": 0.3528, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 1.762402088772846, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 2.836918104132442e-05, | |
| "loss": 0.4571, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.7656657963446474, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 2.8242163964843556e-05, | |
| "loss": 0.4158, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 1.768929503916449, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 2.811525989298081e-05, | |
| "loss": 0.3783, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.7721932114882506, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 2.798847037477483e-05, | |
| "loss": 0.4677, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 1.7754569190600522, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 2.786179695786606e-05, | |
| "loss": 0.3359, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.7787206266318538, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 2.773524118847769e-05, | |
| "loss": 0.5303, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.7819843342036554, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 2.7608804611396926e-05, | |
| "loss": 0.5395, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.785248041775457, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.7482488769956006e-05, | |
| "loss": 0.431, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 1.7885117493472587, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 2.7356295206013497e-05, | |
| "loss": 0.5289, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.79177545691906, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 2.7230225459935333e-05, | |
| "loss": 0.3535, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 1.7950391644908616, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 2.7104281070576102e-05, | |
| "loss": 0.3265, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.7950391644908616, | |
| "eval_loss": 0.42748335003852844, | |
| "eval_runtime": 97.001, | |
| "eval_samples_per_second": 1.711, | |
| "eval_steps_per_second": 0.216, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.798302872062663, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.6978463575260266e-05, | |
| "loss": 0.3462, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 1.8015665796344646, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 2.6852774509763334e-05, | |
| "loss": 0.4593, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.8048302872062663, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 2.672721540829318e-05, | |
| "loss": 0.529, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 1.8080939947780679, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 2.6601787803471277e-05, | |
| "loss": 0.4503, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.8113577023498695, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 2.6476493226313997e-05, | |
| "loss": 0.3689, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.814621409921671, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.635133320621393e-05, | |
| "loss": 0.4557, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.8178851174934727, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 2.6226309270921233e-05, | |
| "loss": 0.474, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 1.8211488250652743, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 2.610142294652491e-05, | |
| "loss": 0.3612, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.8244125326370757, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 2.5976675757434293e-05, | |
| "loss": 0.4681, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 1.8276762402088773, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 2.5852069226360327e-05, | |
| "loss": 0.7216, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.8309399477806787, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 2.572760487429708e-05, | |
| "loss": 0.4456, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 1.8342036553524803, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 2.560328422050307e-05, | |
| "loss": 0.4557, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.837467362924282, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 2.5479108782482853e-05, | |
| "loss": 0.4459, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 1.8407310704960835, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 2.535508007596835e-05, | |
| "loss": 0.4037, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.8439947780678851, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 2.52311996149005e-05, | |
| "loss": 0.3679, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.8472584856396868, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 2.5107468911410642e-05, | |
| "loss": 0.4333, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.8505221932114884, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 2.498388947580212e-05, | |
| "loss": 0.3374, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 1.85378590078329, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 2.4860462816531897e-05, | |
| "loss": 0.381, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.8570496083550914, | |
| "grad_norm": 0.1123046875, | |
| "learning_rate": 2.473719044019203e-05, | |
| "loss": 0.3058, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 1.860313315926893, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 2.461407385149139e-05, | |
| "loss": 0.3647, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.8635770234986944, | |
| "grad_norm": 0.061279296875, | |
| "learning_rate": 2.4491114553237216e-05, | |
| "loss": 0.3203, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 1.866840731070496, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 2.4368314046316812e-05, | |
| "loss": 0.3773, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.8701044386422976, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 2.4245673829679214e-05, | |
| "loss": 0.3324, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 1.8733681462140992, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 2.4123195400316884e-05, | |
| "loss": 0.3101, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.8766318537859008, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 2.4000880253247453e-05, | |
| "loss": 0.3916, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.8798955613577024, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 2.3878729881495504e-05, | |
| "loss": 0.3611, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.883159268929504, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 2.3756745776074235e-05, | |
| "loss": 0.3401, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 1.8864229765013056, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 2.3634929425967417e-05, | |
| "loss": 0.4169, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.889686684073107, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 2.351328231811106e-05, | |
| "loss": 0.4605, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 1.8929503916449086, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 2.3391805937375414e-05, | |
| "loss": 0.3902, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.89621409921671, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 2.3270501766546694e-05, | |
| "loss": 0.4328, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 1.8994778067885116, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 2.3149371286309133e-05, | |
| "loss": 0.4144, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.9027415143603132, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 2.302841597522678e-05, | |
| "loss": 0.3111, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 1.9060052219321149, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 2.2907637309725493e-05, | |
| "loss": 0.5008, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.9092689295039165, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 2.2787036764074987e-05, | |
| "loss": 0.4183, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.912532637075718, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 2.266661581037072e-05, | |
| "loss": 0.3335, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.9157963446475197, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 2.2546375918516026e-05, | |
| "loss": 0.356, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 1.9190600522193213, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 2.2426318556204095e-05, | |
| "loss": 0.4071, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.9223237597911227, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 2.230644518890013e-05, | |
| "loss": 0.2689, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 1.9255874673629243, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 2.2186757279823377e-05, | |
| "loss": 0.3768, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.9288511749347257, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 2.206725628992935e-05, | |
| "loss": 0.422, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 1.9321148825065273, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 2.1947943677891948e-05, | |
| "loss": 0.4165, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.935378590078329, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 2.1828820900085644e-05, | |
| "loss": 0.4499, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 1.9386422976501305, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 2.170988941056772e-05, | |
| "loss": 0.4426, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.9419060052219321, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 2.1591150661060552e-05, | |
| "loss": 0.3248, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.9451697127937337, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 2.147260610093383e-05, | |
| "loss": 0.3365, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.9484334203655354, | |
| "grad_norm": 0.10986328125, | |
| "learning_rate": 2.1354257177186918e-05, | |
| "loss": 0.4688, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 1.951697127937337, | |
| "grad_norm": 0.1015625, | |
| "learning_rate": 2.1236105334431134e-05, | |
| "loss": 0.3508, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.9549608355091384, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 2.11181520148722e-05, | |
| "loss": 0.3929, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 1.95822454308094, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 2.1000398658292544e-05, | |
| "loss": 0.351, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.95822454308094, | |
| "eval_loss": 0.42418473958969116, | |
| "eval_runtime": 97.1482, | |
| "eval_samples_per_second": 1.709, | |
| "eval_steps_per_second": 0.216, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.9614882506527413, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 2.0882846702033817e-05, | |
| "loss": 0.374, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 1.964751958224543, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 2.0765497580979243e-05, | |
| "loss": 0.3603, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 1.9680156657963446, | |
| "grad_norm": 0.061767578125, | |
| "learning_rate": 2.0648352727536215e-05, | |
| "loss": 0.3853, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 1.9712793733681462, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 2.0531413571618748e-05, | |
| "loss": 0.4701, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 1.9745430809399478, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 2.0414681540630008e-05, | |
| "loss": 0.4403, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 1.9778067885117494, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 2.029815805944496e-05, | |
| "loss": 0.4048, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 1.981070496083551, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 2.0181844550392882e-05, | |
| "loss": 0.3307, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 1.9843342036553526, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 2.0065742433240112e-05, | |
| "loss": 0.4144, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 1.987597911227154, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 1.994985312517259e-05, | |
| "loss": 0.3239, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 1.9908616187989556, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.9834178040778708e-05, | |
| "loss": 0.7305, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.9941253263707572, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 1.9718718592031877e-05, | |
| "loss": 0.3846, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 1.9973890339425586, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 1.9603476188273463e-05, | |
| "loss": 0.3161, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.0006527415143602, | |
| "grad_norm": 0.0390625, | |
| "learning_rate": 1.948845223619543e-05, | |
| "loss": 0.4225, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 2.003916449086162, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 1.9373648139823288e-05, | |
| "loss": 0.3153, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.0071801566579635, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 1.9259065300498868e-05, | |
| "loss": 0.3552, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 2.010443864229765, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 1.9144705116863293e-05, | |
| "loss": 0.3486, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.0137075718015667, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 1.9030568984839837e-05, | |
| "loss": 0.3223, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 2.0169712793733683, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 1.8916658297616905e-05, | |
| "loss": 0.3672, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.02023498694517, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 1.88029744456311e-05, | |
| "loss": 0.3068, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 2.023498694516971, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.8689518816550106e-05, | |
| "loss": 0.3569, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.0267624020887727, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 1.8576292795255912e-05, | |
| "loss": 0.3565, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 2.0300261096605743, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 1.8463297763827757e-05, | |
| "loss": 0.3031, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 2.033289817232376, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 1.8350535101525395e-05, | |
| "loss": 0.438, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 2.0365535248041775, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 1.823800618477214e-05, | |
| "loss": 0.2767, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.039817232375979, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 1.8125712387138168e-05, | |
| "loss": 0.2778, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 2.0430809399477807, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 1.8013655079323657e-05, | |
| "loss": 0.4171, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 2.0463446475195823, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.7901835629142152e-05, | |
| "loss": 0.3327, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 2.049608355091384, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 1.779025540150376e-05, | |
| "loss": 0.3382, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.0528720626631856, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 1.7678915758398612e-05, | |
| "loss": 0.3052, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 2.056135770234987, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 1.7567818058880107e-05, | |
| "loss": 0.4037, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.0593994778067883, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 1.745696365904846e-05, | |
| "loss": 0.3383, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 2.06266318537859, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 1.7346353912034024e-05, | |
| "loss": 0.3054, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.0659268929503916, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 1.723599016798083e-05, | |
| "loss": 0.3456, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 2.069190600522193, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 1.7125873774030163e-05, | |
| "loss": 0.3405, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 2.072454308093995, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.7016006074303974e-05, | |
| "loss": 0.3015, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 2.0757180156657964, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.6906388409888627e-05, | |
| "loss": 0.302, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.078981723237598, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 1.6797022118818447e-05, | |
| "loss": 0.4294, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 2.0822454308093996, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 1.6687908536059365e-05, | |
| "loss": 0.3167, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.085509138381201, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 1.6579048993492723e-05, | |
| "loss": 0.3518, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 2.0887728459530024, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 1.6470444819898878e-05, | |
| "loss": 0.2712, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.092036553524804, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 1.636209734094112e-05, | |
| "loss": 0.3356, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 2.0953002610966056, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 1.6254007879149407e-05, | |
| "loss": 0.3221, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.098563968668407, | |
| "grad_norm": 0.08837890625, | |
| "learning_rate": 1.6146177753904224e-05, | |
| "loss": 0.289, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 2.101827676240209, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 1.6038608281420538e-05, | |
| "loss": 0.2737, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.1050913838120104, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.5931300774731636e-05, | |
| "loss": 0.3421, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 2.108355091383812, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 1.5824256543673217e-05, | |
| "loss": 0.3091, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 2.1116187989556137, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 1.5717476894867264e-05, | |
| "loss": 0.325, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 2.1148825065274153, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 1.5610963131706243e-05, | |
| "loss": 0.4947, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.118146214099217, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 1.5504716554337078e-05, | |
| "loss": 0.3466, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 2.1214099216710185, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.539873845964531e-05, | |
| "loss": 0.3074, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.1214099216710185, | |
| "eval_loss": 0.43338489532470703, | |
| "eval_runtime": 97.0849, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.1246736292428197, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.5293030141239342e-05, | |
| "loss": 0.3912, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 2.1279373368146213, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 1.5187592889434507e-05, | |
| "loss": 0.3207, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 2.131201044386423, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 1.5082427991237471e-05, | |
| "loss": 0.3349, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 2.1344647519582245, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 1.4977536730330383e-05, | |
| "loss": 0.3584, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 2.137728459530026, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 1.4872920387055338e-05, | |
| "loss": 0.4087, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 2.1409921671018277, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 1.4768580238398619e-05, | |
| "loss": 0.322, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 2.1442558746736293, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 1.4664517557975236e-05, | |
| "loss": 0.3243, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 2.147519582245431, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 1.456073361601326e-05, | |
| "loss": 0.3427, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 2.1507832898172325, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 1.4457229679338425e-05, | |
| "loss": 0.3055, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 2.1540469973890337, | |
| "grad_norm": 0.103515625, | |
| "learning_rate": 1.435400701135857e-05, | |
| "loss": 0.3251, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.1573107049608353, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.4251066872048312e-05, | |
| "loss": 0.3683, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 2.160574412532637, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 1.4148410517933568e-05, | |
| "loss": 0.3124, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 2.1638381201044385, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 1.4046039202076314e-05, | |
| "loss": 0.3981, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 2.16710182767624, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 1.3943954174059192e-05, | |
| "loss": 0.3228, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 2.1703655352480418, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 1.384215667997036e-05, | |
| "loss": 0.2762, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 2.1736292428198434, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.3740647962388183e-05, | |
| "loss": 0.3592, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 2.176892950391645, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 1.3639429260366128e-05, | |
| "loss": 0.399, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 2.1801566579634466, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 1.3538501809417646e-05, | |
| "loss": 0.2658, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 2.183420365535248, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 1.3437866841501027e-05, | |
| "loss": 0.3383, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 2.18668407310705, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 1.3337525585004444e-05, | |
| "loss": 0.3072, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.189947780678851, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 1.323747926473088e-05, | |
| "loss": 0.3372, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 2.1932114882506526, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 1.3137729101883261e-05, | |
| "loss": 0.3112, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 2.196475195822454, | |
| "grad_norm": 0.10205078125, | |
| "learning_rate": 1.3038276314049449e-05, | |
| "loss": 0.3035, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 2.199738903394256, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 1.2939122115187478e-05, | |
| "loss": 0.2873, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 2.2030026109660574, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 1.2840267715610693e-05, | |
| "loss": 0.3703, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 2.206266318537859, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 1.2741714321972936e-05, | |
| "loss": 0.3806, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 2.2095300261096606, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 1.2643463137253914e-05, | |
| "loss": 0.2866, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 2.2127937336814623, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 1.2545515360744405e-05, | |
| "loss": 0.3115, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 2.216057441253264, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 1.2447872188031712e-05, | |
| "loss": 0.3881, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 2.2193211488250655, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 1.2350534810985007e-05, | |
| "loss": 0.2636, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.2225848563968666, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 1.2253504417740808e-05, | |
| "loss": 0.2787, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 2.2258485639686683, | |
| "grad_norm": 0.125, | |
| "learning_rate": 1.2156782192688502e-05, | |
| "loss": 0.6392, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 2.22911227154047, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 1.206036931645582e-05, | |
| "loss": 0.2901, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 2.2323759791122715, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 1.1964266965894472e-05, | |
| "loss": 0.3881, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 2.235639686684073, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 1.1868476314065808e-05, | |
| "loss": 0.2948, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 2.2389033942558747, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.1772998530226423e-05, | |
| "loss": 0.3676, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 2.2421671018276763, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 1.1677834779813963e-05, | |
| "loss": 0.4046, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 2.245430809399478, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.158298622443283e-05, | |
| "loss": 0.3048, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 2.2486945169712795, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 1.1488454021840085e-05, | |
| "loss": 0.3067, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 2.251958224543081, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 1.1394239325931214e-05, | |
| "loss": 0.3266, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.2552219321148828, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 1.1300343286726145e-05, | |
| "loss": 0.3957, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 2.258485639686684, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 1.120676705035512e-05, | |
| "loss": 0.3257, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 2.2617493472584855, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 1.1113511759044789e-05, | |
| "loss": 0.373, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 2.265013054830287, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 1.1020578551104183e-05, | |
| "loss": 0.3352, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 2.2682767624020888, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.0927968560910915e-05, | |
| "loss": 0.3619, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 2.2715404699738904, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 1.0835682918897221e-05, | |
| "loss": 0.4713, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 2.274804177545692, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 1.074372275153628e-05, | |
| "loss": 0.4476, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 2.2780678851174936, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 1.0652089181328364e-05, | |
| "loss": 0.3008, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 2.281331592689295, | |
| "grad_norm": 0.10595703125, | |
| "learning_rate": 1.0560783326787182e-05, | |
| "loss": 0.3803, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 2.2845953002610964, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 1.0469806302426252e-05, | |
| "loss": 0.3342, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.2845953002610964, | |
| "eval_loss": 0.4299331605434418, | |
| "eval_runtime": 97.0974, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.287859007832898, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 1.0379159218745229e-05, | |
| "loss": 0.3328, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 2.2911227154046996, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.0288843182216424e-05, | |
| "loss": 0.3187, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 2.294386422976501, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 1.0198859295271224e-05, | |
| "loss": 0.3737, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 2.297650130548303, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 1.0109208656286718e-05, | |
| "loss": 0.4332, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 2.3009138381201044, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 1.0019892359572203e-05, | |
| "loss": 0.2869, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 2.304177545691906, | |
| "grad_norm": 0.11181640625, | |
| "learning_rate": 9.930911495355906e-06, | |
| "loss": 0.3341, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 2.3074412532637076, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 9.842267149771599e-06, | |
| "loss": 0.3378, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 2.3107049608355092, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 9.753960404845429e-06, | |
| "loss": 0.3384, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 2.313968668407311, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 9.66599233848261e-06, | |
| "loss": 0.4086, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 2.3172323759791125, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 9.578364024454378e-06, | |
| "loss": 0.2587, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 2.320496083550914, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 9.491076532384757e-06, | |
| "loss": 0.3813, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 2.3237597911227152, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 9.404130927737637e-06, | |
| "loss": 0.3561, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 2.327023498694517, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 9.317528271803645e-06, | |
| "loss": 0.3228, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 2.3302872062663185, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 9.231269621687287e-06, | |
| "loss": 0.3716, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 2.33355091383812, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 9.145356030294e-06, | |
| "loss": 0.3529, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 2.3368146214099217, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 9.059788546317274e-06, | |
| "loss": 0.2741, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 2.3400783289817233, | |
| "grad_norm": 0.09130859375, | |
| "learning_rate": 8.974568214225928e-06, | |
| "loss": 0.3426, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 2.343342036553525, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 8.889696074251269e-06, | |
| "loss": 0.4396, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 2.3466057441253265, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.805173162374476e-06, | |
| "loss": 0.3047, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 2.349869451697128, | |
| "grad_norm": 0.0947265625, | |
| "learning_rate": 8.721000510313898e-06, | |
| "loss": 0.274, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.3531331592689293, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.637179145512498e-06, | |
| "loss": 0.3896, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 2.356396866840731, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 8.553710091125273e-06, | |
| "loss": 0.4092, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 2.3596605744125325, | |
| "grad_norm": 0.11669921875, | |
| "learning_rate": 8.470594366006814e-06, | |
| "loss": 0.2948, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 2.362924281984334, | |
| "grad_norm": 0.0966796875, | |
| "learning_rate": 8.38783298469881e-06, | |
| "loss": 0.3066, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 2.3661879895561357, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.305426957417738e-06, | |
| "loss": 0.3687, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 2.3694516971279374, | |
| "grad_norm": 0.111328125, | |
| "learning_rate": 8.223377290042442e-06, | |
| "loss": 0.3082, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 2.372715404699739, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 8.141684984101953e-06, | |
| "loss": 0.3597, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 2.3759791122715406, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 8.060351036763164e-06, | |
| "loss": 0.3336, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 2.379242819843342, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 7.97937644081876e-06, | |
| "loss": 0.2912, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 2.382506527415144, | |
| "grad_norm": 0.09521484375, | |
| "learning_rate": 7.898762184675001e-06, | |
| "loss": 0.309, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 2.3857702349869454, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 7.818509252339747e-06, | |
| "loss": 0.3665, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 2.3890339425587466, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 7.738618623410369e-06, | |
| "loss": 0.3755, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 2.392297650130548, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 7.659091273061833e-06, | |
| "loss": 0.3564, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 2.39556135770235, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 7.5799281720348175e-06, | |
| "loss": 0.4174, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 2.3988250652741514, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 7.5011302866238025e-06, | |
| "loss": 0.3259, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 2.402088772845953, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 7.422698578665338e-06, | |
| "loss": 0.265, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 2.4053524804177546, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.344634005526242e-06, | |
| "loss": 0.3676, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 2.4086161879895562, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 7.266937520091989e-06, | |
| "loss": 0.3388, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 2.411879895561358, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 7.189610070754998e-06, | |
| "loss": 0.29, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 2.4151436031331595, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 7.112652601403125e-06, | |
| "loss": 0.484, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.4184073107049606, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 7.036066051408078e-06, | |
| "loss": 0.3144, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 2.4216710182767622, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 6.9598513556140285e-06, | |
| "loss": 0.3376, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 2.424934725848564, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 6.884009444326109e-06, | |
| "loss": 0.3488, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 2.4281984334203655, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 6.808541243299148e-06, | |
| "loss": 0.3758, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 2.431462140992167, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 6.733447673726288e-06, | |
| "loss": 0.3327, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 2.4347258485639687, | |
| "grad_norm": 0.09375, | |
| "learning_rate": 6.658729652227815e-06, | |
| "loss": 0.277, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 2.4379895561357703, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 6.584388090839894e-06, | |
| "loss": 0.2496, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 2.441253263707572, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 6.510423897003526e-06, | |
| "loss": 0.43, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 2.4445169712793735, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 6.436837973553377e-06, | |
| "loss": 0.3731, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 2.447780678851175, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 6.36363121870683e-06, | |
| "loss": 0.343, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.447780678851175, | |
| "eval_loss": 0.43050137162208557, | |
| "eval_runtime": 97.0194, | |
| "eval_samples_per_second": 1.711, | |
| "eval_steps_per_second": 0.216, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.4510443864229767, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 6.29080452605298e-06, | |
| "loss": 0.2704, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 2.454308093994778, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 6.218358784541744e-06, | |
| "loss": 0.3681, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 2.4575718015665795, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 6.146294878473022e-06, | |
| "loss": 0.4178, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 2.460835509138381, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 6.074613687485852e-06, | |
| "loss": 0.4002, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 2.4640992167101827, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 6.003316086547756e-06, | |
| "loss": 0.2876, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 2.4673629242819843, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 5.9324029459439625e-06, | |
| "loss": 0.2893, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 2.470626631853786, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 5.8618751312668715e-06, | |
| "loss": 0.3622, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 2.4738903394255876, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 5.791733503405424e-06, | |
| "loss": 0.2832, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 2.477154046997389, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.721978918534642e-06, | |
| "loss": 0.352, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 2.480417754569191, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 5.652612228105132e-06, | |
| "loss": 0.4816, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.483681462140992, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 5.583634278832744e-06, | |
| "loss": 0.4329, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 2.4869451697127936, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 5.515045912688172e-06, | |
| "loss": 0.3659, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 2.490208877284595, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 5.446847966886755e-06, | |
| "loss": 0.3918, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 2.493472584856397, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 5.379041273878177e-06, | |
| "loss": 0.2514, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 2.4967362924281984, | |
| "grad_norm": 0.0810546875, | |
| "learning_rate": 5.311626661336356e-06, | |
| "loss": 0.3803, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 5.244604952149344e-06, | |
| "loss": 0.4072, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 2.5032637075718016, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 5.177976964409239e-06, | |
| "loss": 0.2915, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 2.506527415143603, | |
| "grad_norm": 0.1025390625, | |
| "learning_rate": 5.11174351140226e-06, | |
| "loss": 0.2754, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 2.509791122715405, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 5.045905401598755e-06, | |
| "loss": 0.3264, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 2.5130548302872064, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 4.980463438643395e-06, | |
| "loss": 0.4017, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 2.516318537859008, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 4.915418421345301e-06, | |
| "loss": 0.3507, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 2.5195822454308097, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 4.850771143668365e-06, | |
| "loss": 0.3118, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 2.522845953002611, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 4.78652239472148e-06, | |
| "loss": 0.3528, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 2.5261096605744124, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 4.7226729587489856e-06, | |
| "loss": 0.244, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 2.529373368146214, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 4.65922361512102e-06, | |
| "loss": 0.2996, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 2.5326370757180157, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 4.5961751383240845e-06, | |
| "loss": 0.3682, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 2.5359007832898173, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 4.53352829795151e-06, | |
| "loss": 0.3229, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 2.539164490861619, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 4.47128385869413e-06, | |
| "loss": 0.3573, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 2.5424281984334205, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 4.409442580330907e-06, | |
| "loss": 0.3391, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 2.5456919060052217, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.34800521771968e-06, | |
| "loss": 0.3845, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.5489556135770233, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.2869725207879304e-06, | |
| "loss": 0.3741, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 2.552219321148825, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 4.226345234523634e-06, | |
| "loss": 0.3384, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 2.5554830287206265, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 4.166124098966196e-06, | |
| "loss": 0.3473, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 2.558746736292428, | |
| "grad_norm": 0.07666015625, | |
| "learning_rate": 4.106309849197362e-06, | |
| "loss": 0.2661, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 2.5620104438642297, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.046903215332312e-06, | |
| "loss": 0.3476, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 2.5652741514360313, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 3.987904922510677e-06, | |
| "loss": 0.4288, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 2.568537859007833, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 3.929315690887754e-06, | |
| "loss": 0.2945, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 2.5718015665796345, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 3.871136235625673e-06, | |
| "loss": 0.5041, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 2.575065274151436, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 3.8133672668846745e-06, | |
| "loss": 0.3282, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 2.5783289817232378, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 3.75600948981446e-06, | |
| "loss": 0.4907, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 2.5815926892950394, | |
| "grad_norm": 0.10546875, | |
| "learning_rate": 3.6990636045455574e-06, | |
| "loss": 0.3415, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 2.584856396866841, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 3.6425303061807965e-06, | |
| "loss": 0.3957, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 2.588120104438642, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 3.5864102847868226e-06, | |
| "loss": 0.3348, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 2.5913838120104438, | |
| "grad_norm": 0.09326171875, | |
| "learning_rate": 3.530704225385654e-06, | |
| "loss": 0.2835, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 2.5946475195822454, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 3.4754128079463427e-06, | |
| "loss": 0.3671, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 2.597911227154047, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 3.4205367073766603e-06, | |
| "loss": 0.452, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 2.6011749347258486, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 3.3660765935148744e-06, | |
| "loss": 0.3029, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 2.60443864229765, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 3.312033131121548e-06, | |
| "loss": 0.2667, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 2.607702349869452, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 3.2584069798714474e-06, | |
| "loss": 0.3078, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 2.6109660574412534, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 3.205198794345494e-06, | |
| "loss": 0.3406, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.6109660574412534, | |
| "eval_loss": 0.4306102395057678, | |
| "eval_runtime": 96.9205, | |
| "eval_samples_per_second": 1.713, | |
| "eval_steps_per_second": 0.217, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.6142297650130546, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 3.152409224022734e-06, | |
| "loss": 0.3848, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 2.617493472584856, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 3.100038913272479e-06, | |
| "loss": 0.49, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 2.620757180156658, | |
| "grad_norm": 0.04833984375, | |
| "learning_rate": 3.04808850134636e-06, | |
| "loss": 0.2684, | |
| "step": 4015 | |
| }, | |
| { | |
| "epoch": 2.6240208877284594, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 2.9965586223705996e-06, | |
| "loss": 0.4062, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 2.627284595300261, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 2.9454499053382095e-06, | |
| "loss": 0.3041, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 2.6305483028720626, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 2.8947629741013724e-06, | |
| "loss": 0.8468, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 2.6338120104438643, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 2.844498447363759e-06, | |
| "loss": 0.4113, | |
| "step": 4035 | |
| }, | |
| { | |
| "epoch": 2.637075718015666, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 2.7946569386730413e-06, | |
| "loss": 0.2856, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 2.6403394255874675, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 2.745239056413358e-06, | |
| "loss": 0.3137, | |
| "step": 4045 | |
| }, | |
| { | |
| "epoch": 2.643603133159269, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 2.6962454037979213e-06, | |
| "loss": 0.325, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 2.6468668407310707, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 2.647676578861607e-06, | |
| "loss": 0.5683, | |
| "step": 4055 | |
| }, | |
| { | |
| "epoch": 2.6501305483028723, | |
| "grad_norm": 0.06787109375, | |
| "learning_rate": 2.5995331744537273e-06, | |
| "loss": 0.2632, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 2.6533942558746735, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 2.551815778230709e-06, | |
| "loss": 0.3466, | |
| "step": 4065 | |
| }, | |
| { | |
| "epoch": 2.656657963446475, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 2.504524972648998e-06, | |
| "loss": 0.2886, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 2.6599216710182767, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 2.457661334957893e-06, | |
| "loss": 0.5142, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 2.6631853785900783, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 2.411225437192521e-06, | |
| "loss": 0.4388, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 2.66644908616188, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 2.3652178461668824e-06, | |
| "loss": 0.3042, | |
| "step": 4085 | |
| }, | |
| { | |
| "epoch": 2.6697127937336815, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 2.3196391234668625e-06, | |
| "loss": 0.2691, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 2.672976501305483, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 2.2744898254434567e-06, | |
| "loss": 0.3605, | |
| "step": 4095 | |
| }, | |
| { | |
| "epoch": 2.6762402088772848, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 2.229770503205919e-06, | |
| "loss": 0.4555, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.679503916449086, | |
| "grad_norm": 0.109375, | |
| "learning_rate": 2.185481702615076e-06, | |
| "loss": 0.332, | |
| "step": 4105 | |
| }, | |
| { | |
| "epoch": 2.6827676240208875, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 2.1416239642766177e-06, | |
| "loss": 0.3176, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 2.686031331592689, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 2.0981978235345586e-06, | |
| "loss": 0.2961, | |
| "step": 4115 | |
| }, | |
| { | |
| "epoch": 2.6892950391644908, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 2.0552038104646453e-06, | |
| "loss": 0.2982, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 2.6925587467362924, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 2.0126424498679294e-06, | |
| "loss": 0.2402, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 2.695822454308094, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 1.970514261264343e-06, | |
| "loss": 0.3185, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 2.6990861618798956, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 1.928819758886342e-06, | |
| "loss": 0.2964, | |
| "step": 4135 | |
| }, | |
| { | |
| "epoch": 2.702349869451697, | |
| "grad_norm": 0.09423828125, | |
| "learning_rate": 1.8875594516726641e-06, | |
| "loss": 0.2859, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 2.705613577023499, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 1.8467338432620854e-06, | |
| "loss": 0.2926, | |
| "step": 4145 | |
| }, | |
| { | |
| "epoch": 2.7088772845953004, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.8063434319872977e-06, | |
| "loss": 0.45, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 2.712140992167102, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 1.7663887108688005e-06, | |
| "loss": 0.3227, | |
| "step": 4155 | |
| }, | |
| { | |
| "epoch": 2.7154046997389036, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 1.7268701676089055e-06, | |
| "loss": 0.3024, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 2.718668407310705, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 1.6877882845857671e-06, | |
| "loss": 0.4541, | |
| "step": 4165 | |
| }, | |
| { | |
| "epoch": 2.7219321148825064, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 1.6491435388475068e-06, | |
| "loss": 0.3162, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 2.725195822454308, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 1.6109364021063776e-06, | |
| "loss": 0.3428, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 2.7284595300261096, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 1.5731673407330252e-06, | |
| "loss": 0.3112, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 2.7317232375979112, | |
| "grad_norm": 0.06005859375, | |
| "learning_rate": 1.5358368157507655e-06, | |
| "loss": 0.2769, | |
| "step": 4185 | |
| }, | |
| { | |
| "epoch": 2.734986945169713, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 1.4989452828299956e-06, | |
| "loss": 0.3968, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 2.7382506527415145, | |
| "grad_norm": 0.08837890625, | |
| "learning_rate": 1.4624931922825863e-06, | |
| "loss": 0.2949, | |
| "step": 4195 | |
| }, | |
| { | |
| "epoch": 2.741514360313316, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.4264809890564374e-06, | |
| "loss": 0.3438, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.7447780678851172, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 1.3909091127299932e-06, | |
| "loss": 0.2876, | |
| "step": 4205 | |
| }, | |
| { | |
| "epoch": 2.748041775456919, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 1.3557779975069267e-06, | |
| "loss": 0.2769, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.7513054830287205, | |
| "grad_norm": 0.08935546875, | |
| "learning_rate": 1.3210880722107895e-06, | |
| "loss": 0.2985, | |
| "step": 4215 | |
| }, | |
| { | |
| "epoch": 2.754569190600522, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 1.2868397602798287e-06, | |
| "loss": 0.4005, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 2.7578328981723237, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 1.253033479761774e-06, | |
| "loss": 0.2723, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 2.7610966057441253, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 1.219669643308771e-06, | |
| "loss": 0.4161, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 2.764360313315927, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 1.186748658172311e-06, | |
| "loss": 0.266, | |
| "step": 4235 | |
| }, | |
| { | |
| "epoch": 2.7676240208877285, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 1.1542709261982835e-06, | |
| "loss": 0.3112, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 2.77088772845953, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 1.1222368438220691e-06, | |
| "loss": 0.3189, | |
| "step": 4245 | |
| }, | |
| { | |
| "epoch": 2.7741514360313317, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 1.0906468020636799e-06, | |
| "loss": 0.3175, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.7741514360313317, | |
| "eval_loss": 0.4307728111743927, | |
| "eval_runtime": 97.0566, | |
| "eval_samples_per_second": 1.71, | |
| "eval_steps_per_second": 0.216, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.7774151436031334, | |
| "grad_norm": 0.0986328125, | |
| "learning_rate": 1.0595011865230221e-06, | |
| "loss": 0.3127, | |
| "step": 4255 | |
| }, | |
| { | |
| "epoch": 2.780678851174935, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 1.0288003773751454e-06, | |
| "loss": 0.3544, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 2.783942558746736, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 9.985447493656513e-07, | |
| "loss": 0.38, | |
| "step": 4265 | |
| }, | |
| { | |
| "epoch": 2.7872062663185377, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 9.687346718060707e-07, | |
| "loss": 0.3401, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 2.7904699738903394, | |
| "grad_norm": 0.125, | |
| "learning_rate": 9.39370508569394e-07, | |
| "loss": 0.345, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 2.793733681462141, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 9.104526180856025e-07, | |
| "loss": 0.3241, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 2.7969973890339426, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.819813533373147e-07, | |
| "loss": 0.3457, | |
| "step": 4285 | |
| }, | |
| { | |
| "epoch": 2.800261096605744, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 8.539570618554676e-07, | |
| "loss": 0.3233, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 2.803524804177546, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 8.263800857150671e-07, | |
| "loss": 0.3968, | |
| "step": 4295 | |
| }, | |
| { | |
| "epoch": 2.8067885117493474, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 7.992507615310201e-07, | |
| "loss": 0.3133, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.8100522193211486, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 7.725694204540417e-07, | |
| "loss": 0.2727, | |
| "step": 4305 | |
| }, | |
| { | |
| "epoch": 2.81331592689295, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 7.463363881665754e-07, | |
| "loss": 0.3509, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 2.816579634464752, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 7.205519848788588e-07, | |
| "loss": 0.3307, | |
| "step": 4315 | |
| }, | |
| { | |
| "epoch": 2.8198433420365534, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 6.952165253249805e-07, | |
| "loss": 0.401, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 2.823107049608355, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 6.703303187590631e-07, | |
| "loss": 0.2843, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 2.8263707571801566, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 6.458936689514735e-07, | |
| "loss": 0.3478, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 2.8296344647519582, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 6.219068741851227e-07, | |
| "loss": 0.3569, | |
| "step": 4335 | |
| }, | |
| { | |
| "epoch": 2.83289817232376, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 5.983702272518264e-07, | |
| "loss": 0.4081, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 2.8361618798955615, | |
| "grad_norm": 0.1064453125, | |
| "learning_rate": 5.752840154487168e-07, | |
| "loss": 0.4128, | |
| "step": 4345 | |
| }, | |
| { | |
| "epoch": 2.839425587467363, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 5.526485205747569e-07, | |
| "loss": 0.2948, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.8426892950391647, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 5.304640189272777e-07, | |
| "loss": 0.3386, | |
| "step": 4355 | |
| }, | |
| { | |
| "epoch": 2.8459530026109663, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 5.087307812986343e-07, | |
| "loss": 0.3151, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 2.849216710182768, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 4.874490729728675e-07, | |
| "loss": 0.3458, | |
| "step": 4365 | |
| }, | |
| { | |
| "epoch": 2.852480417754569, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 4.666191537224931e-07, | |
| "loss": 0.3181, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 2.8557441253263707, | |
| "grad_norm": 0.076171875, | |
| "learning_rate": 4.4624127780530925e-07, | |
| "loss": 0.3734, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 2.8590078328981723, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 4.263156939613238e-07, | |
| "loss": 0.4662, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 2.862271540469974, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 4.068426454096732e-07, | |
| "loss": 0.3408, | |
| "step": 4385 | |
| }, | |
| { | |
| "epoch": 2.8655352480417755, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 3.8782236984569173e-07, | |
| "loss": 0.3939, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 2.868798955613577, | |
| "grad_norm": 0.12890625, | |
| "learning_rate": 3.6925509943800113e-07, | |
| "loss": 0.3397, | |
| "step": 4395 | |
| }, | |
| { | |
| "epoch": 2.8720626631853787, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 3.51141060825659e-07, | |
| "loss": 0.3527, | |
| "step": 4400 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4596, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "total_flos": 8.296824683102208e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |