| { | |
| "best_metric": 0.7111111111111111, | |
| "best_model_checkpoint": "CTMAE-P2-V4-S3/checkpoint-9396", | |
| "epoch": 49.02, | |
| "eval_steps": 500, | |
| "global_step": 13050, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007662835249042146, | |
| "grad_norm": 6.327857494354248, | |
| "learning_rate": 7.662835249042146e-08, | |
| "loss": 0.667, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0015325670498084292, | |
| "grad_norm": 5.990856170654297, | |
| "learning_rate": 1.5325670498084292e-07, | |
| "loss": 0.6385, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0022988505747126436, | |
| "grad_norm": 8.174938201904297, | |
| "learning_rate": 2.2988505747126437e-07, | |
| "loss": 0.6968, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0030651340996168583, | |
| "grad_norm": 7.544012546539307, | |
| "learning_rate": 3.0651340996168583e-07, | |
| "loss": 0.6474, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0038314176245210726, | |
| "grad_norm": 5.425698757171631, | |
| "learning_rate": 3.831417624521073e-07, | |
| "loss": 0.6453, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004597701149425287, | |
| "grad_norm": 8.521080017089844, | |
| "learning_rate": 4.5977011494252875e-07, | |
| "loss": 0.6873, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0053639846743295016, | |
| "grad_norm": 7.814248085021973, | |
| "learning_rate": 5.363984674329502e-07, | |
| "loss": 0.6156, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.006130268199233717, | |
| "grad_norm": 7.305485248565674, | |
| "learning_rate": 6.130268199233717e-07, | |
| "loss": 0.6401, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.006896551724137931, | |
| "grad_norm": 4.393548488616943, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.584, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.007662835249042145, | |
| "grad_norm": 10.718528747558594, | |
| "learning_rate": 7.662835249042146e-07, | |
| "loss": 0.8103, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.00842911877394636, | |
| "grad_norm": 7.18550443649292, | |
| "learning_rate": 8.429118773946361e-07, | |
| "loss": 0.5662, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.009195402298850575, | |
| "grad_norm": 8.828963279724121, | |
| "learning_rate": 9.195402298850575e-07, | |
| "loss": 0.5529, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.00996168582375479, | |
| "grad_norm": 5.637415409088135, | |
| "learning_rate": 9.96168582375479e-07, | |
| "loss": 0.4353, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.010727969348659003, | |
| "grad_norm": 7.122509479522705, | |
| "learning_rate": 1.0727969348659004e-06, | |
| "loss": 0.7448, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.011494252873563218, | |
| "grad_norm": 34.192989349365234, | |
| "learning_rate": 1.1494252873563219e-06, | |
| "loss": 0.5278, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.012260536398467433, | |
| "grad_norm": 56.518585205078125, | |
| "learning_rate": 1.2260536398467433e-06, | |
| "loss": 0.5696, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.013026819923371647, | |
| "grad_norm": 7.17815637588501, | |
| "learning_rate": 1.3026819923371648e-06, | |
| "loss": 0.5793, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.013793103448275862, | |
| "grad_norm": 3.756725549697876, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.5662, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.014559386973180077, | |
| "grad_norm": 2.2923433780670166, | |
| "learning_rate": 1.455938697318008e-06, | |
| "loss": 0.8302, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.01532567049808429, | |
| "grad_norm": 2.681919574737549, | |
| "learning_rate": 1.5325670498084292e-06, | |
| "loss": 0.9074, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.016091954022988506, | |
| "grad_norm": 2.827556848526001, | |
| "learning_rate": 1.6091954022988506e-06, | |
| "loss": 0.981, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.01685823754789272, | |
| "grad_norm": 3.9783685207366943, | |
| "learning_rate": 1.6858237547892723e-06, | |
| "loss": 1.4148, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.017624521072796936, | |
| "grad_norm": 1.1288398504257202, | |
| "learning_rate": 1.7624521072796935e-06, | |
| "loss": 0.0274, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.01839080459770115, | |
| "grad_norm": 0.45937296748161316, | |
| "learning_rate": 1.839080459770115e-06, | |
| "loss": 0.473, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.019157088122605363, | |
| "grad_norm": 0.3246555030345917, | |
| "learning_rate": 1.9157088122605367e-06, | |
| "loss": 1.447, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.01992337164750958, | |
| "grad_norm": 0.3196878135204315, | |
| "learning_rate": 1.992337164750958e-06, | |
| "loss": 1.5461, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.1853597164154053, | |
| "eval_runtime": 32.4115, | |
| "eval_samples_per_second": 1.388, | |
| "eval_steps_per_second": 1.388, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0006896551724138, | |
| "grad_norm": 0.25021031498908997, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 1.0533, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.001455938697318, | |
| "grad_norm": 75.3009262084961, | |
| "learning_rate": 2.145593869731801e-06, | |
| "loss": 1.5309, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0022222222222221, | |
| "grad_norm": 0.2906692326068878, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.077, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0029885057471264, | |
| "grad_norm": 0.5273029804229736, | |
| "learning_rate": 2.2988505747126437e-06, | |
| "loss": 1.0467, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0037547892720307, | |
| "grad_norm": 0.49148106575012207, | |
| "learning_rate": 2.3754789272030654e-06, | |
| "loss": 2.7266, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.004521072796935, | |
| "grad_norm": 0.536920428276062, | |
| "learning_rate": 2.4521072796934867e-06, | |
| "loss": 1.4654, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0052873563218392, | |
| "grad_norm": 0.9628154635429382, | |
| "learning_rate": 2.5287356321839083e-06, | |
| "loss": 1.7753, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0060536398467432, | |
| "grad_norm": 0.4243074357509613, | |
| "learning_rate": 2.6053639846743296e-06, | |
| "loss": 0.8896, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0068199233716475, | |
| "grad_norm": 0.8502910733222961, | |
| "learning_rate": 2.6819923371647512e-06, | |
| "loss": 2.0229, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0075862068965518, | |
| "grad_norm": 67.57994842529297, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 1.5019, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.008352490421456, | |
| "grad_norm": 0.69255530834198, | |
| "learning_rate": 2.835249042145594e-06, | |
| "loss": 1.8585, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.00911877394636, | |
| "grad_norm": 0.20654457807540894, | |
| "learning_rate": 2.911877394636016e-06, | |
| "loss": 0.5447, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0098850574712643, | |
| "grad_norm": 84.71825408935547, | |
| "learning_rate": 2.988505747126437e-06, | |
| "loss": 2.6544, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0106513409961686, | |
| "grad_norm": 0.3103322386741638, | |
| "learning_rate": 3.0651340996168583e-06, | |
| "loss": 0.4704, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0114176245210729, | |
| "grad_norm": 45.16764831542969, | |
| "learning_rate": 3.14176245210728e-06, | |
| "loss": 1.1276, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.012183908045977, | |
| "grad_norm": 0.26659080386161804, | |
| "learning_rate": 3.2183908045977012e-06, | |
| "loss": 1.633, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0129501915708812, | |
| "grad_norm": 0.8224875926971436, | |
| "learning_rate": 3.295019157088123e-06, | |
| "loss": 1.5345, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0137164750957854, | |
| "grad_norm": 0.3690725862979889, | |
| "learning_rate": 3.3716475095785446e-06, | |
| "loss": 1.4896, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0144827586206897, | |
| "grad_norm": 0.25970518589019775, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.0055, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.015249042145594, | |
| "grad_norm": 0.13966263830661774, | |
| "learning_rate": 3.524904214559387e-06, | |
| "loss": 1.0795, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.016015325670498, | |
| "grad_norm": 0.3560970425605774, | |
| "learning_rate": 3.6015325670498087e-06, | |
| "loss": 1.6404, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0167816091954023, | |
| "grad_norm": 0.6458178758621216, | |
| "learning_rate": 3.67816091954023e-06, | |
| "loss": 1.9237, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0175478927203065, | |
| "grad_norm": 0.5411605834960938, | |
| "learning_rate": 3.7547892720306517e-06, | |
| "loss": 1.7498, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0183141762452108, | |
| "grad_norm": 0.294920951128006, | |
| "learning_rate": 3.831417624521073e-06, | |
| "loss": 0.4111, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0190804597701149, | |
| "grad_norm": 0.1411299705505371, | |
| "learning_rate": 3.908045977011495e-06, | |
| "loss": 0.4981, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0198467432950191, | |
| "grad_norm": 0.08414328843355179, | |
| "learning_rate": 3.984674329501916e-06, | |
| "loss": 0.6074, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.651780843734741, | |
| "eval_runtime": 31.2312, | |
| "eval_samples_per_second": 1.441, | |
| "eval_steps_per_second": 1.441, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.0006130268199236, | |
| "grad_norm": 39.5939826965332, | |
| "learning_rate": 4.0613026819923375e-06, | |
| "loss": 1.7655, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.0013793103448276, | |
| "grad_norm": 1.111221194267273, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 2.6473, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.0021455938697317, | |
| "grad_norm": 43.87031173706055, | |
| "learning_rate": 4.214559386973181e-06, | |
| "loss": 0.8798, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.002911877394636, | |
| "grad_norm": 43.0674934387207, | |
| "learning_rate": 4.291187739463602e-06, | |
| "loss": 2.2232, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.00367816091954, | |
| "grad_norm": 0.8063564300537109, | |
| "learning_rate": 4.367816091954023e-06, | |
| "loss": 1.638, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.0044444444444443, | |
| "grad_norm": 0.5503166913986206, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.3265, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.0052107279693487, | |
| "grad_norm": 0.3785612881183624, | |
| "learning_rate": 4.521072796934866e-06, | |
| "loss": 0.9416, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.005977011494253, | |
| "grad_norm": 1.303848147392273, | |
| "learning_rate": 4.5977011494252875e-06, | |
| "loss": 2.3919, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.0067432950191573, | |
| "grad_norm": 0.39508482813835144, | |
| "learning_rate": 4.674329501915709e-06, | |
| "loss": 0.3936, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0075095785440613, | |
| "grad_norm": 0.8463417887687683, | |
| "learning_rate": 4.750957854406131e-06, | |
| "loss": 0.9592, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0082758620689654, | |
| "grad_norm": 0.07845330983400345, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.004, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.00904214559387, | |
| "grad_norm": 57.573429107666016, | |
| "learning_rate": 4.904214559386973e-06, | |
| "loss": 2.2964, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.009808429118774, | |
| "grad_norm": 34.41446304321289, | |
| "learning_rate": 4.980842911877395e-06, | |
| "loss": 1.0389, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0105747126436784, | |
| "grad_norm": 0.3542000353336334, | |
| "learning_rate": 5.057471264367817e-06, | |
| "loss": 0.4739, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.0113409961685824, | |
| "grad_norm": 32.84673309326172, | |
| "learning_rate": 5.134099616858238e-06, | |
| "loss": 1.0305, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.0121072796934865, | |
| "grad_norm": 37.61670684814453, | |
| "learning_rate": 5.210727969348659e-06, | |
| "loss": 1.9648, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.012873563218391, | |
| "grad_norm": 0.3119650185108185, | |
| "learning_rate": 5.287356321839081e-06, | |
| "loss": 0.4674, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.013639846743295, | |
| "grad_norm": 31.36629867553711, | |
| "learning_rate": 5.3639846743295025e-06, | |
| "loss": 1.397, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.014406130268199, | |
| "grad_norm": 0.31152790784835815, | |
| "learning_rate": 5.440613026819924e-06, | |
| "loss": 0.9365, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.0151724137931035, | |
| "grad_norm": 0.3041549623012543, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 1.864, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.0159386973180076, | |
| "grad_norm": 31.53572654724121, | |
| "learning_rate": 5.593869731800766e-06, | |
| "loss": 1.6771, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.016704980842912, | |
| "grad_norm": 0.5023788213729858, | |
| "learning_rate": 5.670498084291188e-06, | |
| "loss": 0.3871, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.017471264367816, | |
| "grad_norm": 31.97800064086914, | |
| "learning_rate": 5.747126436781609e-06, | |
| "loss": 0.5553, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.01823754789272, | |
| "grad_norm": 0.13413414359092712, | |
| "learning_rate": 5.823754789272032e-06, | |
| "loss": 0.5476, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.0190038314176246, | |
| "grad_norm": 0.166935995221138, | |
| "learning_rate": 5.9003831417624525e-06, | |
| "loss": 1.0767, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.0197701149425287, | |
| "grad_norm": 0.30724337697029114, | |
| "learning_rate": 5.977011494252874e-06, | |
| "loss": 1.5766, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 1.984305500984192, | |
| "eval_runtime": 30.3151, | |
| "eval_samples_per_second": 1.484, | |
| "eval_steps_per_second": 1.484, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.000536398467433, | |
| "grad_norm": 40.09532165527344, | |
| "learning_rate": 6.053639846743296e-06, | |
| "loss": 1.2763, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.001302681992337, | |
| "grad_norm": 0.40683451294898987, | |
| "learning_rate": 6.130268199233717e-06, | |
| "loss": 0.4937, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.0020689655172412, | |
| "grad_norm": 0.18957696855068207, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 1.0243, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.0028352490421457, | |
| "grad_norm": 0.1001056656241417, | |
| "learning_rate": 6.28352490421456e-06, | |
| "loss": 1.1016, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0036015325670498, | |
| "grad_norm": 0.3062469959259033, | |
| "learning_rate": 6.360153256704982e-06, | |
| "loss": 1.4319, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.004367816091954, | |
| "grad_norm": 32.24515914916992, | |
| "learning_rate": 6.4367816091954025e-06, | |
| "loss": 1.001, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.0051340996168583, | |
| "grad_norm": 0.4738576114177704, | |
| "learning_rate": 6.513409961685824e-06, | |
| "loss": 1.9497, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.0059003831417623, | |
| "grad_norm": 0.5721397399902344, | |
| "learning_rate": 6.590038314176246e-06, | |
| "loss": 0.7872, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.006666666666667, | |
| "grad_norm": 0.3018346130847931, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.392, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.007432950191571, | |
| "grad_norm": 0.13677266240119934, | |
| "learning_rate": 6.743295019157089e-06, | |
| "loss": 0.0059, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.008199233716475, | |
| "grad_norm": 31.005964279174805, | |
| "learning_rate": 6.81992337164751e-06, | |
| "loss": 1.0504, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0089655172413794, | |
| "grad_norm": 0.2759823501110077, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 1.5524, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.0097318007662834, | |
| "grad_norm": 0.18805429339408875, | |
| "learning_rate": 6.973180076628353e-06, | |
| "loss": 0.888, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.010498084291188, | |
| "grad_norm": 21.635663986206055, | |
| "learning_rate": 7.049808429118774e-06, | |
| "loss": 1.9466, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.011264367816092, | |
| "grad_norm": 0.48827672004699707, | |
| "learning_rate": 7.126436781609196e-06, | |
| "loss": 0.1415, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.012030651340996, | |
| "grad_norm": 36.17082214355469, | |
| "learning_rate": 7.2030651340996175e-06, | |
| "loss": 1.9956, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0127969348659005, | |
| "grad_norm": 0.5784490704536438, | |
| "learning_rate": 7.279693486590039e-06, | |
| "loss": 0.9481, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0135632183908045, | |
| "grad_norm": 29.443023681640625, | |
| "learning_rate": 7.35632183908046e-06, | |
| "loss": 0.9678, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.014329501915709, | |
| "grad_norm": 31.519386291503906, | |
| "learning_rate": 7.4329501915708825e-06, | |
| "loss": 2.4512, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.015095785440613, | |
| "grad_norm": 0.29328370094299316, | |
| "learning_rate": 7.509578544061303e-06, | |
| "loss": 0.4548, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.015862068965517, | |
| "grad_norm": 0.40748167037963867, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 1.3566, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.0166283524904216, | |
| "grad_norm": 0.19874916970729828, | |
| "learning_rate": 7.662835249042147e-06, | |
| "loss": 0.4632, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0173946360153256, | |
| "grad_norm": 26.931385040283203, | |
| "learning_rate": 7.739463601532567e-06, | |
| "loss": 2.3728, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.0181609195402297, | |
| "grad_norm": 2.915680408477783, | |
| "learning_rate": 7.81609195402299e-06, | |
| "loss": 1.6149, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.018927203065134, | |
| "grad_norm": 0.7119380235671997, | |
| "learning_rate": 7.89272030651341e-06, | |
| "loss": 0.8573, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.0196934865900382, | |
| "grad_norm": 0.29016897082328796, | |
| "learning_rate": 7.969348659003832e-06, | |
| "loss": 0.7713, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.2332448959350586, | |
| "eval_runtime": 30.6757, | |
| "eval_samples_per_second": 1.467, | |
| "eval_steps_per_second": 1.467, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.000459770114943, | |
| "grad_norm": 0.12648364901542664, | |
| "learning_rate": 8.045977011494253e-06, | |
| "loss": 1.0154, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.001226053639847, | |
| "grad_norm": 27.59795379638672, | |
| "learning_rate": 8.122605363984675e-06, | |
| "loss": 0.5761, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.001992337164751, | |
| "grad_norm": 26.80912208557129, | |
| "learning_rate": 8.199233716475097e-06, | |
| "loss": 1.0388, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.002758620689655, | |
| "grad_norm": 0.41966408491134644, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 1.8333, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.00352490421456, | |
| "grad_norm": 0.567575991153717, | |
| "learning_rate": 8.35249042145594e-06, | |
| "loss": 0.7781, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.004291187739463, | |
| "grad_norm": 25.789730072021484, | |
| "learning_rate": 8.429118773946362e-06, | |
| "loss": 1.3806, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.005057471264368, | |
| "grad_norm": 27.713552474975586, | |
| "learning_rate": 8.505747126436782e-06, | |
| "loss": 1.8523, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.005823754789272, | |
| "grad_norm": 1.5164697170257568, | |
| "learning_rate": 8.582375478927203e-06, | |
| "loss": 1.2296, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.006590038314176, | |
| "grad_norm": 0.36756592988967896, | |
| "learning_rate": 8.659003831417625e-06, | |
| "loss": 0.9391, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.00735632183908, | |
| "grad_norm": 33.70969009399414, | |
| "learning_rate": 8.735632183908047e-06, | |
| "loss": 1.5494, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.008122605363985, | |
| "grad_norm": 0.26646265387535095, | |
| "learning_rate": 8.812260536398468e-06, | |
| "loss": 0.8775, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.0088888888888885, | |
| "grad_norm": 0.15112105011940002, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.0672, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.009655172413793, | |
| "grad_norm": 0.42132970690727234, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 1.3626, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.0104214559386975, | |
| "grad_norm": 0.9489944577217102, | |
| "learning_rate": 9.042145593869732e-06, | |
| "loss": 1.4519, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.011187739463602, | |
| "grad_norm": 0.2976325750350952, | |
| "learning_rate": 9.118773946360155e-06, | |
| "loss": 1.2865, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.011954022988506, | |
| "grad_norm": 0.2896762490272522, | |
| "learning_rate": 9.195402298850575e-06, | |
| "loss": 0.9145, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.01272030651341, | |
| "grad_norm": 0.23707671463489532, | |
| "learning_rate": 9.272030651340997e-06, | |
| "loss": 0.4653, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.0134865900383145, | |
| "grad_norm": 0.10083122551441193, | |
| "learning_rate": 9.348659003831418e-06, | |
| "loss": 0.5373, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.014252873563218, | |
| "grad_norm": 0.3603350520133972, | |
| "learning_rate": 9.42528735632184e-06, | |
| "loss": 1.9906, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.015019157088123, | |
| "grad_norm": 0.6415284872055054, | |
| "learning_rate": 9.501915708812262e-06, | |
| "loss": 1.5834, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.015785440613027, | |
| "grad_norm": 24.561208724975586, | |
| "learning_rate": 9.578544061302683e-06, | |
| "loss": 1.1263, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.016551724137931, | |
| "grad_norm": 0.147970512509346, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.4481, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.017318007662835, | |
| "grad_norm": 0.11870920658111572, | |
| "learning_rate": 9.731800766283525e-06, | |
| "loss": 0.5278, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.01808429118774, | |
| "grad_norm": 0.07480484992265701, | |
| "learning_rate": 9.808429118773947e-06, | |
| "loss": 1.0992, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.018850574712643, | |
| "grad_norm": 0.10927191376686096, | |
| "learning_rate": 9.885057471264368e-06, | |
| "loss": 1.0798, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.019616858237548, | |
| "grad_norm": 24.948057174682617, | |
| "learning_rate": 9.96168582375479e-06, | |
| "loss": 1.797, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 1.706432819366455, | |
| "eval_runtime": 31.3395, | |
| "eval_samples_per_second": 1.436, | |
| "eval_steps_per_second": 1.436, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 5.000383141762452, | |
| "grad_norm": 0.2994377017021179, | |
| "learning_rate": 9.995742869306088e-06, | |
| "loss": 0.765, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.001149425287356, | |
| "grad_norm": 0.39027076959609985, | |
| "learning_rate": 9.987228607918263e-06, | |
| "loss": 1.3461, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.001915708812261, | |
| "grad_norm": 0.19331149756908417, | |
| "learning_rate": 9.97871434653044e-06, | |
| "loss": 0.4567, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.002681992337164, | |
| "grad_norm": 0.20254382491111755, | |
| "learning_rate": 9.970200085142615e-06, | |
| "loss": 1.2892, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.003448275862069, | |
| "grad_norm": 25.492420196533203, | |
| "learning_rate": 9.96168582375479e-06, | |
| "loss": 2.0841, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.004214559386973, | |
| "grad_norm": 25.167016983032227, | |
| "learning_rate": 9.953171562366965e-06, | |
| "loss": 0.9479, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 5.004980842911878, | |
| "grad_norm": 0.15213768184185028, | |
| "learning_rate": 9.944657300979142e-06, | |
| "loss": 0.018, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 5.005747126436781, | |
| "grad_norm": 0.13735659420490265, | |
| "learning_rate": 9.936143039591317e-06, | |
| "loss": 1.5766, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 5.006513409961686, | |
| "grad_norm": 25.63395118713379, | |
| "learning_rate": 9.927628778203492e-06, | |
| "loss": 2.2741, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 5.00727969348659, | |
| "grad_norm": 0.3610183894634247, | |
| "learning_rate": 9.919114516815667e-06, | |
| "loss": 0.4247, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.008045977011494, | |
| "grad_norm": 27.840686798095703, | |
| "learning_rate": 9.910600255427842e-06, | |
| "loss": 2.0585, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 5.0088122605363985, | |
| "grad_norm": 0.32169675827026367, | |
| "learning_rate": 9.902085994040018e-06, | |
| "loss": 0.4628, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 5.009578544061303, | |
| "grad_norm": 26.238645553588867, | |
| "learning_rate": 9.893571732652193e-06, | |
| "loss": 1.2224, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 5.010344827586207, | |
| "grad_norm": 0.6875100135803223, | |
| "learning_rate": 9.885057471264368e-06, | |
| "loss": 1.665, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 5.011111111111111, | |
| "grad_norm": 0.3718373477458954, | |
| "learning_rate": 9.876543209876543e-06, | |
| "loss": 0.6889, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.011877394636016, | |
| "grad_norm": 0.27881282567977905, | |
| "learning_rate": 9.86802894848872e-06, | |
| "loss": 0.8497, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 5.012643678160919, | |
| "grad_norm": 0.32661184668540955, | |
| "learning_rate": 9.859514687100895e-06, | |
| "loss": 0.8933, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 5.013409961685824, | |
| "grad_norm": 0.5657132863998413, | |
| "learning_rate": 9.85100042571307e-06, | |
| "loss": 1.7081, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.014176245210728, | |
| "grad_norm": 0.5560334920883179, | |
| "learning_rate": 9.842486164325245e-06, | |
| "loss": 0.7738, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 5.014942528735633, | |
| "grad_norm": 0.1214025467634201, | |
| "learning_rate": 9.833971902937422e-06, | |
| "loss": 1.3839, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.015708812260536, | |
| "grad_norm": 0.3592962622642517, | |
| "learning_rate": 9.825457641549597e-06, | |
| "loss": 1.394, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 5.016475095785441, | |
| "grad_norm": 0.24445772171020508, | |
| "learning_rate": 9.816943380161772e-06, | |
| "loss": 0.4118, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 5.017241379310345, | |
| "grad_norm": 25.80752944946289, | |
| "learning_rate": 9.808429118773947e-06, | |
| "loss": 1.6126, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.018007662835249, | |
| "grad_norm": 25.0863094329834, | |
| "learning_rate": 9.799914857386122e-06, | |
| "loss": 1.8805, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.018773946360153, | |
| "grad_norm": 0.32739561796188354, | |
| "learning_rate": 9.791400595998298e-06, | |
| "loss": 0.3766, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.019540229885058, | |
| "grad_norm": 0.2028200775384903, | |
| "learning_rate": 9.782886334610473e-06, | |
| "loss": 0.8914, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 1.8977174758911133, | |
| "eval_runtime": 30.2939, | |
| "eval_samples_per_second": 1.485, | |
| "eval_steps_per_second": 1.485, | |
| "step": 1566 | |
| }, | |
| { | |
| "epoch": 6.000306513409962, | |
| "grad_norm": 26.039766311645508, | |
| "learning_rate": 9.774372073222648e-06, | |
| "loss": 0.8297, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 6.001072796934866, | |
| "grad_norm": 25.48565673828125, | |
| "learning_rate": 9.765857811834825e-06, | |
| "loss": 1.2037, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 6.00183908045977, | |
| "grad_norm": 0.8141511678695679, | |
| "learning_rate": 9.757343550447e-06, | |
| "loss": 0.9044, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.002605363984674, | |
| "grad_norm": 0.06729582697153091, | |
| "learning_rate": 9.748829289059175e-06, | |
| "loss": 0.7943, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.003371647509579, | |
| "grad_norm": 0.3494458496570587, | |
| "learning_rate": 9.74031502767135e-06, | |
| "loss": 2.0909, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 6.0041379310344825, | |
| "grad_norm": 0.1301506906747818, | |
| "learning_rate": 9.731800766283525e-06, | |
| "loss": 0.7102, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 6.004904214559387, | |
| "grad_norm": 32.33718490600586, | |
| "learning_rate": 9.723286504895702e-06, | |
| "loss": 1.8136, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 6.005670498084291, | |
| "grad_norm": 0.21746733784675598, | |
| "learning_rate": 9.714772243507877e-06, | |
| "loss": 0.0189, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 6.006436781609195, | |
| "grad_norm": 38.88836669921875, | |
| "learning_rate": 9.706257982120052e-06, | |
| "loss": 1.0149, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.0072030651340995, | |
| "grad_norm": 30.158119201660156, | |
| "learning_rate": 9.697743720732228e-06, | |
| "loss": 0.9491, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 6.007969348659004, | |
| "grad_norm": 26.556350708007812, | |
| "learning_rate": 9.689229459344403e-06, | |
| "loss": 1.5764, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 6.008735632183908, | |
| "grad_norm": 1.7624911069869995, | |
| "learning_rate": 9.680715197956578e-06, | |
| "loss": 0.7505, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 6.009501915708812, | |
| "grad_norm": 0.19809751212596893, | |
| "learning_rate": 9.672200936568753e-06, | |
| "loss": 0.7088, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 6.010268199233717, | |
| "grad_norm": 39.09135818481445, | |
| "learning_rate": 9.663686675180928e-06, | |
| "loss": 2.0235, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 6.011034482758621, | |
| "grad_norm": 1.1116338968276978, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 1.403, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 6.011800766283525, | |
| "grad_norm": 2.2675068378448486, | |
| "learning_rate": 9.64665815240528e-06, | |
| "loss": 1.209, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 6.012567049808429, | |
| "grad_norm": 0.321755051612854, | |
| "learning_rate": 9.638143891017455e-06, | |
| "loss": 0.0205, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 6.013333333333334, | |
| "grad_norm": 0.08931531012058258, | |
| "learning_rate": 9.62962962962963e-06, | |
| "loss": 1.759, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 6.014099616858237, | |
| "grad_norm": 24.562509536743164, | |
| "learning_rate": 9.621115368241805e-06, | |
| "loss": 1.4305, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 6.014865900383142, | |
| "grad_norm": 0.4359224736690521, | |
| "learning_rate": 9.612601106853982e-06, | |
| "loss": 0.7926, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 6.015632183908046, | |
| "grad_norm": 0.09415857493877411, | |
| "learning_rate": 9.604086845466157e-06, | |
| "loss": 0.3808, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 6.01639846743295, | |
| "grad_norm": 0.038484711199998856, | |
| "learning_rate": 9.595572584078332e-06, | |
| "loss": 0.9305, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 6.017164750957854, | |
| "grad_norm": 26.646089553833008, | |
| "learning_rate": 9.587058322690508e-06, | |
| "loss": 2.4958, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 6.017931034482759, | |
| "grad_norm": 31.78369903564453, | |
| "learning_rate": 9.578544061302683e-06, | |
| "loss": 1.4576, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.018697318007663, | |
| "grad_norm": 0.1341467648744583, | |
| "learning_rate": 9.570029799914858e-06, | |
| "loss": 0.3736, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 6.019463601532567, | |
| "grad_norm": 0.6747694611549377, | |
| "learning_rate": 9.561515538527033e-06, | |
| "loss": 0.7372, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.2072386741638184, | |
| "eval_runtime": 30.1859, | |
| "eval_samples_per_second": 1.491, | |
| "eval_steps_per_second": 1.491, | |
| "step": 1827 | |
| }, | |
| { | |
| "epoch": 7.000229885057471, | |
| "grad_norm": 48.89395523071289, | |
| "learning_rate": 9.553001277139208e-06, | |
| "loss": 1.0962, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 7.000996168582375, | |
| "grad_norm": 28.422090530395508, | |
| "learning_rate": 9.544487015751385e-06, | |
| "loss": 1.7933, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 7.00176245210728, | |
| "grad_norm": 0.5705060362815857, | |
| "learning_rate": 9.53597275436356e-06, | |
| "loss": 0.6668, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.0025287356321835, | |
| "grad_norm": 24.384010314941406, | |
| "learning_rate": 9.527458492975735e-06, | |
| "loss": 1.2067, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 7.003295019157088, | |
| "grad_norm": 1.3086849451065063, | |
| "learning_rate": 9.518944231587912e-06, | |
| "loss": 0.6482, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 7.0040613026819925, | |
| "grad_norm": 0.16909146308898926, | |
| "learning_rate": 9.510429970200085e-06, | |
| "loss": 0.5901, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 7.004827586206897, | |
| "grad_norm": 29.166427612304688, | |
| "learning_rate": 9.501915708812262e-06, | |
| "loss": 1.2253, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 7.0055938697318005, | |
| "grad_norm": 49.40794372558594, | |
| "learning_rate": 9.493401447424437e-06, | |
| "loss": 2.0064, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.006360153256705, | |
| "grad_norm": 0.3431980013847351, | |
| "learning_rate": 9.484887186036612e-06, | |
| "loss": 1.069, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 7.0071264367816095, | |
| "grad_norm": 0.1730840653181076, | |
| "learning_rate": 9.476372924648788e-06, | |
| "loss": 0.7683, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 7.007892720306513, | |
| "grad_norm": 0.10903787612915039, | |
| "learning_rate": 9.467858663260963e-06, | |
| "loss": 0.9174, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 7.008659003831418, | |
| "grad_norm": 0.15809372067451477, | |
| "learning_rate": 9.459344401873138e-06, | |
| "loss": 1.647, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 7.009425287356322, | |
| "grad_norm": 25.555727005004883, | |
| "learning_rate": 9.450830140485315e-06, | |
| "loss": 0.8062, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 7.010191570881226, | |
| "grad_norm": 31.12367820739746, | |
| "learning_rate": 9.442315879097488e-06, | |
| "loss": 1.0894, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 7.01095785440613, | |
| "grad_norm": 24.753366470336914, | |
| "learning_rate": 9.433801617709665e-06, | |
| "loss": 1.2923, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 7.011724137931035, | |
| "grad_norm": 0.2812572121620178, | |
| "learning_rate": 9.42528735632184e-06, | |
| "loss": 0.011, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 7.012490421455938, | |
| "grad_norm": 38.690345764160156, | |
| "learning_rate": 9.416773094934015e-06, | |
| "loss": 1.0393, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 7.013256704980843, | |
| "grad_norm": 0.1567247062921524, | |
| "learning_rate": 9.408258833546192e-06, | |
| "loss": 1.0962, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.014022988505747, | |
| "grad_norm": 0.060428839176893234, | |
| "learning_rate": 9.399744572158365e-06, | |
| "loss": 0.8817, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 7.014789272030652, | |
| "grad_norm": 48.183135986328125, | |
| "learning_rate": 9.391230310770542e-06, | |
| "loss": 1.5745, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 7.015555555555555, | |
| "grad_norm": 0.3339151442050934, | |
| "learning_rate": 9.382716049382717e-06, | |
| "loss": 0.9986, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 7.01632183908046, | |
| "grad_norm": 0.03695123642683029, | |
| "learning_rate": 9.374201787994892e-06, | |
| "loss": 0.4187, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 7.017088122605364, | |
| "grad_norm": 0.07724831998348236, | |
| "learning_rate": 9.365687526607068e-06, | |
| "loss": 1.0018, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 7.017854406130268, | |
| "grad_norm": 0.253776490688324, | |
| "learning_rate": 9.357173265219243e-06, | |
| "loss": 2.3753, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 7.018620689655172, | |
| "grad_norm": 0.06786445528268814, | |
| "learning_rate": 9.348659003831418e-06, | |
| "loss": 1.3683, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 7.019386973180077, | |
| "grad_norm": 29.821596145629883, | |
| "learning_rate": 9.340144742443595e-06, | |
| "loss": 1.0467, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 1.7543911933898926, | |
| "eval_runtime": 29.9358, | |
| "eval_samples_per_second": 1.503, | |
| "eval_steps_per_second": 1.503, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 8.00015325670498, | |
| "grad_norm": 0.40903595089912415, | |
| "learning_rate": 9.331630481055768e-06, | |
| "loss": 1.183, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 8.000919540229885, | |
| "grad_norm": 26.776763916015625, | |
| "learning_rate": 9.323116219667945e-06, | |
| "loss": 0.9906, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.001685823754789, | |
| "grad_norm": 0.12063957750797272, | |
| "learning_rate": 9.31460195828012e-06, | |
| "loss": 1.3088, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 8.002452107279694, | |
| "grad_norm": 0.048885323107242584, | |
| "learning_rate": 9.306087696892295e-06, | |
| "loss": 0.7917, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 8.003218390804598, | |
| "grad_norm": 1.3041679859161377, | |
| "learning_rate": 9.297573435504472e-06, | |
| "loss": 1.9487, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 8.003984674329502, | |
| "grad_norm": 0.06798844784498215, | |
| "learning_rate": 9.289059174116647e-06, | |
| "loss": 0.0157, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 8.004750957854407, | |
| "grad_norm": 0.05616142228245735, | |
| "learning_rate": 9.280544912728822e-06, | |
| "loss": 0.467, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 8.00551724137931, | |
| "grad_norm": 27.259878158569336, | |
| "learning_rate": 9.272030651340997e-06, | |
| "loss": 0.5149, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 8.006283524904214, | |
| "grad_norm": 26.025419235229492, | |
| "learning_rate": 9.263516389953172e-06, | |
| "loss": 3.0408, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 8.00704980842912, | |
| "grad_norm": 3.329866886138916, | |
| "learning_rate": 9.255002128565348e-06, | |
| "loss": 2.0258, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 8.007816091954023, | |
| "grad_norm": 37.89010238647461, | |
| "learning_rate": 9.246487867177523e-06, | |
| "loss": 0.8858, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 8.008582375478927, | |
| "grad_norm": 28.047279357910156, | |
| "learning_rate": 9.237973605789698e-06, | |
| "loss": 0.5723, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.009348659003832, | |
| "grad_norm": 0.5687035918235779, | |
| "learning_rate": 9.229459344401875e-06, | |
| "loss": 1.6233, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 8.010114942528736, | |
| "grad_norm": 0.21891818940639496, | |
| "learning_rate": 9.220945083014048e-06, | |
| "loss": 0.7431, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 8.01088122605364, | |
| "grad_norm": 0.2682950496673584, | |
| "learning_rate": 9.212430821626225e-06, | |
| "loss": 0.4155, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 8.011647509578545, | |
| "grad_norm": 0.09466023743152618, | |
| "learning_rate": 9.2039165602384e-06, | |
| "loss": 0.8691, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 8.012413793103448, | |
| "grad_norm": 1.0565741062164307, | |
| "learning_rate": 9.195402298850575e-06, | |
| "loss": 0.6764, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 8.013180076628352, | |
| "grad_norm": 0.14699453115463257, | |
| "learning_rate": 9.186888037462752e-06, | |
| "loss": 1.6452, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 8.013946360153257, | |
| "grad_norm": 39.59352111816406, | |
| "learning_rate": 9.178373776074927e-06, | |
| "loss": 1.1483, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 8.01471264367816, | |
| "grad_norm": 0.045827772468328476, | |
| "learning_rate": 9.169859514687102e-06, | |
| "loss": 0.529, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 8.015478927203064, | |
| "grad_norm": 27.256160736083984, | |
| "learning_rate": 9.161345253299277e-06, | |
| "loss": 0.6336, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 8.01624521072797, | |
| "grad_norm": 0.0258074589073658, | |
| "learning_rate": 9.152830991911452e-06, | |
| "loss": 1.2224, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 8.017011494252873, | |
| "grad_norm": 0.058820076286792755, | |
| "learning_rate": 9.144316730523628e-06, | |
| "loss": 1.1961, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 8.017777777777777, | |
| "grad_norm": 27.617746353149414, | |
| "learning_rate": 9.135802469135803e-06, | |
| "loss": 1.6595, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 8.018544061302682, | |
| "grad_norm": 0.042329080402851105, | |
| "learning_rate": 9.127288207747978e-06, | |
| "loss": 0.3921, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 8.019310344827586, | |
| "grad_norm": 0.029105035588145256, | |
| "learning_rate": 9.118773946360155e-06, | |
| "loss": 1.2248, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.031482696533203, | |
| "eval_runtime": 30.192, | |
| "eval_samples_per_second": 1.49, | |
| "eval_steps_per_second": 1.49, | |
| "step": 2349 | |
| }, | |
| { | |
| "epoch": 9.00007662835249, | |
| "grad_norm": 0.13440681993961334, | |
| "learning_rate": 9.110259684972328e-06, | |
| "loss": 0.9831, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 9.000842911877395, | |
| "grad_norm": 71.88294982910156, | |
| "learning_rate": 9.101745423584505e-06, | |
| "loss": 0.4825, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 9.001609195402299, | |
| "grad_norm": 37.53052520751953, | |
| "learning_rate": 9.09323116219668e-06, | |
| "loss": 1.185, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 9.002375478927203, | |
| "grad_norm": 38.406646728515625, | |
| "learning_rate": 9.084716900808855e-06, | |
| "loss": 0.4177, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 9.003141762452108, | |
| "grad_norm": 0.0962865874171257, | |
| "learning_rate": 9.076202639421032e-06, | |
| "loss": 1.2027, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 9.003908045977012, | |
| "grad_norm": 25.04035758972168, | |
| "learning_rate": 9.067688378033207e-06, | |
| "loss": 1.8004, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.004674329501915, | |
| "grad_norm": 21.876209259033203, | |
| "learning_rate": 9.059174116645382e-06, | |
| "loss": 1.2508, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 9.00544061302682, | |
| "grad_norm": 2.1615195274353027, | |
| "learning_rate": 9.050659855257558e-06, | |
| "loss": 0.7229, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 9.006206896551724, | |
| "grad_norm": 2.2023909091949463, | |
| "learning_rate": 9.042145593869732e-06, | |
| "loss": 1.6846, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 9.006973180076628, | |
| "grad_norm": 45.072322845458984, | |
| "learning_rate": 9.033631332481908e-06, | |
| "loss": 1.3944, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 9.007739463601533, | |
| "grad_norm": 0.5853039026260376, | |
| "learning_rate": 9.025117071094083e-06, | |
| "loss": 0.7244, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 9.008505747126437, | |
| "grad_norm": 0.07648712396621704, | |
| "learning_rate": 9.016602809706258e-06, | |
| "loss": 0.9744, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 9.00927203065134, | |
| "grad_norm": 0.03981802612543106, | |
| "learning_rate": 9.008088548318435e-06, | |
| "loss": 0.2308, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 9.010038314176246, | |
| "grad_norm": 0.02455969899892807, | |
| "learning_rate": 8.999574286930608e-06, | |
| "loss": 0.0019, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 9.01080459770115, | |
| "grad_norm": 0.01998833380639553, | |
| "learning_rate": 8.991060025542785e-06, | |
| "loss": 0.0022, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 9.011570881226053, | |
| "grad_norm": 0.44685810804367065, | |
| "learning_rate": 8.98254576415496e-06, | |
| "loss": 1.4054, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.012337164750958, | |
| "grad_norm": 0.17958851158618927, | |
| "learning_rate": 8.974031502767135e-06, | |
| "loss": 1.1453, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 9.013103448275862, | |
| "grad_norm": 0.5927056074142456, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 2.4858, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 9.013869731800765, | |
| "grad_norm": 0.41714614629745483, | |
| "learning_rate": 8.957002979991487e-06, | |
| "loss": 0.275, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 9.01463601532567, | |
| "grad_norm": 0.22673273086547852, | |
| "learning_rate": 8.948488718603662e-06, | |
| "loss": 2.9377, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 9.015402298850574, | |
| "grad_norm": 1.3472007513046265, | |
| "learning_rate": 8.939974457215838e-06, | |
| "loss": 1.2388, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 9.01616858237548, | |
| "grad_norm": 50.77263259887695, | |
| "learning_rate": 8.931460195828012e-06, | |
| "loss": 0.6684, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 9.016934865900383, | |
| "grad_norm": 36.43789291381836, | |
| "learning_rate": 8.922945934440188e-06, | |
| "loss": 1.4564, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 9.017701149425287, | |
| "grad_norm": 58.35029220581055, | |
| "learning_rate": 8.914431673052363e-06, | |
| "loss": 0.3585, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 9.018467432950192, | |
| "grad_norm": 0.281345933675766, | |
| "learning_rate": 8.905917411664538e-06, | |
| "loss": 0.6004, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 9.019233716475096, | |
| "grad_norm": 0.21588343381881714, | |
| "learning_rate": 8.897403150276715e-06, | |
| "loss": 2.1211, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 2.7813258171081543, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.7126, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 1.7717280387878418, | |
| "eval_runtime": 31.7486, | |
| "eval_samples_per_second": 1.417, | |
| "eval_steps_per_second": 1.417, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 10.000766283524904, | |
| "grad_norm": 0.27461373805999756, | |
| "learning_rate": 8.880374627501065e-06, | |
| "loss": 0.9198, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 10.001532567049809, | |
| "grad_norm": 0.771606981754303, | |
| "learning_rate": 8.87186036611324e-06, | |
| "loss": 1.7329, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 10.002298850574713, | |
| "grad_norm": 0.1488303244113922, | |
| "learning_rate": 8.863346104725415e-06, | |
| "loss": 0.497, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 10.003065134099616, | |
| "grad_norm": 32.737430572509766, | |
| "learning_rate": 8.854831843337592e-06, | |
| "loss": 2.4259, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 10.003831417624522, | |
| "grad_norm": 0.36537179350852966, | |
| "learning_rate": 8.846317581949767e-06, | |
| "loss": 1.3456, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 10.004597701149425, | |
| "grad_norm": 0.03043508343398571, | |
| "learning_rate": 8.837803320561942e-06, | |
| "loss": 0.0356, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 10.005363984674329, | |
| "grad_norm": 1.8493080139160156, | |
| "learning_rate": 8.829289059174118e-06, | |
| "loss": 1.7255, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 10.006130268199234, | |
| "grad_norm": 0.44268864393234253, | |
| "learning_rate": 8.820774797786292e-06, | |
| "loss": 1.0021, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 10.006896551724138, | |
| "grad_norm": 0.10876309126615524, | |
| "learning_rate": 8.812260536398468e-06, | |
| "loss": 0.0146, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 10.007662835249041, | |
| "grad_norm": 0.1778973639011383, | |
| "learning_rate": 8.803746275010643e-06, | |
| "loss": 2.0153, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 10.008429118773947, | |
| "grad_norm": 1.5086232423782349, | |
| "learning_rate": 8.795232013622818e-06, | |
| "loss": 1.0842, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 10.00919540229885, | |
| "grad_norm": 0.23681464791297913, | |
| "learning_rate": 8.786717752234995e-06, | |
| "loss": 0.005, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 10.009961685823756, | |
| "grad_norm": 0.04919436573982239, | |
| "learning_rate": 8.77820349084717e-06, | |
| "loss": 0.963, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 10.01072796934866, | |
| "grad_norm": 34.107696533203125, | |
| "learning_rate": 8.769689229459345e-06, | |
| "loss": 1.5674, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 10.011494252873563, | |
| "grad_norm": 0.39854350686073303, | |
| "learning_rate": 8.76117496807152e-06, | |
| "loss": 1.2748, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 10.012260536398468, | |
| "grad_norm": 1.1435327529907227, | |
| "learning_rate": 8.752660706683695e-06, | |
| "loss": 0.2826, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 10.013026819923372, | |
| "grad_norm": 0.0729794055223465, | |
| "learning_rate": 8.744146445295872e-06, | |
| "loss": 0.9041, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 10.013793103448275, | |
| "grad_norm": 4.273685455322266, | |
| "learning_rate": 8.735632183908047e-06, | |
| "loss": 0.4638, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 10.01455938697318, | |
| "grad_norm": 39.03486251831055, | |
| "learning_rate": 8.727117922520222e-06, | |
| "loss": 1.7522, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 10.015325670498084, | |
| "grad_norm": 0.6265343427658081, | |
| "learning_rate": 8.718603661132398e-06, | |
| "loss": 1.413, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 10.016091954022988, | |
| "grad_norm": 129.6248321533203, | |
| "learning_rate": 8.710089399744572e-06, | |
| "loss": 1.8888, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 10.016858237547893, | |
| "grad_norm": 29.127796173095703, | |
| "learning_rate": 8.701575138356748e-06, | |
| "loss": 0.9942, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 10.017624521072797, | |
| "grad_norm": 0.07597409188747406, | |
| "learning_rate": 8.693060876968923e-06, | |
| "loss": 0.1063, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 10.0183908045977, | |
| "grad_norm": 0.12554273009300232, | |
| "learning_rate": 8.684546615581098e-06, | |
| "loss": 0.4064, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 10.019157088122606, | |
| "grad_norm": 0.20525963604450226, | |
| "learning_rate": 8.676032354193275e-06, | |
| "loss": 1.0527, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 10.01992337164751, | |
| "grad_norm": 0.09247679263353348, | |
| "learning_rate": 8.66751809280545e-06, | |
| "loss": 1.2486, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.044833183288574, | |
| "eval_runtime": 29.149, | |
| "eval_samples_per_second": 1.544, | |
| "eval_steps_per_second": 1.544, | |
| "step": 2871 | |
| }, | |
| { | |
| "epoch": 11.000689655172414, | |
| "grad_norm": 0.19499462842941284, | |
| "learning_rate": 8.659003831417625e-06, | |
| "loss": 0.8874, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 11.001455938697317, | |
| "grad_norm": 0.6056790947914124, | |
| "learning_rate": 8.650489570029802e-06, | |
| "loss": 1.8815, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 11.002222222222223, | |
| "grad_norm": 1.8409892320632935, | |
| "learning_rate": 8.641975308641975e-06, | |
| "loss": 0.3759, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 11.002988505747126, | |
| "grad_norm": 39.48097229003906, | |
| "learning_rate": 8.633461047254152e-06, | |
| "loss": 1.4098, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 11.00375478927203, | |
| "grad_norm": 0.5572229623794556, | |
| "learning_rate": 8.624946785866327e-06, | |
| "loss": 0.4802, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 11.004521072796935, | |
| "grad_norm": 39.57521057128906, | |
| "learning_rate": 8.616432524478502e-06, | |
| "loss": 1.4711, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 11.005287356321839, | |
| "grad_norm": 42.3668327331543, | |
| "learning_rate": 8.607918263090678e-06, | |
| "loss": 1.1544, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 11.006053639846744, | |
| "grad_norm": 40.42279815673828, | |
| "learning_rate": 8.599404001702853e-06, | |
| "loss": 1.5491, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 11.006819923371648, | |
| "grad_norm": 0.818279504776001, | |
| "learning_rate": 8.590889740315028e-06, | |
| "loss": 0.6879, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 11.007586206896551, | |
| "grad_norm": 0.06000624969601631, | |
| "learning_rate": 8.582375478927203e-06, | |
| "loss": 0.7827, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 11.008352490421457, | |
| "grad_norm": 8.923982620239258, | |
| "learning_rate": 8.573861217539378e-06, | |
| "loss": 0.6354, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 11.00911877394636, | |
| "grad_norm": 33.427696228027344, | |
| "learning_rate": 8.565346956151555e-06, | |
| "loss": 0.7101, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 11.009885057471264, | |
| "grad_norm": 0.1918342560529709, | |
| "learning_rate": 8.55683269476373e-06, | |
| "loss": 0.2631, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.01065134099617, | |
| "grad_norm": 31.761737823486328, | |
| "learning_rate": 8.548318433375905e-06, | |
| "loss": 0.1071, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 11.011417624521073, | |
| "grad_norm": 0.03893289342522621, | |
| "learning_rate": 8.539804171988082e-06, | |
| "loss": 0.0013, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 11.012183908045976, | |
| "grad_norm": 42.312713623046875, | |
| "learning_rate": 8.531289910600255e-06, | |
| "loss": 1.3946, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 11.012950191570882, | |
| "grad_norm": 0.08506738394498825, | |
| "learning_rate": 8.522775649212432e-06, | |
| "loss": 1.6674, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 11.013716475095785, | |
| "grad_norm": 43.45698928833008, | |
| "learning_rate": 8.514261387824607e-06, | |
| "loss": 1.991, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 11.014482758620689, | |
| "grad_norm": 34.260616302490234, | |
| "learning_rate": 8.505747126436782e-06, | |
| "loss": 0.3898, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 11.015249042145594, | |
| "grad_norm": 0.24433740973472595, | |
| "learning_rate": 8.497232865048958e-06, | |
| "loss": 0.8675, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 11.016015325670498, | |
| "grad_norm": 1.174355387687683, | |
| "learning_rate": 8.488718603661133e-06, | |
| "loss": 0.443, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 11.016781609195402, | |
| "grad_norm": 0.01762991212308407, | |
| "learning_rate": 8.480204342273308e-06, | |
| "loss": 0.3616, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 11.017547892720307, | |
| "grad_norm": 0.7885949015617371, | |
| "learning_rate": 8.471690080885483e-06, | |
| "loss": 1.423, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 11.01831417624521, | |
| "grad_norm": 122.68183898925781, | |
| "learning_rate": 8.463175819497658e-06, | |
| "loss": 1.5785, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 11.019080459770114, | |
| "grad_norm": 0.02293579652905464, | |
| "learning_rate": 8.454661558109835e-06, | |
| "loss": 0.0335, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 11.01984674329502, | |
| "grad_norm": 85.93800354003906, | |
| "learning_rate": 8.44614729672201e-06, | |
| "loss": 2.2836, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.198768138885498, | |
| "eval_runtime": 28.4521, | |
| "eval_samples_per_second": 1.582, | |
| "eval_steps_per_second": 1.582, | |
| "step": 3132 | |
| }, | |
| { | |
| "epoch": 12.000613026819924, | |
| "grad_norm": 0.14830510318279266, | |
| "learning_rate": 8.437633035334185e-06, | |
| "loss": 0.4864, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 12.001379310344827, | |
| "grad_norm": 0.057821013033390045, | |
| "learning_rate": 8.429118773946362e-06, | |
| "loss": 0.899, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 12.002145593869733, | |
| "grad_norm": 63.5409049987793, | |
| "learning_rate": 8.420604512558537e-06, | |
| "loss": 1.304, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 12.002911877394636, | |
| "grad_norm": 28.738975524902344, | |
| "learning_rate": 8.412090251170712e-06, | |
| "loss": 0.497, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 12.00367816091954, | |
| "grad_norm": 1.074270248413086, | |
| "learning_rate": 8.403575989782887e-06, | |
| "loss": 0.4471, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 12.004444444444445, | |
| "grad_norm": 0.07934243977069855, | |
| "learning_rate": 8.395061728395062e-06, | |
| "loss": 1.0692, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 12.005210727969349, | |
| "grad_norm": 0.1797323077917099, | |
| "learning_rate": 8.386547467007238e-06, | |
| "loss": 0.4376, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 12.005977011494252, | |
| "grad_norm": 1.5944926738739014, | |
| "learning_rate": 8.378033205619413e-06, | |
| "loss": 1.9216, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 12.006743295019158, | |
| "grad_norm": 0.199304461479187, | |
| "learning_rate": 8.369518944231588e-06, | |
| "loss": 0.6527, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 12.007509578544061, | |
| "grad_norm": 0.09717396646738052, | |
| "learning_rate": 8.361004682843763e-06, | |
| "loss": 0.4853, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 12.008275862068965, | |
| "grad_norm": 36.5285530090332, | |
| "learning_rate": 8.35249042145594e-06, | |
| "loss": 1.499, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 12.00904214559387, | |
| "grad_norm": 54.89915466308594, | |
| "learning_rate": 8.343976160068115e-06, | |
| "loss": 0.9393, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 12.009808429118774, | |
| "grad_norm": 1.9105443954467773, | |
| "learning_rate": 8.33546189868029e-06, | |
| "loss": 1.0704, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 12.010574712643677, | |
| "grad_norm": 6.898054122924805, | |
| "learning_rate": 8.326947637292465e-06, | |
| "loss": 0.6557, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 12.011340996168583, | |
| "grad_norm": 0.5106290578842163, | |
| "learning_rate": 8.318433375904642e-06, | |
| "loss": 0.9871, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 12.012107279693486, | |
| "grad_norm": 0.3555498719215393, | |
| "learning_rate": 8.309919114516817e-06, | |
| "loss": 0.4547, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 12.01287356321839, | |
| "grad_norm": 0.0721430778503418, | |
| "learning_rate": 8.301404853128992e-06, | |
| "loss": 1.0625, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 12.013639846743295, | |
| "grad_norm": 1.5174165964126587, | |
| "learning_rate": 8.292890591741167e-06, | |
| "loss": 0.6646, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 12.014406130268199, | |
| "grad_norm": 44.36272048950195, | |
| "learning_rate": 8.284376330353342e-06, | |
| "loss": 1.4613, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 12.015172413793103, | |
| "grad_norm": 0.030098868533968925, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.4604, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 12.015938697318008, | |
| "grad_norm": 0.04015477001667023, | |
| "learning_rate": 8.267347807577693e-06, | |
| "loss": 0.136, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 12.016704980842912, | |
| "grad_norm": 46.45560836791992, | |
| "learning_rate": 8.258833546189868e-06, | |
| "loss": 1.491, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 12.017471264367815, | |
| "grad_norm": 7.361248970031738, | |
| "learning_rate": 8.250319284802043e-06, | |
| "loss": 0.3483, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 12.01823754789272, | |
| "grad_norm": 0.041824039071798325, | |
| "learning_rate": 8.24180502341422e-06, | |
| "loss": 0.1995, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 12.019003831417624, | |
| "grad_norm": 117.7020034790039, | |
| "learning_rate": 8.233290762026395e-06, | |
| "loss": 0.4412, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 12.01977011494253, | |
| "grad_norm": 201.08111572265625, | |
| "learning_rate": 8.22477650063857e-06, | |
| "loss": 0.8409, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.6257597208023071, | |
| "eval_runtime": 30.0728, | |
| "eval_samples_per_second": 1.496, | |
| "eval_steps_per_second": 1.496, | |
| "step": 3393 | |
| }, | |
| { | |
| "epoch": 13.000536398467434, | |
| "grad_norm": 0.027195075526833534, | |
| "learning_rate": 8.216262239250745e-06, | |
| "loss": 0.5696, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 13.001302681992337, | |
| "grad_norm": 0.015136953443288803, | |
| "learning_rate": 8.207747977862922e-06, | |
| "loss": 0.683, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 13.00206896551724, | |
| "grad_norm": 0.07920116931200027, | |
| "learning_rate": 8.199233716475097e-06, | |
| "loss": 1.2076, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 13.002835249042146, | |
| "grad_norm": 60.538002014160156, | |
| "learning_rate": 8.190719455087272e-06, | |
| "loss": 0.3912, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 13.00360153256705, | |
| "grad_norm": 0.32756534218788147, | |
| "learning_rate": 8.182205193699447e-06, | |
| "loss": 0.989, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 13.004367816091953, | |
| "grad_norm": 0.07383251935243607, | |
| "learning_rate": 8.173690932311623e-06, | |
| "loss": 0.5731, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 13.005134099616859, | |
| "grad_norm": 3.3000388145446777, | |
| "learning_rate": 8.165176670923798e-06, | |
| "loss": 2.2572, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 13.005900383141762, | |
| "grad_norm": 0.02051711641252041, | |
| "learning_rate": 8.156662409535973e-06, | |
| "loss": 0.168, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 13.006666666666666, | |
| "grad_norm": 0.09207461774349213, | |
| "learning_rate": 8.148148148148148e-06, | |
| "loss": 1.4929, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 13.007432950191571, | |
| "grad_norm": 0.10528067499399185, | |
| "learning_rate": 8.139633886760325e-06, | |
| "loss": 0.0797, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 13.008199233716475, | |
| "grad_norm": 0.10485323518514633, | |
| "learning_rate": 8.1311196253725e-06, | |
| "loss": 3.8356, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 13.008965517241379, | |
| "grad_norm": 0.09597320109605789, | |
| "learning_rate": 8.122605363984675e-06, | |
| "loss": 1.5876, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 13.009731800766284, | |
| "grad_norm": 143.17591857910156, | |
| "learning_rate": 8.11409110259685e-06, | |
| "loss": 0.8356, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 13.010498084291187, | |
| "grad_norm": 58.587364196777344, | |
| "learning_rate": 8.105576841209027e-06, | |
| "loss": 0.7239, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 13.011264367816091, | |
| "grad_norm": 0.0739077627658844, | |
| "learning_rate": 8.097062579821202e-06, | |
| "loss": 0.7464, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 13.012030651340996, | |
| "grad_norm": 0.06026057153940201, | |
| "learning_rate": 8.088548318433377e-06, | |
| "loss": 0.9616, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 13.0127969348659, | |
| "grad_norm": 17.447830200195312, | |
| "learning_rate": 8.080034057045552e-06, | |
| "loss": 0.2298, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 13.013563218390805, | |
| "grad_norm": 0.13126251101493835, | |
| "learning_rate": 8.071519795657727e-06, | |
| "loss": 1.3408, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 13.014329501915709, | |
| "grad_norm": 0.07087504863739014, | |
| "learning_rate": 8.063005534269903e-06, | |
| "loss": 1.3494, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 13.015095785440613, | |
| "grad_norm": 2.7543046474456787, | |
| "learning_rate": 8.054491272882078e-06, | |
| "loss": 0.2207, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 13.015862068965518, | |
| "grad_norm": 0.016590647399425507, | |
| "learning_rate": 8.045977011494253e-06, | |
| "loss": 0.8871, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 13.016628352490422, | |
| "grad_norm": 0.03712855651974678, | |
| "learning_rate": 8.037462750106428e-06, | |
| "loss": 0.3747, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 13.017394636015325, | |
| "grad_norm": 0.39427733421325684, | |
| "learning_rate": 8.028948488718605e-06, | |
| "loss": 0.5934, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 13.01816091954023, | |
| "grad_norm": 144.2153778076172, | |
| "learning_rate": 8.02043422733078e-06, | |
| "loss": 0.2252, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 13.018927203065134, | |
| "grad_norm": 87.64982604980469, | |
| "learning_rate": 8.011919965942955e-06, | |
| "loss": 1.4615, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 13.019693486590038, | |
| "grad_norm": 0.23163729906082153, | |
| "learning_rate": 8.00340570455513e-06, | |
| "loss": 0.4642, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 13.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.3451073169708252, | |
| "eval_runtime": 27.8647, | |
| "eval_samples_per_second": 1.615, | |
| "eval_steps_per_second": 1.615, | |
| "step": 3654 | |
| }, | |
| { | |
| "epoch": 14.000459770114942, | |
| "grad_norm": 140.87591552734375, | |
| "learning_rate": 7.994891443167307e-06, | |
| "loss": 1.7746, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 14.001226053639847, | |
| "grad_norm": 9.656526565551758, | |
| "learning_rate": 7.986377181779482e-06, | |
| "loss": 0.388, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 14.00199233716475, | |
| "grad_norm": 0.5438821911811829, | |
| "learning_rate": 7.977862920391657e-06, | |
| "loss": 0.971, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 14.002758620689654, | |
| "grad_norm": 0.02026885375380516, | |
| "learning_rate": 7.969348659003832e-06, | |
| "loss": 0.4048, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 14.00352490421456, | |
| "grad_norm": 12.08399772644043, | |
| "learning_rate": 7.960834397616007e-06, | |
| "loss": 0.2787, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 14.004291187739463, | |
| "grad_norm": 0.06703075021505356, | |
| "learning_rate": 7.952320136228183e-06, | |
| "loss": 1.0453, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 14.005057471264367, | |
| "grad_norm": 0.3174547553062439, | |
| "learning_rate": 7.943805874840358e-06, | |
| "loss": 0.009, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 14.005823754789272, | |
| "grad_norm": 0.014533264562487602, | |
| "learning_rate": 7.935291613452533e-06, | |
| "loss": 1.1884, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 14.006590038314176, | |
| "grad_norm": 0.5189967155456543, | |
| "learning_rate": 7.92677735206471e-06, | |
| "loss": 0.5976, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 14.007356321839081, | |
| "grad_norm": 74.87429809570312, | |
| "learning_rate": 7.918263090676885e-06, | |
| "loss": 0.8599, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 14.008122605363985, | |
| "grad_norm": 0.03736857697367668, | |
| "learning_rate": 7.90974882928906e-06, | |
| "loss": 0.3985, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 14.008888888888889, | |
| "grad_norm": 191.43634033203125, | |
| "learning_rate": 7.901234567901235e-06, | |
| "loss": 0.0901, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 14.009655172413794, | |
| "grad_norm": 212.68418884277344, | |
| "learning_rate": 7.89272030651341e-06, | |
| "loss": 0.6925, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 14.010421455938697, | |
| "grad_norm": 0.6265121698379517, | |
| "learning_rate": 7.884206045125587e-06, | |
| "loss": 0.3075, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 14.011187739463601, | |
| "grad_norm": 22.70719337463379, | |
| "learning_rate": 7.875691783737762e-06, | |
| "loss": 0.5691, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 14.011954022988506, | |
| "grad_norm": 5.610614776611328, | |
| "learning_rate": 7.867177522349937e-06, | |
| "loss": 0.5433, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 14.01272030651341, | |
| "grad_norm": 0.08452703058719635, | |
| "learning_rate": 7.858663260962112e-06, | |
| "loss": 1.5302, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 14.013486590038314, | |
| "grad_norm": 0.02380300499498844, | |
| "learning_rate": 7.850148999574287e-06, | |
| "loss": 1.9518, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 14.014252873563219, | |
| "grad_norm": 2.767400026321411, | |
| "learning_rate": 7.841634738186463e-06, | |
| "loss": 0.5367, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 14.015019157088123, | |
| "grad_norm": 0.17338675260543823, | |
| "learning_rate": 7.833120476798638e-06, | |
| "loss": 0.7597, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 14.015785440613026, | |
| "grad_norm": 0.053753260523080826, | |
| "learning_rate": 7.824606215410813e-06, | |
| "loss": 0.0779, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 14.016551724137932, | |
| "grad_norm": 0.021334677934646606, | |
| "learning_rate": 7.81609195402299e-06, | |
| "loss": 0.0797, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 14.017318007662835, | |
| "grad_norm": 0.02414635941386223, | |
| "learning_rate": 7.807577692635165e-06, | |
| "loss": 1.4757, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 14.018084291187739, | |
| "grad_norm": 0.8290309906005859, | |
| "learning_rate": 7.79906343124734e-06, | |
| "loss": 2.1325, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 14.018850574712644, | |
| "grad_norm": 44.017173767089844, | |
| "learning_rate": 7.790549169859515e-06, | |
| "loss": 1.3666, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 14.019616858237548, | |
| "grad_norm": 0.06696108728647232, | |
| "learning_rate": 7.78203490847169e-06, | |
| "loss": 0.007, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "eval_accuracy": 0.5555555555555556, | |
| "eval_loss": 2.2437803745269775, | |
| "eval_runtime": 27.8147, | |
| "eval_samples_per_second": 1.618, | |
| "eval_steps_per_second": 1.618, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 15.000383141762452, | |
| "grad_norm": 0.04292473942041397, | |
| "learning_rate": 7.773520647083867e-06, | |
| "loss": 1.5763, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 15.001149425287357, | |
| "grad_norm": 0.027751317247748375, | |
| "learning_rate": 7.765006385696042e-06, | |
| "loss": 0.0038, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 15.00191570881226, | |
| "grad_norm": 0.09398941695690155, | |
| "learning_rate": 7.756492124308217e-06, | |
| "loss": 1.0784, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 15.002681992337164, | |
| "grad_norm": 30.546510696411133, | |
| "learning_rate": 7.747977862920393e-06, | |
| "loss": 0.4579, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 15.00344827586207, | |
| "grad_norm": 0.6451389789581299, | |
| "learning_rate": 7.739463601532567e-06, | |
| "loss": 0.8586, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 15.004214559386973, | |
| "grad_norm": 0.5304000377655029, | |
| "learning_rate": 7.730949340144743e-06, | |
| "loss": 1.5245, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 15.004980842911877, | |
| "grad_norm": 1.0518193244934082, | |
| "learning_rate": 7.722435078756918e-06, | |
| "loss": 0.6112, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 15.005747126436782, | |
| "grad_norm": 8.345175743103027, | |
| "learning_rate": 7.713920817369093e-06, | |
| "loss": 0.5924, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 15.006513409961686, | |
| "grad_norm": 0.19540780782699585, | |
| "learning_rate": 7.70540655598127e-06, | |
| "loss": 1.059, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 15.00727969348659, | |
| "grad_norm": 0.9799060821533203, | |
| "learning_rate": 7.696892294593445e-06, | |
| "loss": 0.6304, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 15.008045977011495, | |
| "grad_norm": 0.0840282067656517, | |
| "learning_rate": 7.68837803320562e-06, | |
| "loss": 0.4809, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 15.008812260536398, | |
| "grad_norm": 0.012157630175352097, | |
| "learning_rate": 7.679863771817797e-06, | |
| "loss": 1.4427, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 15.009578544061302, | |
| "grad_norm": 42.577362060546875, | |
| "learning_rate": 7.67134951042997e-06, | |
| "loss": 1.5545, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 15.010344827586207, | |
| "grad_norm": 1.662129282951355, | |
| "learning_rate": 7.662835249042147e-06, | |
| "loss": 1.2053, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 15.011111111111111, | |
| "grad_norm": 0.23481321334838867, | |
| "learning_rate": 7.654320987654322e-06, | |
| "loss": 0.4214, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 15.011877394636015, | |
| "grad_norm": 0.3130255937576294, | |
| "learning_rate": 7.645806726266497e-06, | |
| "loss": 0.3643, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 15.01264367816092, | |
| "grad_norm": 0.6060452461242676, | |
| "learning_rate": 7.637292464878673e-06, | |
| "loss": 1.6084, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 15.013409961685824, | |
| "grad_norm": 0.07070709019899368, | |
| "learning_rate": 7.6287782034908475e-06, | |
| "loss": 0.571, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 15.014176245210727, | |
| "grad_norm": 0.015648258849978447, | |
| "learning_rate": 7.620263942103023e-06, | |
| "loss": 1.3485, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 15.014942528735633, | |
| "grad_norm": 0.038527365773916245, | |
| "learning_rate": 7.611749680715198e-06, | |
| "loss": 0.5748, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 15.015708812260536, | |
| "grad_norm": 0.01983085460960865, | |
| "learning_rate": 7.603235419327374e-06, | |
| "loss": 0.4425, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 15.01647509578544, | |
| "grad_norm": 0.517955482006073, | |
| "learning_rate": 7.59472115793955e-06, | |
| "loss": 1.9229, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 15.017241379310345, | |
| "grad_norm": 1.5724751949310303, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.2957, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 15.018007662835249, | |
| "grad_norm": 0.3141676187515259, | |
| "learning_rate": 7.5776926351639e-06, | |
| "loss": 0.7588, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 15.018773946360152, | |
| "grad_norm": 0.3598669171333313, | |
| "learning_rate": 7.569178373776076e-06, | |
| "loss": 1.3462, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 15.019540229885058, | |
| "grad_norm": 117.66388702392578, | |
| "learning_rate": 7.560664112388251e-06, | |
| "loss": 0.9377, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.1870836019515991, | |
| "eval_runtime": 28.3965, | |
| "eval_samples_per_second": 1.585, | |
| "eval_steps_per_second": 1.585, | |
| "step": 4176 | |
| }, | |
| { | |
| "epoch": 16.00030651340996, | |
| "grad_norm": 54.328304290771484, | |
| "learning_rate": 7.552149851000427e-06, | |
| "loss": 0.4261, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 16.001072796934867, | |
| "grad_norm": 44.30015182495117, | |
| "learning_rate": 7.543635589612601e-06, | |
| "loss": 0.2114, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 16.00183908045977, | |
| "grad_norm": 0.014054322615265846, | |
| "learning_rate": 7.535121328224777e-06, | |
| "loss": 1.953, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 16.002605363984674, | |
| "grad_norm": 0.13047707080841064, | |
| "learning_rate": 7.5266070668369525e-06, | |
| "loss": 0.0269, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 16.003371647509578, | |
| "grad_norm": 0.7880622148513794, | |
| "learning_rate": 7.5180928054491275e-06, | |
| "loss": 0.0439, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 16.00413793103448, | |
| "grad_norm": 0.022288668900728226, | |
| "learning_rate": 7.509578544061303e-06, | |
| "loss": 0.4317, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 16.00490421455939, | |
| "grad_norm": 0.10423079133033752, | |
| "learning_rate": 7.501064282673479e-06, | |
| "loss": 0.005, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 16.005670498084292, | |
| "grad_norm": 0.05330587178468704, | |
| "learning_rate": 7.492550021285654e-06, | |
| "loss": 0.4634, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 16.006436781609196, | |
| "grad_norm": 0.013447528705000877, | |
| "learning_rate": 7.48403575989783e-06, | |
| "loss": 0.1265, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 16.0072030651341, | |
| "grad_norm": 3.4366865158081055, | |
| "learning_rate": 7.475521498510004e-06, | |
| "loss": 1.016, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 16.007969348659003, | |
| "grad_norm": 0.8176273703575134, | |
| "learning_rate": 7.46700723712218e-06, | |
| "loss": 1.0061, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 16.008735632183907, | |
| "grad_norm": 0.04355312138795853, | |
| "learning_rate": 7.458492975734356e-06, | |
| "loss": 0.7972, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 16.009501915708814, | |
| "grad_norm": 43.31974411010742, | |
| "learning_rate": 7.449978714346531e-06, | |
| "loss": 1.8093, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 16.010268199233717, | |
| "grad_norm": 1.6316255331039429, | |
| "learning_rate": 7.441464452958707e-06, | |
| "loss": 1.4982, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 16.01103448275862, | |
| "grad_norm": 0.24711428582668304, | |
| "learning_rate": 7.4329501915708825e-06, | |
| "loss": 0.1721, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 16.011800766283525, | |
| "grad_norm": 0.040599048137664795, | |
| "learning_rate": 7.4244359301830575e-06, | |
| "loss": 0.1528, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 16.01256704980843, | |
| "grad_norm": 0.18860793113708496, | |
| "learning_rate": 7.4159216687952325e-06, | |
| "loss": 0.9181, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 16.013333333333332, | |
| "grad_norm": 195.70016479492188, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 1.3522, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 16.01409961685824, | |
| "grad_norm": 0.02612440474331379, | |
| "learning_rate": 7.398893146019583e-06, | |
| "loss": 0.7378, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 16.014865900383143, | |
| "grad_norm": 0.011697613634169102, | |
| "learning_rate": 7.390378884631759e-06, | |
| "loss": 0.7663, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 16.015632183908046, | |
| "grad_norm": 1.438512921333313, | |
| "learning_rate": 7.381864623243934e-06, | |
| "loss": 1.1566, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 16.01639846743295, | |
| "grad_norm": 170.9693145751953, | |
| "learning_rate": 7.37335036185611e-06, | |
| "loss": 0.8363, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 16.017164750957853, | |
| "grad_norm": 0.030791645869612694, | |
| "learning_rate": 7.364836100468284e-06, | |
| "loss": 0.5086, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 16.017931034482757, | |
| "grad_norm": 49.93655014038086, | |
| "learning_rate": 7.35632183908046e-06, | |
| "loss": 1.8686, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 16.018697318007664, | |
| "grad_norm": 1.1999695301055908, | |
| "learning_rate": 7.347807577692636e-06, | |
| "loss": 0.5041, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 16.019463601532568, | |
| "grad_norm": 0.043441757559776306, | |
| "learning_rate": 7.339293316304811e-06, | |
| "loss": 0.7025, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.8904989957809448, | |
| "eval_runtime": 28.9409, | |
| "eval_samples_per_second": 1.555, | |
| "eval_steps_per_second": 1.555, | |
| "step": 4437 | |
| }, | |
| { | |
| "epoch": 17.000229885057472, | |
| "grad_norm": 0.038414884358644485, | |
| "learning_rate": 7.330779054916987e-06, | |
| "loss": 0.2607, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 17.000996168582375, | |
| "grad_norm": 47.00775909423828, | |
| "learning_rate": 7.3222647935291625e-06, | |
| "loss": 1.9866, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 17.00176245210728, | |
| "grad_norm": 0.062096305191516876, | |
| "learning_rate": 7.3137505321413375e-06, | |
| "loss": 0.9869, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 17.002528735632183, | |
| "grad_norm": 0.2115720957517624, | |
| "learning_rate": 7.305236270753513e-06, | |
| "loss": 0.3168, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 17.00329501915709, | |
| "grad_norm": 0.06847666949033737, | |
| "learning_rate": 7.2967220093656875e-06, | |
| "loss": 0.173, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 17.004061302681993, | |
| "grad_norm": 81.696533203125, | |
| "learning_rate": 7.288207747977863e-06, | |
| "loss": 1.3976, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 17.004827586206897, | |
| "grad_norm": 0.02751876972615719, | |
| "learning_rate": 7.279693486590039e-06, | |
| "loss": 0.7855, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 17.0055938697318, | |
| "grad_norm": 196.77980041503906, | |
| "learning_rate": 7.271179225202214e-06, | |
| "loss": 0.2874, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 17.006360153256704, | |
| "grad_norm": 0.02160465344786644, | |
| "learning_rate": 7.26266496381439e-06, | |
| "loss": 1.3808, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 17.007126436781608, | |
| "grad_norm": 2.6331965923309326, | |
| "learning_rate": 7.254150702426566e-06, | |
| "loss": 0.2896, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 17.007892720306515, | |
| "grad_norm": 59.360233306884766, | |
| "learning_rate": 7.24563644103874e-06, | |
| "loss": 1.1669, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 17.00865900383142, | |
| "grad_norm": 0.4970775544643402, | |
| "learning_rate": 7.237122179650916e-06, | |
| "loss": 0.4606, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 17.009425287356322, | |
| "grad_norm": 39.951873779296875, | |
| "learning_rate": 7.228607918263091e-06, | |
| "loss": 0.6738, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 17.010191570881226, | |
| "grad_norm": 0.010161004029214382, | |
| "learning_rate": 7.220093656875267e-06, | |
| "loss": 0.4168, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 17.01095785440613, | |
| "grad_norm": 0.05762038007378578, | |
| "learning_rate": 7.2115793954874425e-06, | |
| "loss": 0.0201, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 17.011724137931033, | |
| "grad_norm": 0.03370929881930351, | |
| "learning_rate": 7.2030651340996175e-06, | |
| "loss": 0.0009, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 17.01249042145594, | |
| "grad_norm": 0.0068419617600739, | |
| "learning_rate": 7.194550872711793e-06, | |
| "loss": 1.1103, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 17.013256704980844, | |
| "grad_norm": 50.3115234375, | |
| "learning_rate": 7.1860366113239675e-06, | |
| "loss": 1.3907, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 17.014022988505747, | |
| "grad_norm": 82.64256286621094, | |
| "learning_rate": 7.177522349936143e-06, | |
| "loss": 1.1197, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 17.01478927203065, | |
| "grad_norm": 0.03214224800467491, | |
| "learning_rate": 7.169008088548319e-06, | |
| "loss": 0.8854, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 17.015555555555554, | |
| "grad_norm": 0.6222575306892395, | |
| "learning_rate": 7.160493827160494e-06, | |
| "loss": 0.5726, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 17.016321839080458, | |
| "grad_norm": 0.08235086500644684, | |
| "learning_rate": 7.15197956577267e-06, | |
| "loss": 1.1434, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 17.017088122605365, | |
| "grad_norm": 0.31722962856292725, | |
| "learning_rate": 7.143465304384846e-06, | |
| "loss": 0.6517, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 17.01785440613027, | |
| "grad_norm": 63.441280364990234, | |
| "learning_rate": 7.13495104299702e-06, | |
| "loss": 1.0804, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 17.018620689655172, | |
| "grad_norm": 81.03401947021484, | |
| "learning_rate": 7.126436781609196e-06, | |
| "loss": 0.5098, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 17.019386973180076, | |
| "grad_norm": 185.1552734375, | |
| "learning_rate": 7.117922520221371e-06, | |
| "loss": 0.2657, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 17.02, | |
| "eval_accuracy": 0.6222222222222222, | |
| "eval_loss": 2.1760308742523193, | |
| "eval_runtime": 29.1189, | |
| "eval_samples_per_second": 1.545, | |
| "eval_steps_per_second": 1.545, | |
| "step": 4698 | |
| }, | |
| { | |
| "epoch": 18.00015325670498, | |
| "grad_norm": 0.019183212891221046, | |
| "learning_rate": 7.109408258833547e-06, | |
| "loss": 0.121, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 18.000919540229884, | |
| "grad_norm": 0.019621726125478745, | |
| "learning_rate": 7.1008939974457225e-06, | |
| "loss": 0.3266, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 18.00168582375479, | |
| "grad_norm": 0.17553330957889557, | |
| "learning_rate": 7.0923797360578975e-06, | |
| "loss": 0.4775, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 18.002452107279694, | |
| "grad_norm": 0.009523698128759861, | |
| "learning_rate": 7.083865474670073e-06, | |
| "loss": 0.9904, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 18.003218390804598, | |
| "grad_norm": 232.20355224609375, | |
| "learning_rate": 7.075351213282249e-06, | |
| "loss": 2.1995, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 18.0039846743295, | |
| "grad_norm": 0.007395245600491762, | |
| "learning_rate": 7.066836951894423e-06, | |
| "loss": 0.4372, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 18.004750957854405, | |
| "grad_norm": 0.0052565704099833965, | |
| "learning_rate": 7.058322690506599e-06, | |
| "loss": 1.3121, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 18.00551724137931, | |
| "grad_norm": 0.17322950065135956, | |
| "learning_rate": 7.049808429118774e-06, | |
| "loss": 0.8167, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 18.006283524904216, | |
| "grad_norm": 0.01216431800276041, | |
| "learning_rate": 7.04129416773095e-06, | |
| "loss": 0.6295, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 18.00704980842912, | |
| "grad_norm": 0.008243502117693424, | |
| "learning_rate": 7.032779906343126e-06, | |
| "loss": 0.7316, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 18.007816091954023, | |
| "grad_norm": 0.15859942138195038, | |
| "learning_rate": 7.0242656449553e-06, | |
| "loss": 0.5048, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 18.008582375478927, | |
| "grad_norm": 136.9736328125, | |
| "learning_rate": 7.015751383567476e-06, | |
| "loss": 0.0581, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 18.00934865900383, | |
| "grad_norm": 0.049381256103515625, | |
| "learning_rate": 7.007237122179652e-06, | |
| "loss": 0.2104, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 18.010114942528734, | |
| "grad_norm": 0.059466905891895294, | |
| "learning_rate": 6.998722860791827e-06, | |
| "loss": 0.4287, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 18.01088122605364, | |
| "grad_norm": 0.45760491490364075, | |
| "learning_rate": 6.9902085994040025e-06, | |
| "loss": 0.5963, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 18.011647509578545, | |
| "grad_norm": 0.27239811420440674, | |
| "learning_rate": 6.9816943380161775e-06, | |
| "loss": 0.8202, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 18.01241379310345, | |
| "grad_norm": 0.004633226431906223, | |
| "learning_rate": 6.973180076628353e-06, | |
| "loss": 0.1138, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 18.013180076628352, | |
| "grad_norm": 0.015148923732340336, | |
| "learning_rate": 6.964665815240529e-06, | |
| "loss": 0.9596, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 18.013946360153255, | |
| "grad_norm": 0.00948005635291338, | |
| "learning_rate": 6.956151553852703e-06, | |
| "loss": 1.2887, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 18.014712643678163, | |
| "grad_norm": 2.433624744415283, | |
| "learning_rate": 6.947637292464879e-06, | |
| "loss": 0.6749, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 18.015478927203066, | |
| "grad_norm": 0.01259034313261509, | |
| "learning_rate": 6.939123031077054e-06, | |
| "loss": 0.4976, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 18.01624521072797, | |
| "grad_norm": 0.015765152871608734, | |
| "learning_rate": 6.93060876968923e-06, | |
| "loss": 0.0015, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 18.017011494252873, | |
| "grad_norm": 0.9589321613311768, | |
| "learning_rate": 6.922094508301406e-06, | |
| "loss": 1.4013, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 18.017777777777777, | |
| "grad_norm": 313.9700622558594, | |
| "learning_rate": 6.913580246913581e-06, | |
| "loss": 0.2812, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 18.01854406130268, | |
| "grad_norm": 0.07902565598487854, | |
| "learning_rate": 6.905065985525757e-06, | |
| "loss": 0.1061, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 18.019310344827588, | |
| "grad_norm": 0.023706253618001938, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 1.3937, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "eval_accuracy": 0.6, | |
| "eval_loss": 2.062213897705078, | |
| "eval_runtime": 28.5715, | |
| "eval_samples_per_second": 1.575, | |
| "eval_steps_per_second": 1.575, | |
| "step": 4959 | |
| }, | |
| { | |
| "epoch": 19.000076628352492, | |
| "grad_norm": 0.23811280727386475, | |
| "learning_rate": 6.888037462750107e-06, | |
| "loss": 0.0007, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 19.000842911877395, | |
| "grad_norm": 0.280170738697052, | |
| "learning_rate": 6.8795232013622825e-06, | |
| "loss": 0.6486, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 19.0016091954023, | |
| "grad_norm": 0.01795949414372444, | |
| "learning_rate": 6.8710089399744575e-06, | |
| "loss": 0.1405, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 19.002375478927203, | |
| "grad_norm": 0.011958453804254532, | |
| "learning_rate": 6.862494678586633e-06, | |
| "loss": 0.5572, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 19.003141762452106, | |
| "grad_norm": 0.0037818371783941984, | |
| "learning_rate": 6.853980417198809e-06, | |
| "loss": 0.8434, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 19.00390804597701, | |
| "grad_norm": 0.05979347229003906, | |
| "learning_rate": 6.845466155810983e-06, | |
| "loss": 1.2717, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 19.004674329501917, | |
| "grad_norm": 0.11836342513561249, | |
| "learning_rate": 6.836951894423159e-06, | |
| "loss": 0.0318, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 19.00544061302682, | |
| "grad_norm": 0.6222385764122009, | |
| "learning_rate": 6.828437633035335e-06, | |
| "loss": 0.0619, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 19.006206896551724, | |
| "grad_norm": 268.2494201660156, | |
| "learning_rate": 6.81992337164751e-06, | |
| "loss": 1.2077, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 19.006973180076628, | |
| "grad_norm": 168.15304565429688, | |
| "learning_rate": 6.811409110259686e-06, | |
| "loss": 1.2017, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 19.00773946360153, | |
| "grad_norm": 0.0629284679889679, | |
| "learning_rate": 6.802894848871861e-06, | |
| "loss": 0.1261, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 19.00850574712644, | |
| "grad_norm": 0.025880135595798492, | |
| "learning_rate": 6.794380587484037e-06, | |
| "loss": 0.5958, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 19.009272030651342, | |
| "grad_norm": 0.004736622795462608, | |
| "learning_rate": 6.7858663260962125e-06, | |
| "loss": 1.5298, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 19.010038314176246, | |
| "grad_norm": 5.028242111206055, | |
| "learning_rate": 6.777352064708387e-06, | |
| "loss": 0.6431, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 19.01080459770115, | |
| "grad_norm": 5.767208099365234, | |
| "learning_rate": 6.7688378033205625e-06, | |
| "loss": 1.0098, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 19.011570881226053, | |
| "grad_norm": 0.006427829619497061, | |
| "learning_rate": 6.760323541932738e-06, | |
| "loss": 0.0017, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 19.012337164750956, | |
| "grad_norm": 0.012779129669070244, | |
| "learning_rate": 6.751809280544913e-06, | |
| "loss": 0.0961, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 19.013103448275864, | |
| "grad_norm": 0.3548179268836975, | |
| "learning_rate": 6.743295019157089e-06, | |
| "loss": 0.7176, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 19.013869731800767, | |
| "grad_norm": 5.676109790802002, | |
| "learning_rate": 6.734780757769263e-06, | |
| "loss": 0.7148, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 19.01463601532567, | |
| "grad_norm": 56.71847152709961, | |
| "learning_rate": 6.726266496381439e-06, | |
| "loss": 0.5807, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 19.015402298850574, | |
| "grad_norm": 0.05567074194550514, | |
| "learning_rate": 6.717752234993615e-06, | |
| "loss": 0.2452, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 19.016168582375478, | |
| "grad_norm": 0.003458831226453185, | |
| "learning_rate": 6.70923797360579e-06, | |
| "loss": 0.0097, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 19.01693486590038, | |
| "grad_norm": 0.0063134911470115185, | |
| "learning_rate": 6.700723712217966e-06, | |
| "loss": 1.0321, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 19.01770114942529, | |
| "grad_norm": 0.003901480231434107, | |
| "learning_rate": 6.692209450830141e-06, | |
| "loss": 0.0016, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 19.018467432950192, | |
| "grad_norm": 0.007703134790062904, | |
| "learning_rate": 6.683695189442317e-06, | |
| "loss": 1.1552, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 19.019233716475096, | |
| "grad_norm": 0.15453700721263885, | |
| "learning_rate": 6.6751809280544925e-06, | |
| "loss": 1.04, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "grad_norm": 81.24810028076172, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.9924, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.8415968418121338, | |
| "eval_runtime": 28.5446, | |
| "eval_samples_per_second": 1.576, | |
| "eval_steps_per_second": 1.576, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 20.000766283524904, | |
| "grad_norm": 0.011713034473359585, | |
| "learning_rate": 6.6581524052788425e-06, | |
| "loss": 0.0226, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 20.001532567049807, | |
| "grad_norm": 0.040813036262989044, | |
| "learning_rate": 6.649638143891018e-06, | |
| "loss": 0.683, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 20.002298850574714, | |
| "grad_norm": 2.034975290298462, | |
| "learning_rate": 6.641123882503193e-06, | |
| "loss": 0.4885, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 20.003065134099618, | |
| "grad_norm": 0.25844502449035645, | |
| "learning_rate": 6.632609621115369e-06, | |
| "loss": 0.3225, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 20.00383141762452, | |
| "grad_norm": 0.01276817824691534, | |
| "learning_rate": 6.624095359727543e-06, | |
| "loss": 0.0538, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 20.004597701149425, | |
| "grad_norm": 158.18614196777344, | |
| "learning_rate": 6.615581098339719e-06, | |
| "loss": 0.5745, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 20.00536398467433, | |
| "grad_norm": 0.04244688153266907, | |
| "learning_rate": 6.607066836951895e-06, | |
| "loss": 0.7604, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 20.006130268199232, | |
| "grad_norm": 2.517895460128784, | |
| "learning_rate": 6.59855257556407e-06, | |
| "loss": 0.5931, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 20.00689655172414, | |
| "grad_norm": 0.005115671548992395, | |
| "learning_rate": 6.590038314176246e-06, | |
| "loss": 0.4569, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 20.007662835249043, | |
| "grad_norm": 215.1130828857422, | |
| "learning_rate": 6.581524052788422e-06, | |
| "loss": 1.1023, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 20.008429118773947, | |
| "grad_norm": 191.2766876220703, | |
| "learning_rate": 6.573009791400597e-06, | |
| "loss": 0.095, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 20.00919540229885, | |
| "grad_norm": 159.36575317382812, | |
| "learning_rate": 6.5644955300127725e-06, | |
| "loss": 0.6151, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 20.009961685823754, | |
| "grad_norm": 0.3148025572299957, | |
| "learning_rate": 6.555981268624947e-06, | |
| "loss": 1.1696, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 20.010727969348657, | |
| "grad_norm": 0.004226139280945063, | |
| "learning_rate": 6.5474670072371225e-06, | |
| "loss": 0.2537, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 20.011494252873565, | |
| "grad_norm": 0.1940159648656845, | |
| "learning_rate": 6.538952745849298e-06, | |
| "loss": 0.5355, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 20.01226053639847, | |
| "grad_norm": 180.20297241210938, | |
| "learning_rate": 6.530438484461473e-06, | |
| "loss": 0.0471, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 20.013026819923372, | |
| "grad_norm": 0.03982972726225853, | |
| "learning_rate": 6.521924223073649e-06, | |
| "loss": 0.7464, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 20.013793103448275, | |
| "grad_norm": 0.03196590766310692, | |
| "learning_rate": 6.513409961685824e-06, | |
| "loss": 1.0496, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 20.01455938697318, | |
| "grad_norm": 0.0029116757214069366, | |
| "learning_rate": 6.504895700297999e-06, | |
| "loss": 0.0574, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 20.015325670498083, | |
| "grad_norm": 0.014451466500759125, | |
| "learning_rate": 6.496381438910175e-06, | |
| "loss": 0.0944, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 20.01609195402299, | |
| "grad_norm": 6.072531223297119, | |
| "learning_rate": 6.48786717752235e-06, | |
| "loss": 0.2592, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 20.016858237547893, | |
| "grad_norm": 0.5176710486412048, | |
| "learning_rate": 6.479352916134526e-06, | |
| "loss": 1.1315, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 20.017624521072797, | |
| "grad_norm": 52.90974807739258, | |
| "learning_rate": 6.470838654746702e-06, | |
| "loss": 1.3085, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 20.0183908045977, | |
| "grad_norm": 0.0049968864768743515, | |
| "learning_rate": 6.462324393358877e-06, | |
| "loss": 1.4412, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 20.019157088122604, | |
| "grad_norm": 0.005467898678034544, | |
| "learning_rate": 6.4538101319710525e-06, | |
| "loss": 0.0056, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 20.01992337164751, | |
| "grad_norm": 0.0032753855921328068, | |
| "learning_rate": 6.445295870583227e-06, | |
| "loss": 0.0009, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 20.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.906760811805725, | |
| "eval_runtime": 28.6797, | |
| "eval_samples_per_second": 1.569, | |
| "eval_steps_per_second": 1.569, | |
| "step": 5481 | |
| }, | |
| { | |
| "epoch": 21.000689655172415, | |
| "grad_norm": 16.41851234436035, | |
| "learning_rate": 6.4367816091954025e-06, | |
| "loss": 1.0387, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 21.00145593869732, | |
| "grad_norm": 0.43349307775497437, | |
| "learning_rate": 6.428267347807578e-06, | |
| "loss": 0.4997, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 21.002222222222223, | |
| "grad_norm": 0.011135280132293701, | |
| "learning_rate": 6.419753086419753e-06, | |
| "loss": 1.3034, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 21.002988505747126, | |
| "grad_norm": 0.0223165862262249, | |
| "learning_rate": 6.411238825031929e-06, | |
| "loss": 1.339, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 21.00375478927203, | |
| "grad_norm": 0.00240329559892416, | |
| "learning_rate": 6.402724563644105e-06, | |
| "loss": 0.5079, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 21.004521072796933, | |
| "grad_norm": 106.79464721679688, | |
| "learning_rate": 6.39421030225628e-06, | |
| "loss": 1.1305, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 21.00528735632184, | |
| "grad_norm": 0.4033883213996887, | |
| "learning_rate": 6.385696040868455e-06, | |
| "loss": 1.2739, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 21.006053639846744, | |
| "grad_norm": 40.97711181640625, | |
| "learning_rate": 6.37718177948063e-06, | |
| "loss": 2.3358, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 21.006819923371648, | |
| "grad_norm": 0.10723423212766647, | |
| "learning_rate": 6.368667518092806e-06, | |
| "loss": 0.3916, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 21.00758620689655, | |
| "grad_norm": 0.027807606384158134, | |
| "learning_rate": 6.360153256704982e-06, | |
| "loss": 0.0045, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 21.008352490421455, | |
| "grad_norm": 0.024614203721284866, | |
| "learning_rate": 6.351638995317157e-06, | |
| "loss": 0.0085, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 21.00911877394636, | |
| "grad_norm": 11.62214183807373, | |
| "learning_rate": 6.3431247339293325e-06, | |
| "loss": 0.0058, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 21.009885057471266, | |
| "grad_norm": 0.05164700374007225, | |
| "learning_rate": 6.334610472541508e-06, | |
| "loss": 1.0253, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 21.01065134099617, | |
| "grad_norm": 36.70790481567383, | |
| "learning_rate": 6.3260962111536825e-06, | |
| "loss": 0.0111, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 21.011417624521073, | |
| "grad_norm": 0.048200830817222595, | |
| "learning_rate": 6.317581949765858e-06, | |
| "loss": 0.4736, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 21.012183908045976, | |
| "grad_norm": 0.007409057579934597, | |
| "learning_rate": 6.309067688378033e-06, | |
| "loss": 0.5227, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 21.01295019157088, | |
| "grad_norm": 0.05748462304472923, | |
| "learning_rate": 6.300553426990209e-06, | |
| "loss": 0.0005, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 21.013716475095784, | |
| "grad_norm": 0.00826114322990179, | |
| "learning_rate": 6.292039165602385e-06, | |
| "loss": 0.3745, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 21.01448275862069, | |
| "grad_norm": 0.0020533306524157524, | |
| "learning_rate": 6.28352490421456e-06, | |
| "loss": 0.0901, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 21.015249042145594, | |
| "grad_norm": 0.7792718410491943, | |
| "learning_rate": 6.275010642826736e-06, | |
| "loss": 0.0015, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 21.016015325670498, | |
| "grad_norm": 0.04261056333780289, | |
| "learning_rate": 6.26649638143891e-06, | |
| "loss": 0.012, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 21.0167816091954, | |
| "grad_norm": 0.08563796430826187, | |
| "learning_rate": 6.257982120051086e-06, | |
| "loss": 1.0467, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 21.017547892720305, | |
| "grad_norm": 487.4284362792969, | |
| "learning_rate": 6.249467858663262e-06, | |
| "loss": 0.8018, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 21.018314176245212, | |
| "grad_norm": 5.818784713745117, | |
| "learning_rate": 6.240953597275437e-06, | |
| "loss": 0.729, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 21.019080459770116, | |
| "grad_norm": 0.005066063720732927, | |
| "learning_rate": 6.2324393358876125e-06, | |
| "loss": 0.0062, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 21.01984674329502, | |
| "grad_norm": 0.0023419486824423075, | |
| "learning_rate": 6.223925074499788e-06, | |
| "loss": 1.0231, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.8427561521530151, | |
| "eval_runtime": 29.6258, | |
| "eval_samples_per_second": 1.519, | |
| "eval_steps_per_second": 1.519, | |
| "step": 5742 | |
| }, | |
| { | |
| "epoch": 22.000613026819924, | |
| "grad_norm": 1.9332926273345947, | |
| "learning_rate": 6.2154108131119625e-06, | |
| "loss": 0.0014, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 22.001379310344827, | |
| "grad_norm": 0.0077827018685638905, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.0708, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 22.00214559386973, | |
| "grad_norm": 0.0115917744114995, | |
| "learning_rate": 6.198382290336313e-06, | |
| "loss": 0.3427, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 22.002911877394634, | |
| "grad_norm": 0.023211613297462463, | |
| "learning_rate": 6.189868028948489e-06, | |
| "loss": 0.5441, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 22.00367816091954, | |
| "grad_norm": 0.48048824071884155, | |
| "learning_rate": 6.181353767560665e-06, | |
| "loss": 0.7479, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 22.004444444444445, | |
| "grad_norm": 0.0017794081941246986, | |
| "learning_rate": 6.17283950617284e-06, | |
| "loss": 1.1613, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 22.00521072796935, | |
| "grad_norm": 205.38832092285156, | |
| "learning_rate": 6.164325244785016e-06, | |
| "loss": 1.2724, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 22.005977011494252, | |
| "grad_norm": 0.03380270302295685, | |
| "learning_rate": 6.155810983397192e-06, | |
| "loss": 0.1442, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 22.006743295019156, | |
| "grad_norm": 0.25575459003448486, | |
| "learning_rate": 6.147296722009366e-06, | |
| "loss": 0.0053, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 22.00750957854406, | |
| "grad_norm": 0.06640543788671494, | |
| "learning_rate": 6.138782460621542e-06, | |
| "loss": 0.0224, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 22.008275862068967, | |
| "grad_norm": 0.08240759372711182, | |
| "learning_rate": 6.130268199233717e-06, | |
| "loss": 1.3887, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 22.00904214559387, | |
| "grad_norm": 0.10972102731466293, | |
| "learning_rate": 6.1217539378458925e-06, | |
| "loss": 0.9687, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 22.009808429118774, | |
| "grad_norm": 0.25319355726242065, | |
| "learning_rate": 6.113239676458068e-06, | |
| "loss": 0.547, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 22.010574712643677, | |
| "grad_norm": 0.0020377521868795156, | |
| "learning_rate": 6.1047254150702425e-06, | |
| "loss": 0.72, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 22.01134099616858, | |
| "grad_norm": 2.363307476043701, | |
| "learning_rate": 6.096211153682418e-06, | |
| "loss": 0.5225, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 22.01210727969349, | |
| "grad_norm": 0.9951919317245483, | |
| "learning_rate": 6.087696892294594e-06, | |
| "loss": 0.0024, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 22.012873563218392, | |
| "grad_norm": 0.02694072388112545, | |
| "learning_rate": 6.079182630906769e-06, | |
| "loss": 1.2668, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 22.013639846743295, | |
| "grad_norm": 0.03728149086236954, | |
| "learning_rate": 6.070668369518945e-06, | |
| "loss": 0.1775, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 22.0144061302682, | |
| "grad_norm": 5.004837512969971, | |
| "learning_rate": 6.06215410813112e-06, | |
| "loss": 0.2987, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 22.015172413793103, | |
| "grad_norm": 0.056994613260030746, | |
| "learning_rate": 6.053639846743296e-06, | |
| "loss": 0.6439, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 22.015938697318006, | |
| "grad_norm": 0.0020670308731496334, | |
| "learning_rate": 6.045125585355472e-06, | |
| "loss": 0.0004, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 22.016704980842913, | |
| "grad_norm": 0.0823218896985054, | |
| "learning_rate": 6.036611323967646e-06, | |
| "loss": 0.6288, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 22.017471264367817, | |
| "grad_norm": 0.015304329805076122, | |
| "learning_rate": 6.028097062579822e-06, | |
| "loss": 0.1777, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 22.01823754789272, | |
| "grad_norm": 142.00123596191406, | |
| "learning_rate": 6.019582801191997e-06, | |
| "loss": 1.3363, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 22.019003831417624, | |
| "grad_norm": 0.25328370928764343, | |
| "learning_rate": 6.0110685398041725e-06, | |
| "loss": 0.5996, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 22.019770114942528, | |
| "grad_norm": 0.02333473414182663, | |
| "learning_rate": 6.002554278416348e-06, | |
| "loss": 0.7099, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 22.02, | |
| "eval_accuracy": 0.6, | |
| "eval_loss": 2.310786724090576, | |
| "eval_runtime": 28.8965, | |
| "eval_samples_per_second": 1.557, | |
| "eval_steps_per_second": 1.557, | |
| "step": 6003 | |
| }, | |
| { | |
| "epoch": 23.000536398467432, | |
| "grad_norm": 264.2043762207031, | |
| "learning_rate": 5.9940400170285225e-06, | |
| "loss": 0.6163, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 23.001302681992335, | |
| "grad_norm": 16.014562606811523, | |
| "learning_rate": 5.985525755640698e-06, | |
| "loss": 0.0064, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 23.002068965517243, | |
| "grad_norm": 0.02582295797765255, | |
| "learning_rate": 5.977011494252874e-06, | |
| "loss": 0.2809, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 23.002835249042146, | |
| "grad_norm": 0.0813617929816246, | |
| "learning_rate": 5.968497232865049e-06, | |
| "loss": 0.0039, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 23.00360153256705, | |
| "grad_norm": 0.033921342343091965, | |
| "learning_rate": 5.959982971477225e-06, | |
| "loss": 0.77, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 23.004367816091953, | |
| "grad_norm": 118.4710922241211, | |
| "learning_rate": 5.9514687100894e-06, | |
| "loss": 0.655, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 23.005134099616857, | |
| "grad_norm": 0.005723563954234123, | |
| "learning_rate": 5.942954448701576e-06, | |
| "loss": 1.0035, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 23.005900383141764, | |
| "grad_norm": 0.04950481280684471, | |
| "learning_rate": 5.934440187313752e-06, | |
| "loss": 0.0008, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 23.006666666666668, | |
| "grad_norm": 0.02758442424237728, | |
| "learning_rate": 5.925925925925926e-06, | |
| "loss": 0.6145, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 23.00743295019157, | |
| "grad_norm": 191.37852478027344, | |
| "learning_rate": 5.917411664538102e-06, | |
| "loss": 0.9092, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 23.008199233716475, | |
| "grad_norm": 0.030716927722096443, | |
| "learning_rate": 5.9088974031502775e-06, | |
| "loss": 0.809, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 23.00896551724138, | |
| "grad_norm": 32.45998764038086, | |
| "learning_rate": 5.9003831417624525e-06, | |
| "loss": 0.0049, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 23.009731800766282, | |
| "grad_norm": 0.2010127305984497, | |
| "learning_rate": 5.891868880374628e-06, | |
| "loss": 1.007, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 23.01049808429119, | |
| "grad_norm": 0.01684950478374958, | |
| "learning_rate": 5.883354618986803e-06, | |
| "loss": 0.0194, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 23.011264367816093, | |
| "grad_norm": 263.1514892578125, | |
| "learning_rate": 5.874840357598979e-06, | |
| "loss": 0.1132, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 23.012030651340996, | |
| "grad_norm": 0.03413116931915283, | |
| "learning_rate": 5.866326096211154e-06, | |
| "loss": 0.6505, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 23.0127969348659, | |
| "grad_norm": 0.08276499062776566, | |
| "learning_rate": 5.857811834823329e-06, | |
| "loss": 0.0275, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 23.013563218390804, | |
| "grad_norm": 0.034385498613119125, | |
| "learning_rate": 5.849297573435505e-06, | |
| "loss": 0.0785, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 23.014329501915707, | |
| "grad_norm": 0.14270183444023132, | |
| "learning_rate": 5.84078331204768e-06, | |
| "loss": 2.0852, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 23.015095785440614, | |
| "grad_norm": 0.05579046159982681, | |
| "learning_rate": 5.832269050659856e-06, | |
| "loss": 1.4851, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 23.015862068965518, | |
| "grad_norm": 0.009700673632323742, | |
| "learning_rate": 5.823754789272032e-06, | |
| "loss": 0.4295, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 23.01662835249042, | |
| "grad_norm": 2.6064889430999756, | |
| "learning_rate": 5.815240527884206e-06, | |
| "loss": 0.816, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 23.017394636015325, | |
| "grad_norm": 38.40084457397461, | |
| "learning_rate": 5.806726266496382e-06, | |
| "loss": 1.3302, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 23.01816091954023, | |
| "grad_norm": 0.016617679968476295, | |
| "learning_rate": 5.7982120051085575e-06, | |
| "loss": 0.6474, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 23.018927203065132, | |
| "grad_norm": 74.77079010009766, | |
| "learning_rate": 5.7896977437207325e-06, | |
| "loss": 1.3786, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 23.01969348659004, | |
| "grad_norm": 0.05970478057861328, | |
| "learning_rate": 5.781183482332908e-06, | |
| "loss": 0.3243, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 23.02, | |
| "eval_accuracy": 0.5777777777777777, | |
| "eval_loss": 2.2084178924560547, | |
| "eval_runtime": 28.5847, | |
| "eval_samples_per_second": 1.574, | |
| "eval_steps_per_second": 1.574, | |
| "step": 6264 | |
| }, | |
| { | |
| "epoch": 24.000459770114944, | |
| "grad_norm": 65.56672668457031, | |
| "learning_rate": 5.772669220945083e-06, | |
| "loss": 1.2783, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 24.001226053639847, | |
| "grad_norm": 155.61605834960938, | |
| "learning_rate": 5.764154959557259e-06, | |
| "loss": 0.4797, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 24.00199233716475, | |
| "grad_norm": 178.5623321533203, | |
| "learning_rate": 5.755640698169435e-06, | |
| "loss": 1.1384, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 24.002758620689654, | |
| "grad_norm": 0.0167319905012846, | |
| "learning_rate": 5.747126436781609e-06, | |
| "loss": 0.6776, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 24.003524904214558, | |
| "grad_norm": 0.00547564122825861, | |
| "learning_rate": 5.738612175393785e-06, | |
| "loss": 0.0013, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 24.004291187739465, | |
| "grad_norm": 0.0025295685045421124, | |
| "learning_rate": 5.730097914005961e-06, | |
| "loss": 0.4849, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 24.00505747126437, | |
| "grad_norm": 258.21630859375, | |
| "learning_rate": 5.721583652618136e-06, | |
| "loss": 2.3456, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 24.005823754789272, | |
| "grad_norm": 0.06951650232076645, | |
| "learning_rate": 5.713069391230312e-06, | |
| "loss": 0.846, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 24.006590038314176, | |
| "grad_norm": 0.292102187871933, | |
| "learning_rate": 5.704555129842486e-06, | |
| "loss": 0.5234, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 24.00735632183908, | |
| "grad_norm": 0.003594489535316825, | |
| "learning_rate": 5.696040868454662e-06, | |
| "loss": 0.3726, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 24.008122605363983, | |
| "grad_norm": 0.14738978445529938, | |
| "learning_rate": 5.6875266070668375e-06, | |
| "loss": 0.6347, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 24.00888888888889, | |
| "grad_norm": 0.005358193535357714, | |
| "learning_rate": 5.6790123456790125e-06, | |
| "loss": 0.0458, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 24.009655172413794, | |
| "grad_norm": 0.09735672920942307, | |
| "learning_rate": 5.670498084291188e-06, | |
| "loss": 0.4581, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 24.010421455938697, | |
| "grad_norm": 0.0025942621286958456, | |
| "learning_rate": 5.661983822903364e-06, | |
| "loss": 0.4006, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 24.0111877394636, | |
| "grad_norm": 0.01855681836605072, | |
| "learning_rate": 5.653469561515539e-06, | |
| "loss": 0.3958, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 24.011954022988505, | |
| "grad_norm": 0.016031399369239807, | |
| "learning_rate": 5.644955300127715e-06, | |
| "loss": 0.0013, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 24.01272030651341, | |
| "grad_norm": 0.005647106096148491, | |
| "learning_rate": 5.636441038739889e-06, | |
| "loss": 0.5694, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 24.013486590038315, | |
| "grad_norm": 0.012183071114122868, | |
| "learning_rate": 5.627926777352065e-06, | |
| "loss": 0.713, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 24.01425287356322, | |
| "grad_norm": 1.213383674621582, | |
| "learning_rate": 5.619412515964241e-06, | |
| "loss": 0.5549, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 24.015019157088123, | |
| "grad_norm": 0.16394241154193878, | |
| "learning_rate": 5.610898254576416e-06, | |
| "loss": 1.1725, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 24.015785440613026, | |
| "grad_norm": 0.0020893006585538387, | |
| "learning_rate": 5.602383993188592e-06, | |
| "loss": 0.4155, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 24.01655172413793, | |
| "grad_norm": 0.01788010448217392, | |
| "learning_rate": 5.593869731800766e-06, | |
| "loss": 0.0024, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 24.017318007662837, | |
| "grad_norm": 0.005845135543495417, | |
| "learning_rate": 5.585355470412942e-06, | |
| "loss": 0.0005, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 24.01808429118774, | |
| "grad_norm": 0.015230582095682621, | |
| "learning_rate": 5.5768412090251175e-06, | |
| "loss": 0.2815, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 24.018850574712644, | |
| "grad_norm": 510.5647888183594, | |
| "learning_rate": 5.5683269476372925e-06, | |
| "loss": 0.6476, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 24.019616858237548, | |
| "grad_norm": 258.3885803222656, | |
| "learning_rate": 5.559812686249468e-06, | |
| "loss": 2.748, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 24.02, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.8854796886444092, | |
| "eval_runtime": 27.805, | |
| "eval_samples_per_second": 1.618, | |
| "eval_steps_per_second": 1.618, | |
| "step": 6525 | |
| }, | |
| { | |
| "epoch": 25.000383141762452, | |
| "grad_norm": 1.6865376234054565, | |
| "learning_rate": 5.551298424861644e-06, | |
| "loss": 0.7207, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 25.001149425287355, | |
| "grad_norm": 0.03704597428441048, | |
| "learning_rate": 5.542784163473819e-06, | |
| "loss": 0.0084, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 25.00191570881226, | |
| "grad_norm": 0.3662070035934448, | |
| "learning_rate": 5.534269902085995e-06, | |
| "loss": 0.0758, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 25.002681992337166, | |
| "grad_norm": 0.0085606649518013, | |
| "learning_rate": 5.525755640698169e-06, | |
| "loss": 0.7967, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 25.00344827586207, | |
| "grad_norm": 363.8013916015625, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.9311, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 25.004214559386973, | |
| "grad_norm": 0.005399974528700113, | |
| "learning_rate": 5.508727117922521e-06, | |
| "loss": 1.1345, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 25.004980842911877, | |
| "grad_norm": 0.015188188292086124, | |
| "learning_rate": 5.500212856534696e-06, | |
| "loss": 0.5137, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 25.00574712643678, | |
| "grad_norm": 0.008783702738583088, | |
| "learning_rate": 5.491698595146872e-06, | |
| "loss": 0.3957, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 25.006513409961684, | |
| "grad_norm": 0.006709863897413015, | |
| "learning_rate": 5.4831843337590475e-06, | |
| "loss": 0.0008, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 25.00727969348659, | |
| "grad_norm": 0.009334388189017773, | |
| "learning_rate": 5.474670072371222e-06, | |
| "loss": 0.6295, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 25.008045977011495, | |
| "grad_norm": 0.0069431099109351635, | |
| "learning_rate": 5.4661558109833975e-06, | |
| "loss": 0.0375, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 25.0088122605364, | |
| "grad_norm": 0.0025293382350355387, | |
| "learning_rate": 5.4576415495955725e-06, | |
| "loss": 1.8296, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 25.009578544061302, | |
| "grad_norm": 0.026507409289479256, | |
| "learning_rate": 5.449127288207748e-06, | |
| "loss": 0.1752, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 25.010344827586206, | |
| "grad_norm": 0.12145673483610153, | |
| "learning_rate": 5.440613026819924e-06, | |
| "loss": 0.1511, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 25.011111111111113, | |
| "grad_norm": 0.005514368414878845, | |
| "learning_rate": 5.432098765432099e-06, | |
| "loss": 0.9555, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 25.011877394636016, | |
| "grad_norm": 275.9948425292969, | |
| "learning_rate": 5.423584504044275e-06, | |
| "loss": 0.3381, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 25.01264367816092, | |
| "grad_norm": 0.0019095117459073663, | |
| "learning_rate": 5.415070242656451e-06, | |
| "loss": 0.1394, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 25.013409961685824, | |
| "grad_norm": 0.030332105234265327, | |
| "learning_rate": 5.406555981268625e-06, | |
| "loss": 0.2852, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 25.014176245210727, | |
| "grad_norm": 0.0034307390451431274, | |
| "learning_rate": 5.398041719880801e-06, | |
| "loss": 1.0962, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 25.01494252873563, | |
| "grad_norm": 0.02897072024643421, | |
| "learning_rate": 5.389527458492976e-06, | |
| "loss": 0.0025, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 25.015708812260538, | |
| "grad_norm": 0.024984782561659813, | |
| "learning_rate": 5.381013197105152e-06, | |
| "loss": 0.0137, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 25.01647509578544, | |
| "grad_norm": 0.15218405425548553, | |
| "learning_rate": 5.3724989357173275e-06, | |
| "loss": 2.2803, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 25.017241379310345, | |
| "grad_norm": 0.013546308502554893, | |
| "learning_rate": 5.3639846743295025e-06, | |
| "loss": 0.9654, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 25.01800766283525, | |
| "grad_norm": 0.30212756991386414, | |
| "learning_rate": 5.3554704129416775e-06, | |
| "loss": 0.0033, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 25.018773946360152, | |
| "grad_norm": 173.64010620117188, | |
| "learning_rate": 5.3469561515538525e-06, | |
| "loss": 1.3313, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 25.019540229885056, | |
| "grad_norm": 0.016651857644319534, | |
| "learning_rate": 5.338441890166028e-06, | |
| "loss": 0.0002, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 25.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.9442589282989502, | |
| "eval_runtime": 27.8613, | |
| "eval_samples_per_second": 1.615, | |
| "eval_steps_per_second": 1.615, | |
| "step": 6786 | |
| }, | |
| { | |
| "epoch": 26.00030651340996, | |
| "grad_norm": 0.05307227000594139, | |
| "learning_rate": 5.329927628778204e-06, | |
| "loss": 0.5638, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 26.001072796934867, | |
| "grad_norm": 0.015339828096330166, | |
| "learning_rate": 5.321413367390379e-06, | |
| "loss": 0.3586, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 26.00183908045977, | |
| "grad_norm": 0.05491383746266365, | |
| "learning_rate": 5.312899106002555e-06, | |
| "loss": 0.0109, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 26.002605363984674, | |
| "grad_norm": 0.00443470012396574, | |
| "learning_rate": 5.304384844614731e-06, | |
| "loss": 0.003, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 26.003371647509578, | |
| "grad_norm": 0.43860122561454773, | |
| "learning_rate": 5.295870583226905e-06, | |
| "loss": 0.6822, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 26.00413793103448, | |
| "grad_norm": 0.024054815992712975, | |
| "learning_rate": 5.287356321839081e-06, | |
| "loss": 0.0067, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 26.00490421455939, | |
| "grad_norm": 0.0055478704161942005, | |
| "learning_rate": 5.278842060451256e-06, | |
| "loss": 0.535, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 26.005670498084292, | |
| "grad_norm": 22.662864685058594, | |
| "learning_rate": 5.270327799063432e-06, | |
| "loss": 0.1925, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 26.006436781609196, | |
| "grad_norm": 0.003003709949553013, | |
| "learning_rate": 5.2618135376756075e-06, | |
| "loss": 0.5777, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 26.0072030651341, | |
| "grad_norm": 0.0020995778031647205, | |
| "learning_rate": 5.2532992762877825e-06, | |
| "loss": 0.658, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 26.007969348659003, | |
| "grad_norm": 70.04608154296875, | |
| "learning_rate": 5.244785014899958e-06, | |
| "loss": 0.0501, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 26.008735632183907, | |
| "grad_norm": 0.003194736083969474, | |
| "learning_rate": 5.236270753512134e-06, | |
| "loss": 0.2504, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 26.009501915708814, | |
| "grad_norm": 102.48612976074219, | |
| "learning_rate": 5.227756492124308e-06, | |
| "loss": 1.3037, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 26.010268199233717, | |
| "grad_norm": 47.21868896484375, | |
| "learning_rate": 5.219242230736484e-06, | |
| "loss": 0.9629, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 26.01103448275862, | |
| "grad_norm": 233.42538452148438, | |
| "learning_rate": 5.210727969348659e-06, | |
| "loss": 0.5676, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 26.011800766283525, | |
| "grad_norm": 0.03407863900065422, | |
| "learning_rate": 5.202213707960835e-06, | |
| "loss": 0.0005, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 26.01256704980843, | |
| "grad_norm": 0.002790221245959401, | |
| "learning_rate": 5.193699446573011e-06, | |
| "loss": 0.0002, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 26.013333333333332, | |
| "grad_norm": 47.864707946777344, | |
| "learning_rate": 5.185185185185185e-06, | |
| "loss": 2.1923, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 26.01409961685824, | |
| "grad_norm": 0.0036934618838131428, | |
| "learning_rate": 5.176670923797361e-06, | |
| "loss": 0.0143, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 26.014865900383143, | |
| "grad_norm": 0.016530748456716537, | |
| "learning_rate": 5.168156662409536e-06, | |
| "loss": 1.5982, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 26.015632183908046, | |
| "grad_norm": 0.1699918806552887, | |
| "learning_rate": 5.159642401021712e-06, | |
| "loss": 0.0018, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 26.01639846743295, | |
| "grad_norm": 0.07781951874494553, | |
| "learning_rate": 5.1511281396338875e-06, | |
| "loss": 0.2596, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 26.017164750957853, | |
| "grad_norm": 0.003624899545684457, | |
| "learning_rate": 5.1426138782460625e-06, | |
| "loss": 0.0094, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 26.017931034482757, | |
| "grad_norm": 15.835947036743164, | |
| "learning_rate": 5.134099616858238e-06, | |
| "loss": 0.6469, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 26.018697318007664, | |
| "grad_norm": 0.0449938103556633, | |
| "learning_rate": 5.125585355470414e-06, | |
| "loss": 0.4909, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 26.019463601532568, | |
| "grad_norm": 0.5268478393554688, | |
| "learning_rate": 5.117071094082588e-06, | |
| "loss": 1.1288, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 1.6372127532958984, | |
| "eval_runtime": 28.2494, | |
| "eval_samples_per_second": 1.593, | |
| "eval_steps_per_second": 1.593, | |
| "step": 7047 | |
| }, | |
| { | |
| "epoch": 27.000229885057472, | |
| "grad_norm": 0.0025100375059992075, | |
| "learning_rate": 5.108556832694764e-06, | |
| "loss": 1.3201, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 27.000996168582375, | |
| "grad_norm": 0.001371032907627523, | |
| "learning_rate": 5.100042571306939e-06, | |
| "loss": 0.4515, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 27.00176245210728, | |
| "grad_norm": 0.034923654049634933, | |
| "learning_rate": 5.091528309919115e-06, | |
| "loss": 1.1018, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 27.002528735632183, | |
| "grad_norm": 0.020795369520783424, | |
| "learning_rate": 5.083014048531291e-06, | |
| "loss": 0.3081, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 27.00329501915709, | |
| "grad_norm": 0.0011883447878062725, | |
| "learning_rate": 5.074499787143465e-06, | |
| "loss": 0.6769, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 27.004061302681993, | |
| "grad_norm": 0.0698217898607254, | |
| "learning_rate": 5.065985525755641e-06, | |
| "loss": 0.0032, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 27.004827586206897, | |
| "grad_norm": 0.01318784523755312, | |
| "learning_rate": 5.057471264367817e-06, | |
| "loss": 0.0014, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 27.0055938697318, | |
| "grad_norm": 0.0032538459636271, | |
| "learning_rate": 5.048957002979992e-06, | |
| "loss": 1.2139, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 27.006360153256704, | |
| "grad_norm": 0.005463645793497562, | |
| "learning_rate": 5.0404427415921675e-06, | |
| "loss": 0.0029, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 27.007126436781608, | |
| "grad_norm": 0.002044873544946313, | |
| "learning_rate": 5.0319284802043425e-06, | |
| "loss": 0.2554, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 27.007892720306515, | |
| "grad_norm": 150.93966674804688, | |
| "learning_rate": 5.023414218816518e-06, | |
| "loss": 0.5414, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 27.00865900383142, | |
| "grad_norm": 585.3448486328125, | |
| "learning_rate": 5.014899957428694e-06, | |
| "loss": 1.0977, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 27.009425287356322, | |
| "grad_norm": 0.21827629208564758, | |
| "learning_rate": 5.006385696040868e-06, | |
| "loss": 0.0026, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 27.010191570881226, | |
| "grad_norm": 0.14477881789207458, | |
| "learning_rate": 4.997871434653044e-06, | |
| "loss": 0.9725, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 27.01095785440613, | |
| "grad_norm": 0.012054833583533764, | |
| "learning_rate": 4.98935717326522e-06, | |
| "loss": 1.4246, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 27.011724137931033, | |
| "grad_norm": 0.007552944589406252, | |
| "learning_rate": 4.980842911877395e-06, | |
| "loss": 0.0006, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 27.01249042145594, | |
| "grad_norm": 0.04953211545944214, | |
| "learning_rate": 4.972328650489571e-06, | |
| "loss": 0.2401, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 27.013256704980844, | |
| "grad_norm": 0.004737659823149443, | |
| "learning_rate": 4.963814389101746e-06, | |
| "loss": 0.4782, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 27.014022988505747, | |
| "grad_norm": 0.0042915926314890385, | |
| "learning_rate": 4.955300127713921e-06, | |
| "loss": 0.0003, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 27.01478927203065, | |
| "grad_norm": 0.0024593069683760405, | |
| "learning_rate": 4.946785866326097e-06, | |
| "loss": 0.0004, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 27.015555555555554, | |
| "grad_norm": 0.0017103638965636492, | |
| "learning_rate": 4.938271604938272e-06, | |
| "loss": 0.1948, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 27.016321839080458, | |
| "grad_norm": 0.9441583752632141, | |
| "learning_rate": 4.9297573435504475e-06, | |
| "loss": 0.4695, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 27.017088122605365, | |
| "grad_norm": 0.042855050414800644, | |
| "learning_rate": 4.9212430821626225e-06, | |
| "loss": 1.0336, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 27.01785440613027, | |
| "grad_norm": 0.09851084649562836, | |
| "learning_rate": 4.912728820774798e-06, | |
| "loss": 0.1355, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 27.018620689655172, | |
| "grad_norm": 0.004743156488984823, | |
| "learning_rate": 4.904214559386973e-06, | |
| "loss": 0.0014, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 27.019386973180076, | |
| "grad_norm": 12.480242729187012, | |
| "learning_rate": 4.895700297999149e-06, | |
| "loss": 0.0024, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 27.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 2.0813233852386475, | |
| "eval_runtime": 27.9067, | |
| "eval_samples_per_second": 1.613, | |
| "eval_steps_per_second": 1.613, | |
| "step": 7308 | |
| }, | |
| { | |
| "epoch": 28.00015325670498, | |
| "grad_norm": 0.001908393925987184, | |
| "learning_rate": 4.887186036611324e-06, | |
| "loss": 0.3898, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 28.000919540229884, | |
| "grad_norm": 0.000980175449512899, | |
| "learning_rate": 4.8786717752235e-06, | |
| "loss": 0.6533, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 28.00168582375479, | |
| "grad_norm": 0.007126258220523596, | |
| "learning_rate": 4.870157513835675e-06, | |
| "loss": 0.2288, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 28.002452107279694, | |
| "grad_norm": 0.38791441917419434, | |
| "learning_rate": 4.861643252447851e-06, | |
| "loss": 1.6321, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 28.003218390804598, | |
| "grad_norm": 0.009356196038424969, | |
| "learning_rate": 4.853128991060026e-06, | |
| "loss": 0.7451, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 28.0039846743295, | |
| "grad_norm": 0.027178501710295677, | |
| "learning_rate": 4.844614729672202e-06, | |
| "loss": 0.0056, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 28.004750957854405, | |
| "grad_norm": 0.03039945475757122, | |
| "learning_rate": 4.836100468284377e-06, | |
| "loss": 0.0028, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 28.00551724137931, | |
| "grad_norm": 0.03347291797399521, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.0009, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 28.006283524904216, | |
| "grad_norm": 0.015016093850135803, | |
| "learning_rate": 4.8190719455087275e-06, | |
| "loss": 0.3988, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 28.00704980842912, | |
| "grad_norm": 0.004948149900883436, | |
| "learning_rate": 4.8105576841209025e-06, | |
| "loss": 0.0003, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 28.007816091954023, | |
| "grad_norm": 3.105440378189087, | |
| "learning_rate": 4.802043422733078e-06, | |
| "loss": 0.0025, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 28.008582375478927, | |
| "grad_norm": 1.1475921869277954, | |
| "learning_rate": 4.793529161345254e-06, | |
| "loss": 0.0019, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 28.00934865900383, | |
| "grad_norm": 0.013491490855813026, | |
| "learning_rate": 4.785014899957429e-06, | |
| "loss": 0.0005, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 28.010114942528734, | |
| "grad_norm": 0.0009091476676985621, | |
| "learning_rate": 4.776500638569604e-06, | |
| "loss": 0.6723, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 28.01088122605364, | |
| "grad_norm": 0.0025523528456687927, | |
| "learning_rate": 4.76798637718178e-06, | |
| "loss": 0.0002, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 28.011647509578545, | |
| "grad_norm": 0.00285366945900023, | |
| "learning_rate": 4.759472115793956e-06, | |
| "loss": 0.7087, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 28.01241379310345, | |
| "grad_norm": 0.004684296902269125, | |
| "learning_rate": 4.750957854406131e-06, | |
| "loss": 0.3238, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 28.013180076628352, | |
| "grad_norm": 0.0008410373702645302, | |
| "learning_rate": 4.742443593018306e-06, | |
| "loss": 0.0003, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 28.013946360153255, | |
| "grad_norm": 0.030775954946875572, | |
| "learning_rate": 4.733929331630482e-06, | |
| "loss": 0.0073, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 28.014712643678163, | |
| "grad_norm": 0.005375114735215902, | |
| "learning_rate": 4.7254150702426575e-06, | |
| "loss": 0.7178, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 28.015478927203066, | |
| "grad_norm": 0.05854375287890434, | |
| "learning_rate": 4.7169008088548325e-06, | |
| "loss": 0.0007, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 28.01624521072797, | |
| "grad_norm": 0.02918917126953602, | |
| "learning_rate": 4.7083865474670075e-06, | |
| "loss": 0.0003, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 28.017011494252873, | |
| "grad_norm": 344.5948181152344, | |
| "learning_rate": 4.6998722860791825e-06, | |
| "loss": 0.5259, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 28.017777777777777, | |
| "grad_norm": 521.097412109375, | |
| "learning_rate": 4.691358024691358e-06, | |
| "loss": 0.8263, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 28.01854406130268, | |
| "grad_norm": 0.0011964929290115833, | |
| "learning_rate": 4.682843763303534e-06, | |
| "loss": 1.0728, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 28.019310344827588, | |
| "grad_norm": 0.7627772688865662, | |
| "learning_rate": 4.674329501915709e-06, | |
| "loss": 1.3731, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 28.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 2.184565782546997, | |
| "eval_runtime": 30.1621, | |
| "eval_samples_per_second": 1.492, | |
| "eval_steps_per_second": 1.492, | |
| "step": 7569 | |
| }, | |
| { | |
| "epoch": 29.000076628352492, | |
| "grad_norm": 0.02546844072639942, | |
| "learning_rate": 4.665815240527884e-06, | |
| "loss": 0.0006, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 29.000842911877395, | |
| "grad_norm": 0.0161290280520916, | |
| "learning_rate": 4.65730097914006e-06, | |
| "loss": 0.601, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 29.0016091954023, | |
| "grad_norm": 0.0015779354143887758, | |
| "learning_rate": 4.648786717752236e-06, | |
| "loss": 0.6467, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 29.002375478927203, | |
| "grad_norm": 0.0008723331266082823, | |
| "learning_rate": 4.640272456364411e-06, | |
| "loss": 0.4102, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 29.003141762452106, | |
| "grad_norm": 0.8687963485717773, | |
| "learning_rate": 4.631758194976586e-06, | |
| "loss": 0.0015, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 29.00390804597701, | |
| "grad_norm": 0.002515859203413129, | |
| "learning_rate": 4.623243933588762e-06, | |
| "loss": 1.1717, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 29.004674329501917, | |
| "grad_norm": 0.007008945569396019, | |
| "learning_rate": 4.6147296722009375e-06, | |
| "loss": 0.4936, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 29.00544061302682, | |
| "grad_norm": 0.0009232726297341287, | |
| "learning_rate": 4.6062154108131125e-06, | |
| "loss": 0.3088, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 29.006206896551724, | |
| "grad_norm": 0.0009513779659755528, | |
| "learning_rate": 4.5977011494252875e-06, | |
| "loss": 0.7389, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 29.006973180076628, | |
| "grad_norm": 0.22152279317378998, | |
| "learning_rate": 4.589186888037463e-06, | |
| "loss": 0.28, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 29.00773946360153, | |
| "grad_norm": 0.0011534243822097778, | |
| "learning_rate": 4.580672626649638e-06, | |
| "loss": 0.5633, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 29.00850574712644, | |
| "grad_norm": 0.000674307462759316, | |
| "learning_rate": 4.572158365261814e-06, | |
| "loss": 0.0018, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 29.009272030651342, | |
| "grad_norm": 0.016125526279211044, | |
| "learning_rate": 4.563644103873989e-06, | |
| "loss": 0.0006, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 29.010038314176246, | |
| "grad_norm": 0.044381022453308105, | |
| "learning_rate": 4.555129842486164e-06, | |
| "loss": 0.0303, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 29.01080459770115, | |
| "grad_norm": 0.028124311938881874, | |
| "learning_rate": 4.54661558109834e-06, | |
| "loss": 0.3185, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 29.011570881226053, | |
| "grad_norm": 0.0007618298986926675, | |
| "learning_rate": 4.538101319710516e-06, | |
| "loss": 0.0004, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 29.012337164750956, | |
| "grad_norm": 458.9365234375, | |
| "learning_rate": 4.529587058322691e-06, | |
| "loss": 0.6445, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 29.013103448275864, | |
| "grad_norm": 0.0006640357314608991, | |
| "learning_rate": 4.521072796934866e-06, | |
| "loss": 0.5112, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 29.013869731800767, | |
| "grad_norm": 0.02549567073583603, | |
| "learning_rate": 4.512558535547042e-06, | |
| "loss": 0.1595, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 29.01463601532567, | |
| "grad_norm": 77.58110809326172, | |
| "learning_rate": 4.5040442741592175e-06, | |
| "loss": 1.3087, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 29.015402298850574, | |
| "grad_norm": 121.54621124267578, | |
| "learning_rate": 4.4955300127713925e-06, | |
| "loss": 0.6784, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 29.016168582375478, | |
| "grad_norm": 0.055833850055933, | |
| "learning_rate": 4.4870157513835675e-06, | |
| "loss": 0.001, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 29.01693486590038, | |
| "grad_norm": 0.0011547345202416182, | |
| "learning_rate": 4.478501489995743e-06, | |
| "loss": 0.3893, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 29.01770114942529, | |
| "grad_norm": 0.09646335989236832, | |
| "learning_rate": 4.469987228607919e-06, | |
| "loss": 0.0004, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 29.018467432950192, | |
| "grad_norm": 0.002942258957773447, | |
| "learning_rate": 4.461472967220094e-06, | |
| "loss": 0.0424, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 29.019233716475096, | |
| "grad_norm": 0.36769601702690125, | |
| "learning_rate": 4.452958705832269e-06, | |
| "loss": 0.8931, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "grad_norm": 0.0013867371017113328, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.0085, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 29.02, | |
| "eval_accuracy": 0.6222222222222222, | |
| "eval_loss": 2.2413625717163086, | |
| "eval_runtime": 30.1507, | |
| "eval_samples_per_second": 1.493, | |
| "eval_steps_per_second": 1.493, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 30.000766283524904, | |
| "grad_norm": 0.0009211709257215261, | |
| "learning_rate": 4.43593018305662e-06, | |
| "loss": 0.7572, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 30.001532567049807, | |
| "grad_norm": 0.008097738027572632, | |
| "learning_rate": 4.427415921668796e-06, | |
| "loss": 0.5891, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 30.002298850574714, | |
| "grad_norm": 0.06377111375331879, | |
| "learning_rate": 4.418901660280971e-06, | |
| "loss": 0.0083, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 30.003065134099618, | |
| "grad_norm": 0.007994234561920166, | |
| "learning_rate": 4.410387398893146e-06, | |
| "loss": 0.0353, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 30.00383141762452, | |
| "grad_norm": 0.08883940428495407, | |
| "learning_rate": 4.401873137505322e-06, | |
| "loss": 0.0014, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 30.004597701149425, | |
| "grad_norm": 0.014959491789340973, | |
| "learning_rate": 4.3933588761174975e-06, | |
| "loss": 0.0041, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 30.00536398467433, | |
| "grad_norm": 0.0016909199766814709, | |
| "learning_rate": 4.3848446147296725e-06, | |
| "loss": 0.0019, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 30.006130268199232, | |
| "grad_norm": 0.178978830575943, | |
| "learning_rate": 4.3763303533418475e-06, | |
| "loss": 0.3294, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 30.00689655172414, | |
| "grad_norm": 0.2797927260398865, | |
| "learning_rate": 4.367816091954023e-06, | |
| "loss": 0.0003, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 30.007662835249043, | |
| "grad_norm": 0.016016194596886635, | |
| "learning_rate": 4.359301830566199e-06, | |
| "loss": 0.0002, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 30.008429118773947, | |
| "grad_norm": 0.04488089680671692, | |
| "learning_rate": 4.350787569178374e-06, | |
| "loss": 0.5719, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 30.00919540229885, | |
| "grad_norm": 0.0010439414763823152, | |
| "learning_rate": 4.342273307790549e-06, | |
| "loss": 0.5534, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 30.009961685823754, | |
| "grad_norm": 0.03713122010231018, | |
| "learning_rate": 4.333759046402725e-06, | |
| "loss": 0.19, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 30.010727969348657, | |
| "grad_norm": 0.048736151307821274, | |
| "learning_rate": 4.325244785014901e-06, | |
| "loss": 1.0971, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 30.011494252873565, | |
| "grad_norm": 637.2117919921875, | |
| "learning_rate": 4.316730523627076e-06, | |
| "loss": 0.0891, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 30.01226053639847, | |
| "grad_norm": 0.0008081755950115621, | |
| "learning_rate": 4.308216262239251e-06, | |
| "loss": 0.0005, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 30.013026819923372, | |
| "grad_norm": 0.021055513992905617, | |
| "learning_rate": 4.299702000851427e-06, | |
| "loss": 0.0001, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 30.013793103448275, | |
| "grad_norm": 0.04090442508459091, | |
| "learning_rate": 4.291187739463602e-06, | |
| "loss": 0.0042, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 30.01455938697318, | |
| "grad_norm": 0.0018981373868882656, | |
| "learning_rate": 4.2826734780757775e-06, | |
| "loss": 0.8071, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 30.015325670498083, | |
| "grad_norm": 0.0011597869452089071, | |
| "learning_rate": 4.2741592166879525e-06, | |
| "loss": 0.0004, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 30.01609195402299, | |
| "grad_norm": 0.04882820323109627, | |
| "learning_rate": 4.2656449553001275e-06, | |
| "loss": 0.0004, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 30.016858237547893, | |
| "grad_norm": 0.0015442122239619493, | |
| "learning_rate": 4.257130693912303e-06, | |
| "loss": 0.0021, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 30.017624521072797, | |
| "grad_norm": 0.003497874364256859, | |
| "learning_rate": 4.248616432524479e-06, | |
| "loss": 0.8706, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 30.0183908045977, | |
| "grad_norm": 0.012243164703249931, | |
| "learning_rate": 4.240102171136654e-06, | |
| "loss": 0.9448, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 30.019157088122604, | |
| "grad_norm": 0.00077412388054654, | |
| "learning_rate": 4.231587909748829e-06, | |
| "loss": 0.0003, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 30.01992337164751, | |
| "grad_norm": 0.01615084521472454, | |
| "learning_rate": 4.223073648361005e-06, | |
| "loss": 0.0004, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 30.02, | |
| "eval_accuracy": 0.5777777777777777, | |
| "eval_loss": 2.5363144874572754, | |
| "eval_runtime": 31.5231, | |
| "eval_samples_per_second": 1.428, | |
| "eval_steps_per_second": 1.428, | |
| "step": 8091 | |
| }, | |
| { | |
| "epoch": 31.000689655172415, | |
| "grad_norm": 0.8562591075897217, | |
| "learning_rate": 4.214559386973181e-06, | |
| "loss": 0.0006, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 31.00145593869732, | |
| "grad_norm": 0.0006197905167937279, | |
| "learning_rate": 4.206045125585356e-06, | |
| "loss": 0.0013, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 31.002222222222223, | |
| "grad_norm": 0.005202660337090492, | |
| "learning_rate": 4.197530864197531e-06, | |
| "loss": 0.0004, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 31.002988505747126, | |
| "grad_norm": 0.1686737984418869, | |
| "learning_rate": 4.189016602809707e-06, | |
| "loss": 0.0009, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 31.00375478927203, | |
| "grad_norm": 0.011581555008888245, | |
| "learning_rate": 4.180502341421882e-06, | |
| "loss": 0.0003, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 31.004521072796933, | |
| "grad_norm": 0.08714058250188828, | |
| "learning_rate": 4.1719880800340575e-06, | |
| "loss": 0.0018, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 31.00528735632184, | |
| "grad_norm": 0.0007840145844966173, | |
| "learning_rate": 4.1634738186462325e-06, | |
| "loss": 0.0188, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 31.006053639846744, | |
| "grad_norm": 0.001840552780777216, | |
| "learning_rate": 4.154959557258408e-06, | |
| "loss": 0.7423, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 31.006819923371648, | |
| "grad_norm": 0.018084323033690453, | |
| "learning_rate": 4.146445295870583e-06, | |
| "loss": 0.2636, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 31.00758620689655, | |
| "grad_norm": 0.005171739961951971, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.1457, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 31.008352490421455, | |
| "grad_norm": 0.002356515498831868, | |
| "learning_rate": 4.129416773094934e-06, | |
| "loss": 0.0005, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 31.00911877394636, | |
| "grad_norm": 0.009225085377693176, | |
| "learning_rate": 4.12090251170711e-06, | |
| "loss": 0.0002, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 31.009885057471266, | |
| "grad_norm": 0.020877808332443237, | |
| "learning_rate": 4.112388250319285e-06, | |
| "loss": 0.0001, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 31.01065134099617, | |
| "grad_norm": 0.0016082045622169971, | |
| "learning_rate": 4.103873988931461e-06, | |
| "loss": 0.0002, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 31.011417624521073, | |
| "grad_norm": 0.015390790067613125, | |
| "learning_rate": 4.095359727543636e-06, | |
| "loss": 0.5927, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 31.012183908045976, | |
| "grad_norm": 0.004826823249459267, | |
| "learning_rate": 4.086845466155812e-06, | |
| "loss": 1.3201, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 31.01295019157088, | |
| "grad_norm": 0.014385450631380081, | |
| "learning_rate": 4.078331204767987e-06, | |
| "loss": 1.0118, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 31.013716475095784, | |
| "grad_norm": 0.022623060271143913, | |
| "learning_rate": 4.0698169433801625e-06, | |
| "loss": 0.0006, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 31.01448275862069, | |
| "grad_norm": 0.041507456451654434, | |
| "learning_rate": 4.0613026819923375e-06, | |
| "loss": 0.0011, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 31.015249042145594, | |
| "grad_norm": 0.0038991193287074566, | |
| "learning_rate": 4.052788420604513e-06, | |
| "loss": 0.0004, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 31.016015325670498, | |
| "grad_norm": 0.1972963660955429, | |
| "learning_rate": 4.044274159216688e-06, | |
| "loss": 0.0006, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 31.0167816091954, | |
| "grad_norm": 0.0011404335964471102, | |
| "learning_rate": 4.035759897828863e-06, | |
| "loss": 0.1034, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 31.017547892720305, | |
| "grad_norm": 0.0016230113105848432, | |
| "learning_rate": 4.027245636441039e-06, | |
| "loss": 0.0007, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 31.018314176245212, | |
| "grad_norm": 0.013988662511110306, | |
| "learning_rate": 4.018731375053214e-06, | |
| "loss": 0.1782, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 31.019080459770116, | |
| "grad_norm": 0.0008143529412336648, | |
| "learning_rate": 4.01021711366539e-06, | |
| "loss": 0.0002, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 31.01984674329502, | |
| "grad_norm": 0.0058710016310215, | |
| "learning_rate": 4.001702852277565e-06, | |
| "loss": 0.7817, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 31.02, | |
| "eval_accuracy": 0.5777777777777777, | |
| "eval_loss": 2.8433194160461426, | |
| "eval_runtime": 31.5649, | |
| "eval_samples_per_second": 1.426, | |
| "eval_steps_per_second": 1.426, | |
| "step": 8352 | |
| }, | |
| { | |
| "epoch": 32.00061302681992, | |
| "grad_norm": 0.028439415618777275, | |
| "learning_rate": 3.993188590889741e-06, | |
| "loss": 0.0026, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 32.00137931034483, | |
| "grad_norm": 0.013596491888165474, | |
| "learning_rate": 3.984674329501916e-06, | |
| "loss": 0.6117, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 32.002145593869734, | |
| "grad_norm": 0.010585892014205456, | |
| "learning_rate": 3.976160068114092e-06, | |
| "loss": 0.0006, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 32.00291187739464, | |
| "grad_norm": 1.2051737308502197, | |
| "learning_rate": 3.967645806726267e-06, | |
| "loss": 0.0269, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 32.00367816091954, | |
| "grad_norm": 0.010174227878451347, | |
| "learning_rate": 3.9591315453384425e-06, | |
| "loss": 0.0006, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 32.004444444444445, | |
| "grad_norm": 0.021464044228196144, | |
| "learning_rate": 3.9506172839506175e-06, | |
| "loss": 0.0002, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 32.00521072796935, | |
| "grad_norm": 0.013936766423285007, | |
| "learning_rate": 3.942103022562793e-06, | |
| "loss": 0.0003, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 32.00597701149425, | |
| "grad_norm": 0.021049097180366516, | |
| "learning_rate": 3.933588761174968e-06, | |
| "loss": 0.755, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 32.006743295019156, | |
| "grad_norm": 0.01206669770181179, | |
| "learning_rate": 3.925074499787143e-06, | |
| "loss": 0.0002, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 32.00750957854406, | |
| "grad_norm": 0.00617943424731493, | |
| "learning_rate": 3.916560238399319e-06, | |
| "loss": 0.0005, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 32.00827586206896, | |
| "grad_norm": 0.017838966101408005, | |
| "learning_rate": 3.908045977011495e-06, | |
| "loss": 0.0002, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 32.00904214559387, | |
| "grad_norm": 0.0030338692013174295, | |
| "learning_rate": 3.89953171562367e-06, | |
| "loss": 0.0002, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 32.00980842911878, | |
| "grad_norm": 0.0018925362965092063, | |
| "learning_rate": 3.891017454235845e-06, | |
| "loss": 0.0001, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 32.01057471264368, | |
| "grad_norm": 0.015851356089115143, | |
| "learning_rate": 3.882503192848021e-06, | |
| "loss": 0.0001, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 32.011340996168585, | |
| "grad_norm": 0.019674142822623253, | |
| "learning_rate": 3.873988931460197e-06, | |
| "loss": 1.179, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 32.01210727969349, | |
| "grad_norm": 0.46721845865249634, | |
| "learning_rate": 3.865474670072372e-06, | |
| "loss": 0.1115, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 32.01287356321839, | |
| "grad_norm": 0.003226376138627529, | |
| "learning_rate": 3.856960408684547e-06, | |
| "loss": 0.0001, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 32.013639846743295, | |
| "grad_norm": 0.0005900561809539795, | |
| "learning_rate": 3.8484461472967225e-06, | |
| "loss": 0.1018, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 32.0144061302682, | |
| "grad_norm": 0.0007590750465169549, | |
| "learning_rate": 3.839931885908898e-06, | |
| "loss": 0.1272, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 32.0151724137931, | |
| "grad_norm": 12.848638534545898, | |
| "learning_rate": 3.831417624521073e-06, | |
| "loss": 0.0112, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 32.015938697318006, | |
| "grad_norm": 0.001263161888346076, | |
| "learning_rate": 3.822903363133248e-06, | |
| "loss": 0.0002, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 32.01670498084291, | |
| "grad_norm": 47.603431701660156, | |
| "learning_rate": 3.8143891017454237e-06, | |
| "loss": 1.0699, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 32.01747126436781, | |
| "grad_norm": 0.0034020866733044386, | |
| "learning_rate": 3.805874840357599e-06, | |
| "loss": 0.0256, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 32.01823754789272, | |
| "grad_norm": 0.031109264120459557, | |
| "learning_rate": 3.797360578969775e-06, | |
| "loss": 0.0441, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 32.01900383141763, | |
| "grad_norm": 0.0019936468452215195, | |
| "learning_rate": 3.78884631758195e-06, | |
| "loss": 0.0002, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 32.01977011494253, | |
| "grad_norm": 1.7935079336166382, | |
| "learning_rate": 3.7803320561941254e-06, | |
| "loss": 0.3487, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 32.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 2.637408494949341, | |
| "eval_runtime": 27.8621, | |
| "eval_samples_per_second": 1.615, | |
| "eval_steps_per_second": 1.615, | |
| "step": 8613 | |
| }, | |
| { | |
| "epoch": 33.00053639846743, | |
| "grad_norm": 0.2326209396123886, | |
| "learning_rate": 3.7718177948063004e-06, | |
| "loss": 0.2937, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 33.001302681992335, | |
| "grad_norm": 502.8989562988281, | |
| "learning_rate": 3.7633035334184762e-06, | |
| "loss": 1.1396, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 33.00206896551724, | |
| "grad_norm": 0.05000709369778633, | |
| "learning_rate": 3.7547892720306517e-06, | |
| "loss": 0.0988, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 33.00283524904214, | |
| "grad_norm": 0.01394747942686081, | |
| "learning_rate": 3.746275010642827e-06, | |
| "loss": 0.6976, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 33.00360153256705, | |
| "grad_norm": 0.0006377548561431468, | |
| "learning_rate": 3.737760749255002e-06, | |
| "loss": 0.0003, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 33.00436781609196, | |
| "grad_norm": 0.0005171005614101887, | |
| "learning_rate": 3.729246487867178e-06, | |
| "loss": 1.0241, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 33.00513409961686, | |
| "grad_norm": 1.197045087814331, | |
| "learning_rate": 3.7207322264793533e-06, | |
| "loss": 0.0082, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 33.005900383141764, | |
| "grad_norm": 616.587890625, | |
| "learning_rate": 3.7122179650915287e-06, | |
| "loss": 0.2803, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 33.00666666666667, | |
| "grad_norm": 0.011728745885193348, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.009, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 33.00743295019157, | |
| "grad_norm": 0.08511758595705032, | |
| "learning_rate": 3.6951894423158796e-06, | |
| "loss": 0.0031, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 33.008199233716475, | |
| "grad_norm": 0.0005724794464185834, | |
| "learning_rate": 3.686675180928055e-06, | |
| "loss": 0.0004, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 33.00896551724138, | |
| "grad_norm": 0.008813084103167057, | |
| "learning_rate": 3.67816091954023e-06, | |
| "loss": 0.0004, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 33.00973180076628, | |
| "grad_norm": 0.009110772982239723, | |
| "learning_rate": 3.6696466581524054e-06, | |
| "loss": 0.0006, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 33.010498084291186, | |
| "grad_norm": 0.001931531704030931, | |
| "learning_rate": 3.6611323967645812e-06, | |
| "loss": 0.0051, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 33.01126436781609, | |
| "grad_norm": 0.0009206047398038208, | |
| "learning_rate": 3.6526181353767567e-06, | |
| "loss": 0.0003, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 33.01203065134099, | |
| "grad_norm": 636.8836059570312, | |
| "learning_rate": 3.6441038739889317e-06, | |
| "loss": 0.6667, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 33.012796934865904, | |
| "grad_norm": 0.02224240079522133, | |
| "learning_rate": 3.635589612601107e-06, | |
| "loss": 0.7683, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 33.01356321839081, | |
| "grad_norm": 0.0008644562331028283, | |
| "learning_rate": 3.627075351213283e-06, | |
| "loss": 0.011, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 33.01432950191571, | |
| "grad_norm": 0.000989140709862113, | |
| "learning_rate": 3.618561089825458e-06, | |
| "loss": 0.4723, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 33.015095785440614, | |
| "grad_norm": 0.004592487588524818, | |
| "learning_rate": 3.6100468284376333e-06, | |
| "loss": 0.0001, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 33.01586206896552, | |
| "grad_norm": 0.0018623805372044444, | |
| "learning_rate": 3.6015325670498087e-06, | |
| "loss": 0.0007, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 33.01662835249042, | |
| "grad_norm": 0.46955156326293945, | |
| "learning_rate": 3.5930183056619837e-06, | |
| "loss": 0.0005, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 33.017394636015325, | |
| "grad_norm": 0.0008697701268829405, | |
| "learning_rate": 3.5845040442741596e-06, | |
| "loss": 0.01, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 33.01816091954023, | |
| "grad_norm": 0.01529706735163927, | |
| "learning_rate": 3.575989782886335e-06, | |
| "loss": 0.0004, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 33.01892720306513, | |
| "grad_norm": 498.3216857910156, | |
| "learning_rate": 3.56747552149851e-06, | |
| "loss": 0.3644, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 33.019693486590036, | |
| "grad_norm": 0.0025238273665308952, | |
| "learning_rate": 3.5589612601106854e-06, | |
| "loss": 0.0014, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "eval_accuracy": 0.5777777777777777, | |
| "eval_loss": 3.0313196182250977, | |
| "eval_runtime": 27.976, | |
| "eval_samples_per_second": 1.609, | |
| "eval_steps_per_second": 1.609, | |
| "step": 8874 | |
| }, | |
| { | |
| "epoch": 34.000459770114944, | |
| "grad_norm": 141.0464630126953, | |
| "learning_rate": 3.5504469987228612e-06, | |
| "loss": 0.0134, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 34.00122605363985, | |
| "grad_norm": 0.021763743832707405, | |
| "learning_rate": 3.5419327373350367e-06, | |
| "loss": 0.0042, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 34.00199233716475, | |
| "grad_norm": 0.008113621734082699, | |
| "learning_rate": 3.5334184759472117e-06, | |
| "loss": 0.8664, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 34.002758620689654, | |
| "grad_norm": 0.009396526962518692, | |
| "learning_rate": 3.524904214559387e-06, | |
| "loss": 0.0002, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 34.00352490421456, | |
| "grad_norm": 0.0757775753736496, | |
| "learning_rate": 3.516389953171563e-06, | |
| "loss": 0.0001, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 34.00429118773946, | |
| "grad_norm": 0.011029227636754513, | |
| "learning_rate": 3.507875691783738e-06, | |
| "loss": 0.5062, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 34.005057471264365, | |
| "grad_norm": 0.015409487299621105, | |
| "learning_rate": 3.4993614303959133e-06, | |
| "loss": 0.0002, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 34.00582375478927, | |
| "grad_norm": 0.026873435825109482, | |
| "learning_rate": 3.4908471690080887e-06, | |
| "loss": 0.0003, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 34.00659003831418, | |
| "grad_norm": 0.0005174085381440818, | |
| "learning_rate": 3.4823329076202646e-06, | |
| "loss": 0.0002, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 34.00735632183908, | |
| "grad_norm": 0.008040793240070343, | |
| "learning_rate": 3.4738186462324396e-06, | |
| "loss": 0.0004, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 34.00812260536399, | |
| "grad_norm": 0.001422680914402008, | |
| "learning_rate": 3.465304384844615e-06, | |
| "loss": 0.0336, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 34.00888888888889, | |
| "grad_norm": 0.002646015491336584, | |
| "learning_rate": 3.4567901234567904e-06, | |
| "loss": 0.0003, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 34.009655172413794, | |
| "grad_norm": 0.0038285779301077127, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.0052, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 34.0104214559387, | |
| "grad_norm": 0.0006650602445006371, | |
| "learning_rate": 3.4397616006811412e-06, | |
| "loss": 0.0013, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 34.0111877394636, | |
| "grad_norm": 0.007349758874624968, | |
| "learning_rate": 3.4312473392933167e-06, | |
| "loss": 0.0002, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 34.011954022988505, | |
| "grad_norm": 0.0035567977465689182, | |
| "learning_rate": 3.4227330779054917e-06, | |
| "loss": 0.0049, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 34.01272030651341, | |
| "grad_norm": 0.008664336055517197, | |
| "learning_rate": 3.4142188165176675e-06, | |
| "loss": 0.6437, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 34.01348659003831, | |
| "grad_norm": 0.0005684501957148314, | |
| "learning_rate": 3.405704555129843e-06, | |
| "loss": 0.0265, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 34.014252873563215, | |
| "grad_norm": 399.8719177246094, | |
| "learning_rate": 3.3971902937420183e-06, | |
| "loss": 0.0279, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 34.01501915708812, | |
| "grad_norm": 1463.30322265625, | |
| "learning_rate": 3.3886760323541933e-06, | |
| "loss": 0.6656, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 34.01578544061303, | |
| "grad_norm": 0.012521703727543354, | |
| "learning_rate": 3.380161770966369e-06, | |
| "loss": 0.0337, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 34.01655172413793, | |
| "grad_norm": 0.014144524931907654, | |
| "learning_rate": 3.3716475095785446e-06, | |
| "loss": 0.0001, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 34.01731800766284, | |
| "grad_norm": 1.548913836479187, | |
| "learning_rate": 3.3631332481907196e-06, | |
| "loss": 0.3669, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 34.01808429118774, | |
| "grad_norm": 0.0010734919924288988, | |
| "learning_rate": 3.354618986802895e-06, | |
| "loss": 0.7933, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 34.018850574712644, | |
| "grad_norm": 0.0018491506343707442, | |
| "learning_rate": 3.3461047254150704e-06, | |
| "loss": 0.0002, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 34.01961685823755, | |
| "grad_norm": 5.716326713562012, | |
| "learning_rate": 3.3375904640272463e-06, | |
| "loss": 0.0009, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 34.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.6186773777008057, | |
| "eval_runtime": 27.9487, | |
| "eval_samples_per_second": 1.61, | |
| "eval_steps_per_second": 1.61, | |
| "step": 9135 | |
| }, | |
| { | |
| "epoch": 35.000383141762455, | |
| "grad_norm": 0.005345072131603956, | |
| "learning_rate": 3.3290762026394212e-06, | |
| "loss": 0.2087, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 35.00114942528736, | |
| "grad_norm": 0.007061213720589876, | |
| "learning_rate": 3.3205619412515967e-06, | |
| "loss": 0.0001, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 35.00191570881226, | |
| "grad_norm": 0.002428262960165739, | |
| "learning_rate": 3.3120476798637717e-06, | |
| "loss": 0.028, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 35.002681992337166, | |
| "grad_norm": 0.0005560162826441228, | |
| "learning_rate": 3.3035334184759475e-06, | |
| "loss": 0.0002, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 35.00344827586207, | |
| "grad_norm": 106.02903747558594, | |
| "learning_rate": 3.295019157088123e-06, | |
| "loss": 0.0078, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 35.00421455938697, | |
| "grad_norm": 0.0006189548294059932, | |
| "learning_rate": 3.2865048957002983e-06, | |
| "loss": 0.0002, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 35.00498084291188, | |
| "grad_norm": 0.0013160452945157886, | |
| "learning_rate": 3.2779906343124733e-06, | |
| "loss": 0.0002, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 35.00574712643678, | |
| "grad_norm": 0.0006706579006277025, | |
| "learning_rate": 3.269476372924649e-06, | |
| "loss": 0.6169, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 35.006513409961684, | |
| "grad_norm": 86.67324829101562, | |
| "learning_rate": 3.2609621115368246e-06, | |
| "loss": 0.8886, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 35.00727969348659, | |
| "grad_norm": 0.008850634098052979, | |
| "learning_rate": 3.2524478501489996e-06, | |
| "loss": 0.0026, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 35.00804597701149, | |
| "grad_norm": 0.007037854287773371, | |
| "learning_rate": 3.243933588761175e-06, | |
| "loss": 0.0014, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 35.008812260536395, | |
| "grad_norm": 0.05180385708808899, | |
| "learning_rate": 3.235419327373351e-06, | |
| "loss": 0.0636, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 35.009578544061306, | |
| "grad_norm": 0.036820054054260254, | |
| "learning_rate": 3.2269050659855262e-06, | |
| "loss": 0.9398, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 35.01034482758621, | |
| "grad_norm": 0.019011739641427994, | |
| "learning_rate": 3.2183908045977012e-06, | |
| "loss": 0.348, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 35.01111111111111, | |
| "grad_norm": 0.0077316779643297195, | |
| "learning_rate": 3.2098765432098767e-06, | |
| "loss": 0.5693, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 35.01187739463602, | |
| "grad_norm": 0.008710850961506367, | |
| "learning_rate": 3.2013622818220525e-06, | |
| "loss": 0.5292, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 35.01264367816092, | |
| "grad_norm": 0.018290938809514046, | |
| "learning_rate": 3.1928480204342275e-06, | |
| "loss": 0.0002, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 35.013409961685824, | |
| "grad_norm": 0.0012147327652201056, | |
| "learning_rate": 3.184333759046403e-06, | |
| "loss": 0.0019, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 35.01417624521073, | |
| "grad_norm": 0.017071552574634552, | |
| "learning_rate": 3.1758194976585783e-06, | |
| "loss": 0.0006, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 35.01494252873563, | |
| "grad_norm": 0.0005288755637593567, | |
| "learning_rate": 3.167305236270754e-06, | |
| "loss": 0.0002, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 35.015708812260534, | |
| "grad_norm": 0.0023952096235007048, | |
| "learning_rate": 3.158790974882929e-06, | |
| "loss": 0.0001, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 35.01647509578544, | |
| "grad_norm": 0.0009895041584968567, | |
| "learning_rate": 3.1502767134951046e-06, | |
| "loss": 0.0001, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 35.01724137931034, | |
| "grad_norm": 0.003969644661992788, | |
| "learning_rate": 3.14176245210728e-06, | |
| "loss": 0.1839, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 35.01800766283525, | |
| "grad_norm": 0.007862897589802742, | |
| "learning_rate": 3.133248190719455e-06, | |
| "loss": 0.0001, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 35.018773946360156, | |
| "grad_norm": 0.0016002559568732977, | |
| "learning_rate": 3.124733929331631e-06, | |
| "loss": 0.0001, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 35.01954022988506, | |
| "grad_norm": 0.000823653768748045, | |
| "learning_rate": 3.1162196679438062e-06, | |
| "loss": 0.014, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 35.02, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.109400510787964, | |
| "eval_runtime": 29.0543, | |
| "eval_samples_per_second": 1.549, | |
| "eval_steps_per_second": 1.549, | |
| "step": 9396 | |
| }, | |
| { | |
| "epoch": 36.00030651340996, | |
| "grad_norm": 0.01986868306994438, | |
| "learning_rate": 3.1077054065559812e-06, | |
| "loss": 0.0002, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 36.001072796934864, | |
| "grad_norm": 863.7614135742188, | |
| "learning_rate": 3.0991911451681567e-06, | |
| "loss": 0.6115, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 36.00183908045977, | |
| "grad_norm": 0.01149124838411808, | |
| "learning_rate": 3.0906768837803325e-06, | |
| "loss": 0.8093, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 36.00260536398467, | |
| "grad_norm": 0.001022035488858819, | |
| "learning_rate": 3.082162622392508e-06, | |
| "loss": 1.536, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 36.00337164750958, | |
| "grad_norm": 0.029342470690608025, | |
| "learning_rate": 3.073648361004683e-06, | |
| "loss": 0.0008, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 36.004137931034485, | |
| "grad_norm": 0.000798724009655416, | |
| "learning_rate": 3.0651340996168583e-06, | |
| "loss": 0.7052, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 36.00490421455939, | |
| "grad_norm": 0.0022243347484618425, | |
| "learning_rate": 3.056619838229034e-06, | |
| "loss": 0.0001, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 36.00567049808429, | |
| "grad_norm": 0.0008435220806859434, | |
| "learning_rate": 3.048105576841209e-06, | |
| "loss": 0.6712, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 36.006436781609196, | |
| "grad_norm": 0.003984359093010426, | |
| "learning_rate": 3.0395913154533846e-06, | |
| "loss": 0.0002, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 36.0072030651341, | |
| "grad_norm": 0.015606824308633804, | |
| "learning_rate": 3.03107705406556e-06, | |
| "loss": 0.0002, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 36.007969348659, | |
| "grad_norm": 305.4211730957031, | |
| "learning_rate": 3.022562792677736e-06, | |
| "loss": 0.5422, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 36.00873563218391, | |
| "grad_norm": 0.0029582574497908354, | |
| "learning_rate": 3.014048531289911e-06, | |
| "loss": 0.0001, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 36.00950191570881, | |
| "grad_norm": 0.009127511642873287, | |
| "learning_rate": 3.0055342699020862e-06, | |
| "loss": 0.0003, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 36.010268199233714, | |
| "grad_norm": 0.044698331505060196, | |
| "learning_rate": 2.9970200085142612e-06, | |
| "loss": 0.0012, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 36.01103448275862, | |
| "grad_norm": 1064.9903564453125, | |
| "learning_rate": 2.988505747126437e-06, | |
| "loss": 0.5409, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 36.01180076628353, | |
| "grad_norm": 0.0006466146442107856, | |
| "learning_rate": 2.9799914857386125e-06, | |
| "loss": 0.912, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 36.01256704980843, | |
| "grad_norm": 0.002113508293405175, | |
| "learning_rate": 2.971477224350788e-06, | |
| "loss": 0.0237, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 36.013333333333335, | |
| "grad_norm": 0.0005975261447019875, | |
| "learning_rate": 2.962962962962963e-06, | |
| "loss": 0.0001, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 36.01409961685824, | |
| "grad_norm": 0.0025878562591969967, | |
| "learning_rate": 2.9544487015751387e-06, | |
| "loss": 0.0001, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 36.01486590038314, | |
| "grad_norm": 0.12519130110740662, | |
| "learning_rate": 2.945934440187314e-06, | |
| "loss": 0.3203, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 36.015632183908046, | |
| "grad_norm": 0.004954866599291563, | |
| "learning_rate": 2.9374201787994896e-06, | |
| "loss": 0.0024, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 36.01639846743295, | |
| "grad_norm": 0.7206712961196899, | |
| "learning_rate": 2.9289059174116646e-06, | |
| "loss": 0.0503, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 36.01716475095785, | |
| "grad_norm": 0.00043419323628768325, | |
| "learning_rate": 2.92039165602384e-06, | |
| "loss": 0.0001, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 36.01793103448276, | |
| "grad_norm": 0.006202781107276678, | |
| "learning_rate": 2.911877394636016e-06, | |
| "loss": 0.0002, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 36.01869731800766, | |
| "grad_norm": 99.42586517333984, | |
| "learning_rate": 2.903363133248191e-06, | |
| "loss": 0.01, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 36.019463601532564, | |
| "grad_norm": 0.0005115497042424977, | |
| "learning_rate": 2.8948488718603662e-06, | |
| "loss": 0.512, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 36.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.1110174655914307, | |
| "eval_runtime": 27.8759, | |
| "eval_samples_per_second": 1.614, | |
| "eval_steps_per_second": 1.614, | |
| "step": 9657 | |
| }, | |
| { | |
| "epoch": 37.00022988505747, | |
| "grad_norm": 0.0038163657300174236, | |
| "learning_rate": 2.8863346104725417e-06, | |
| "loss": 0.8013, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 37.000996168582375, | |
| "grad_norm": 0.00041584885912016034, | |
| "learning_rate": 2.8778203490847175e-06, | |
| "loss": 0.0001, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 37.00176245210728, | |
| "grad_norm": 0.0018008637707680464, | |
| "learning_rate": 2.8693060876968925e-06, | |
| "loss": 0.7253, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 37.00252873563218, | |
| "grad_norm": 0.0014151118230074644, | |
| "learning_rate": 2.860791826309068e-06, | |
| "loss": 0.0011, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 37.003295019157086, | |
| "grad_norm": 0.07184292376041412, | |
| "learning_rate": 2.852277564921243e-06, | |
| "loss": 0.0007, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 37.00406130268199, | |
| "grad_norm": 0.0006791798514313996, | |
| "learning_rate": 2.8437633035334187e-06, | |
| "loss": 0.0241, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 37.00482758620689, | |
| "grad_norm": 0.0007547545828856528, | |
| "learning_rate": 2.835249042145594e-06, | |
| "loss": 0.0001, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 37.005593869731804, | |
| "grad_norm": 0.00040744239231571555, | |
| "learning_rate": 2.8267347807577696e-06, | |
| "loss": 0.3058, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 37.00636015325671, | |
| "grad_norm": 0.009197317063808441, | |
| "learning_rate": 2.8182205193699446e-06, | |
| "loss": 0.7546, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 37.00712643678161, | |
| "grad_norm": 0.006815429776906967, | |
| "learning_rate": 2.8097062579821204e-06, | |
| "loss": 0.0006, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 37.007892720306515, | |
| "grad_norm": 0.0010986275738105178, | |
| "learning_rate": 2.801191996594296e-06, | |
| "loss": 0.0002, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 37.00865900383142, | |
| "grad_norm": 0.0017043360276147723, | |
| "learning_rate": 2.792677735206471e-06, | |
| "loss": 0.0001, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 37.00942528735632, | |
| "grad_norm": 25.789806365966797, | |
| "learning_rate": 2.7841634738186462e-06, | |
| "loss": 0.0028, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 37.010191570881226, | |
| "grad_norm": 0.02588670514523983, | |
| "learning_rate": 2.775649212430822e-06, | |
| "loss": 0.0001, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 37.01095785440613, | |
| "grad_norm": 0.4962689280509949, | |
| "learning_rate": 2.7671349510429975e-06, | |
| "loss": 0.0002, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 37.01172413793103, | |
| "grad_norm": 0.0005121955182403326, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.1431, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 37.01249042145594, | |
| "grad_norm": 0.0004625523288268596, | |
| "learning_rate": 2.750106428267348e-06, | |
| "loss": 0.657, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 37.01325670498084, | |
| "grad_norm": 0.0032494685146957636, | |
| "learning_rate": 2.7415921668795238e-06, | |
| "loss": 0.5249, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 37.014022988505744, | |
| "grad_norm": 0.003939610905945301, | |
| "learning_rate": 2.7330779054916987e-06, | |
| "loss": 0.0002, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 37.014789272030654, | |
| "grad_norm": 0.0006820214912295341, | |
| "learning_rate": 2.724563644103874e-06, | |
| "loss": 0.0001, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 37.01555555555556, | |
| "grad_norm": 5.725392818450928, | |
| "learning_rate": 2.7160493827160496e-06, | |
| "loss": 0.5215, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 37.01632183908046, | |
| "grad_norm": 0.0006258029025048018, | |
| "learning_rate": 2.7075351213282254e-06, | |
| "loss": 0.0006, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 37.017088122605365, | |
| "grad_norm": 1426.855224609375, | |
| "learning_rate": 2.6990208599404004e-06, | |
| "loss": 0.9009, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 37.01785440613027, | |
| "grad_norm": 890.5499267578125, | |
| "learning_rate": 2.690506598552576e-06, | |
| "loss": 0.1271, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 37.01862068965517, | |
| "grad_norm": 0.04052882269024849, | |
| "learning_rate": 2.6819923371647512e-06, | |
| "loss": 0.7701, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 37.019386973180076, | |
| "grad_norm": 0.001093729748390615, | |
| "learning_rate": 2.6734780757769262e-06, | |
| "loss": 0.0003, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 37.02, | |
| "eval_accuracy": 0.5777777777777777, | |
| "eval_loss": 3.0441110134124756, | |
| "eval_runtime": 28.8351, | |
| "eval_samples_per_second": 1.561, | |
| "eval_steps_per_second": 1.561, | |
| "step": 9918 | |
| }, | |
| { | |
| "epoch": 38.000153256704984, | |
| "grad_norm": 1204.0167236328125, | |
| "learning_rate": 2.664963814389102e-06, | |
| "loss": 0.6456, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 38.00091954022989, | |
| "grad_norm": 0.014917056076228619, | |
| "learning_rate": 2.6564495530012775e-06, | |
| "loss": 0.0003, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 38.00168582375479, | |
| "grad_norm": 0.005296202842146158, | |
| "learning_rate": 2.6479352916134525e-06, | |
| "loss": 0.2374, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 38.002452107279694, | |
| "grad_norm": 0.003745126770809293, | |
| "learning_rate": 2.639421030225628e-06, | |
| "loss": 0.0003, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 38.0032183908046, | |
| "grad_norm": 0.003593859262764454, | |
| "learning_rate": 2.6309067688378037e-06, | |
| "loss": 0.0001, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 38.0039846743295, | |
| "grad_norm": 929.7574462890625, | |
| "learning_rate": 2.622392507449979e-06, | |
| "loss": 0.5339, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 38.004750957854405, | |
| "grad_norm": 0.641494631767273, | |
| "learning_rate": 2.613878246062154e-06, | |
| "loss": 0.0003, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 38.00551724137931, | |
| "grad_norm": 0.0007532311719842255, | |
| "learning_rate": 2.6053639846743296e-06, | |
| "loss": 0.8401, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 38.00628352490421, | |
| "grad_norm": 0.0004510098951868713, | |
| "learning_rate": 2.5968497232865054e-06, | |
| "loss": 0.0002, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 38.007049808429116, | |
| "grad_norm": 0.0011898913653567433, | |
| "learning_rate": 2.5883354618986804e-06, | |
| "loss": 0.0305, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 38.00781609195402, | |
| "grad_norm": 0.11280697584152222, | |
| "learning_rate": 2.579821200510856e-06, | |
| "loss": 0.0004, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 38.00858237547893, | |
| "grad_norm": 0.0005606805207207799, | |
| "learning_rate": 2.5713069391230312e-06, | |
| "loss": 0.0002, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 38.009348659003834, | |
| "grad_norm": 11.6032075881958, | |
| "learning_rate": 2.562792677735207e-06, | |
| "loss": 0.0013, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 38.01011494252874, | |
| "grad_norm": 0.0013480938505381346, | |
| "learning_rate": 2.554278416347382e-06, | |
| "loss": 0.0001, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 38.01088122605364, | |
| "grad_norm": 0.00067657366162166, | |
| "learning_rate": 2.5457641549595575e-06, | |
| "loss": 0.0003, | |
| "step": 10060 | |
| }, | |
| { | |
| "epoch": 38.011647509578545, | |
| "grad_norm": 0.029493281617760658, | |
| "learning_rate": 2.5372498935717325e-06, | |
| "loss": 0.0003, | |
| "step": 10070 | |
| }, | |
| { | |
| "epoch": 38.01241379310345, | |
| "grad_norm": 0.05872733145952225, | |
| "learning_rate": 2.5287356321839083e-06, | |
| "loss": 0.0001, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 38.01318007662835, | |
| "grad_norm": 0.017297696322202682, | |
| "learning_rate": 2.5202213707960837e-06, | |
| "loss": 0.0002, | |
| "step": 10090 | |
| }, | |
| { | |
| "epoch": 38.013946360153255, | |
| "grad_norm": 0.001809093402698636, | |
| "learning_rate": 2.511707109408259e-06, | |
| "loss": 0.0, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 38.01471264367816, | |
| "grad_norm": 0.00425736466422677, | |
| "learning_rate": 2.503192848020434e-06, | |
| "loss": 0.0176, | |
| "step": 10110 | |
| }, | |
| { | |
| "epoch": 38.01547892720306, | |
| "grad_norm": 0.0014015929773449898, | |
| "learning_rate": 2.49467858663261e-06, | |
| "loss": 0.0002, | |
| "step": 10120 | |
| }, | |
| { | |
| "epoch": 38.016245210727966, | |
| "grad_norm": 417.6300048828125, | |
| "learning_rate": 2.4861643252447854e-06, | |
| "loss": 0.7875, | |
| "step": 10130 | |
| }, | |
| { | |
| "epoch": 38.01701149425288, | |
| "grad_norm": 0.012668310664594173, | |
| "learning_rate": 2.4776500638569604e-06, | |
| "loss": 0.0316, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 38.01777777777778, | |
| "grad_norm": 0.0010990687878802419, | |
| "learning_rate": 2.469135802469136e-06, | |
| "loss": 0.6725, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 38.018544061302684, | |
| "grad_norm": 0.0004859846376348287, | |
| "learning_rate": 2.4606215410813112e-06, | |
| "loss": 0.0001, | |
| "step": 10160 | |
| }, | |
| { | |
| "epoch": 38.01931034482759, | |
| "grad_norm": 0.018242767080664635, | |
| "learning_rate": 2.4521072796934867e-06, | |
| "loss": 0.0001, | |
| "step": 10170 | |
| }, | |
| { | |
| "epoch": 38.02, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.4423201084136963, | |
| "eval_runtime": 27.8017, | |
| "eval_samples_per_second": 1.619, | |
| "eval_steps_per_second": 1.619, | |
| "step": 10179 | |
| }, | |
| { | |
| "epoch": 39.00007662835249, | |
| "grad_norm": 0.0008480402757413685, | |
| "learning_rate": 2.443593018305662e-06, | |
| "loss": 0.0001, | |
| "step": 10180 | |
| }, | |
| { | |
| "epoch": 39.00084291187739, | |
| "grad_norm": 0.0028732302598655224, | |
| "learning_rate": 2.4350787569178375e-06, | |
| "loss": 0.0001, | |
| "step": 10190 | |
| }, | |
| { | |
| "epoch": 39.001609195402295, | |
| "grad_norm": 448.4722900390625, | |
| "learning_rate": 2.426564495530013e-06, | |
| "loss": 0.6154, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 39.002375478927206, | |
| "grad_norm": 0.007314682938158512, | |
| "learning_rate": 2.4180502341421883e-06, | |
| "loss": 0.0001, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 39.00314176245211, | |
| "grad_norm": 0.0008429197478108108, | |
| "learning_rate": 2.4095359727543637e-06, | |
| "loss": 0.4188, | |
| "step": 10220 | |
| }, | |
| { | |
| "epoch": 39.00390804597701, | |
| "grad_norm": 789.7183227539062, | |
| "learning_rate": 2.401021711366539e-06, | |
| "loss": 0.3612, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 39.00467432950192, | |
| "grad_norm": 0.0009932577377185225, | |
| "learning_rate": 2.3925074499787146e-06, | |
| "loss": 0.8489, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 39.00544061302682, | |
| "grad_norm": 0.0006202096701599658, | |
| "learning_rate": 2.38399318859089e-06, | |
| "loss": 0.0001, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 39.006206896551724, | |
| "grad_norm": 0.006974675226956606, | |
| "learning_rate": 2.3754789272030654e-06, | |
| "loss": 0.0121, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 39.00697318007663, | |
| "grad_norm": 0.01899573765695095, | |
| "learning_rate": 2.366964665815241e-06, | |
| "loss": 0.8677, | |
| "step": 10270 | |
| }, | |
| { | |
| "epoch": 39.00773946360153, | |
| "grad_norm": 0.005227435380220413, | |
| "learning_rate": 2.3584504044274162e-06, | |
| "loss": 0.0004, | |
| "step": 10280 | |
| }, | |
| { | |
| "epoch": 39.008505747126435, | |
| "grad_norm": 0.017451848834753036, | |
| "learning_rate": 2.3499361430395912e-06, | |
| "loss": 0.0001, | |
| "step": 10290 | |
| }, | |
| { | |
| "epoch": 39.00927203065134, | |
| "grad_norm": 0.00589205976575613, | |
| "learning_rate": 2.341421881651767e-06, | |
| "loss": 0.6305, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 39.01003831417624, | |
| "grad_norm": 0.0006803887663409114, | |
| "learning_rate": 2.332907620263942e-06, | |
| "loss": 0.0001, | |
| "step": 10310 | |
| }, | |
| { | |
| "epoch": 39.01080459770115, | |
| "grad_norm": 0.007072710897773504, | |
| "learning_rate": 2.324393358876118e-06, | |
| "loss": 0.3078, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 39.011570881226056, | |
| "grad_norm": 0.0009751072502695024, | |
| "learning_rate": 2.315879097488293e-06, | |
| "loss": 0.0001, | |
| "step": 10330 | |
| }, | |
| { | |
| "epoch": 39.01233716475096, | |
| "grad_norm": 0.00029019732028245926, | |
| "learning_rate": 2.3073648361004688e-06, | |
| "loss": 0.0004, | |
| "step": 10340 | |
| }, | |
| { | |
| "epoch": 39.013103448275864, | |
| "grad_norm": 0.008286180905997753, | |
| "learning_rate": 2.2988505747126437e-06, | |
| "loss": 0.0003, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 39.01386973180077, | |
| "grad_norm": 0.007536804303526878, | |
| "learning_rate": 2.290336313324819e-06, | |
| "loss": 0.0752, | |
| "step": 10360 | |
| }, | |
| { | |
| "epoch": 39.01463601532567, | |
| "grad_norm": 0.0009000327554531395, | |
| "learning_rate": 2.2818220519369946e-06, | |
| "loss": 0.0002, | |
| "step": 10370 | |
| }, | |
| { | |
| "epoch": 39.015402298850574, | |
| "grad_norm": 0.0007068035774864256, | |
| "learning_rate": 2.27330779054917e-06, | |
| "loss": 0.0002, | |
| "step": 10380 | |
| }, | |
| { | |
| "epoch": 39.01616858237548, | |
| "grad_norm": 0.0009844456799328327, | |
| "learning_rate": 2.2647935291613454e-06, | |
| "loss": 0.0001, | |
| "step": 10390 | |
| }, | |
| { | |
| "epoch": 39.01693486590038, | |
| "grad_norm": 0.7614496946334839, | |
| "learning_rate": 2.256279267773521e-06, | |
| "loss": 0.0005, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 39.017701149425285, | |
| "grad_norm": 0.010708021931350231, | |
| "learning_rate": 2.2477650063856962e-06, | |
| "loss": 0.7864, | |
| "step": 10410 | |
| }, | |
| { | |
| "epoch": 39.01846743295019, | |
| "grad_norm": 0.0005570295616053045, | |
| "learning_rate": 2.2392507449978717e-06, | |
| "loss": 0.0001, | |
| "step": 10420 | |
| }, | |
| { | |
| "epoch": 39.01923371647509, | |
| "grad_norm": 0.0003596442984417081, | |
| "learning_rate": 2.230736483610047e-06, | |
| "loss": 0.0002, | |
| "step": 10430 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "grad_norm": 0.011120975017547607, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.0009, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 39.02, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.353762149810791, | |
| "eval_runtime": 28.4953, | |
| "eval_samples_per_second": 1.579, | |
| "eval_steps_per_second": 1.579, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 40.000766283524904, | |
| "grad_norm": 0.0003600206400733441, | |
| "learning_rate": 2.213707960834398e-06, | |
| "loss": 0.0001, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 40.00153256704981, | |
| "grad_norm": 0.00020263415353838354, | |
| "learning_rate": 2.205193699446573e-06, | |
| "loss": 0.0001, | |
| "step": 10460 | |
| }, | |
| { | |
| "epoch": 40.00229885057471, | |
| "grad_norm": 0.0002682032936718315, | |
| "learning_rate": 2.1966794380587487e-06, | |
| "loss": 0.0001, | |
| "step": 10470 | |
| }, | |
| { | |
| "epoch": 40.003065134099614, | |
| "grad_norm": 0.0013818013248965144, | |
| "learning_rate": 2.1881651766709237e-06, | |
| "loss": 0.0005, | |
| "step": 10480 | |
| }, | |
| { | |
| "epoch": 40.00383141762452, | |
| "grad_norm": 0.0014766278909519315, | |
| "learning_rate": 2.1796509152830996e-06, | |
| "loss": 0.0001, | |
| "step": 10490 | |
| }, | |
| { | |
| "epoch": 40.00459770114943, | |
| "grad_norm": 0.0006407822365872562, | |
| "learning_rate": 2.1711366538952746e-06, | |
| "loss": 0.0002, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 40.00536398467433, | |
| "grad_norm": 0.003510398091748357, | |
| "learning_rate": 2.1626223925074504e-06, | |
| "loss": 0.8606, | |
| "step": 10510 | |
| }, | |
| { | |
| "epoch": 40.006130268199236, | |
| "grad_norm": 0.015534725971519947, | |
| "learning_rate": 2.1541081311196254e-06, | |
| "loss": 0.7498, | |
| "step": 10520 | |
| }, | |
| { | |
| "epoch": 40.00689655172414, | |
| "grad_norm": 0.012120231054723263, | |
| "learning_rate": 2.145593869731801e-06, | |
| "loss": 0.0001, | |
| "step": 10530 | |
| }, | |
| { | |
| "epoch": 40.00766283524904, | |
| "grad_norm": 0.3469342589378357, | |
| "learning_rate": 2.1370796083439762e-06, | |
| "loss": 0.0004, | |
| "step": 10540 | |
| }, | |
| { | |
| "epoch": 40.00842911877395, | |
| "grad_norm": 0.00825453456491232, | |
| "learning_rate": 2.1285653469561517e-06, | |
| "loss": 0.0003, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 40.00919540229885, | |
| "grad_norm": 0.00035154339275322855, | |
| "learning_rate": 2.120051085568327e-06, | |
| "loss": 0.0002, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 40.009961685823754, | |
| "grad_norm": 0.0005046504666097462, | |
| "learning_rate": 2.1115368241805025e-06, | |
| "loss": 0.0002, | |
| "step": 10570 | |
| }, | |
| { | |
| "epoch": 40.01072796934866, | |
| "grad_norm": 0.001045138924382627, | |
| "learning_rate": 2.103022562792678e-06, | |
| "loss": 0.0006, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 40.01149425287356, | |
| "grad_norm": 0.004084183368831873, | |
| "learning_rate": 2.0945083014048533e-06, | |
| "loss": 0.567, | |
| "step": 10590 | |
| }, | |
| { | |
| "epoch": 40.012260536398465, | |
| "grad_norm": 0.000454683176940307, | |
| "learning_rate": 2.0859940400170287e-06, | |
| "loss": 0.0002, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 40.01302681992337, | |
| "grad_norm": 0.024498263373970985, | |
| "learning_rate": 2.077479778629204e-06, | |
| "loss": 0.0001, | |
| "step": 10610 | |
| }, | |
| { | |
| "epoch": 40.01379310344828, | |
| "grad_norm": 0.0032995757646858692, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.0098, | |
| "step": 10620 | |
| }, | |
| { | |
| "epoch": 40.01455938697318, | |
| "grad_norm": 0.2157309204339981, | |
| "learning_rate": 2.060451255853555e-06, | |
| "loss": 0.0035, | |
| "step": 10630 | |
| }, | |
| { | |
| "epoch": 40.015325670498086, | |
| "grad_norm": 0.00028585558175109327, | |
| "learning_rate": 2.0519369944657304e-06, | |
| "loss": 0.0001, | |
| "step": 10640 | |
| }, | |
| { | |
| "epoch": 40.01609195402299, | |
| "grad_norm": 0.00040348677430301905, | |
| "learning_rate": 2.043422733077906e-06, | |
| "loss": 0.0144, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 40.01685823754789, | |
| "grad_norm": 0.001562313991598785, | |
| "learning_rate": 2.0349084716900813e-06, | |
| "loss": 0.0001, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 40.0176245210728, | |
| "grad_norm": 159.2068328857422, | |
| "learning_rate": 2.0263942103022567e-06, | |
| "loss": 0.5783, | |
| "step": 10670 | |
| }, | |
| { | |
| "epoch": 40.0183908045977, | |
| "grad_norm": 0.0005799325881525874, | |
| "learning_rate": 2.0178799489144317e-06, | |
| "loss": 1.0221, | |
| "step": 10680 | |
| }, | |
| { | |
| "epoch": 40.019157088122604, | |
| "grad_norm": 0.0030735430773347616, | |
| "learning_rate": 2.009365687526607e-06, | |
| "loss": 0.0003, | |
| "step": 10690 | |
| }, | |
| { | |
| "epoch": 40.01992337164751, | |
| "grad_norm": 0.0018771645845845342, | |
| "learning_rate": 2.0008514261387825e-06, | |
| "loss": 0.0001, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 40.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.481161594390869, | |
| "eval_runtime": 27.8586, | |
| "eval_samples_per_second": 1.615, | |
| "eval_steps_per_second": 1.615, | |
| "step": 10701 | |
| }, | |
| { | |
| "epoch": 41.000689655172415, | |
| "grad_norm": 0.012024985626339912, | |
| "learning_rate": 1.992337164750958e-06, | |
| "loss": 0.0021, | |
| "step": 10710 | |
| }, | |
| { | |
| "epoch": 41.00145593869732, | |
| "grad_norm": 0.0026518628001213074, | |
| "learning_rate": 1.9838229033631333e-06, | |
| "loss": 0.0006, | |
| "step": 10720 | |
| }, | |
| { | |
| "epoch": 41.00222222222222, | |
| "grad_norm": 0.05540211498737335, | |
| "learning_rate": 1.9753086419753087e-06, | |
| "loss": 0.0002, | |
| "step": 10730 | |
| }, | |
| { | |
| "epoch": 41.002988505747126, | |
| "grad_norm": 0.0030381560791283846, | |
| "learning_rate": 1.966794380587484e-06, | |
| "loss": 0.6141, | |
| "step": 10740 | |
| }, | |
| { | |
| "epoch": 41.00375478927203, | |
| "grad_norm": 0.0010614864295348525, | |
| "learning_rate": 1.9582801191996596e-06, | |
| "loss": 0.0001, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 41.00452107279693, | |
| "grad_norm": 0.0018325786804780364, | |
| "learning_rate": 1.949765857811835e-06, | |
| "loss": 0.0376, | |
| "step": 10760 | |
| }, | |
| { | |
| "epoch": 41.00528735632184, | |
| "grad_norm": 0.007499290630221367, | |
| "learning_rate": 1.9412515964240104e-06, | |
| "loss": 0.0002, | |
| "step": 10770 | |
| }, | |
| { | |
| "epoch": 41.00605363984674, | |
| "grad_norm": 0.0025593435857445, | |
| "learning_rate": 1.932737335036186e-06, | |
| "loss": 0.0002, | |
| "step": 10780 | |
| }, | |
| { | |
| "epoch": 41.006819923371644, | |
| "grad_norm": 0.00169795798137784, | |
| "learning_rate": 1.9242230736483612e-06, | |
| "loss": 0.0001, | |
| "step": 10790 | |
| }, | |
| { | |
| "epoch": 41.007586206896555, | |
| "grad_norm": 0.03861542046070099, | |
| "learning_rate": 1.9157088122605367e-06, | |
| "loss": 0.2676, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 41.00835249042146, | |
| "grad_norm": 0.0034034252166748047, | |
| "learning_rate": 1.9071945508727119e-06, | |
| "loss": 0.1978, | |
| "step": 10810 | |
| }, | |
| { | |
| "epoch": 41.00911877394636, | |
| "grad_norm": 0.0061155627481639385, | |
| "learning_rate": 1.8986802894848875e-06, | |
| "loss": 0.0001, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 41.009885057471266, | |
| "grad_norm": 0.00044847006211057305, | |
| "learning_rate": 1.8901660280970627e-06, | |
| "loss": 0.0, | |
| "step": 10830 | |
| }, | |
| { | |
| "epoch": 41.01065134099617, | |
| "grad_norm": 0.0002485412987880409, | |
| "learning_rate": 1.8816517667092381e-06, | |
| "loss": 0.0, | |
| "step": 10840 | |
| }, | |
| { | |
| "epoch": 41.01141762452107, | |
| "grad_norm": 0.0023819031193852425, | |
| "learning_rate": 1.8731375053214135e-06, | |
| "loss": 0.0002, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 41.01218390804598, | |
| "grad_norm": 0.05877603963017464, | |
| "learning_rate": 1.864623243933589e-06, | |
| "loss": 0.0002, | |
| "step": 10860 | |
| }, | |
| { | |
| "epoch": 41.01295019157088, | |
| "grad_norm": 0.022863805294036865, | |
| "learning_rate": 1.8561089825457644e-06, | |
| "loss": 0.0002, | |
| "step": 10870 | |
| }, | |
| { | |
| "epoch": 41.013716475095784, | |
| "grad_norm": 0.008092855103313923, | |
| "learning_rate": 1.8475947211579398e-06, | |
| "loss": 0.0001, | |
| "step": 10880 | |
| }, | |
| { | |
| "epoch": 41.01448275862069, | |
| "grad_norm": 0.0005834649200551212, | |
| "learning_rate": 1.839080459770115e-06, | |
| "loss": 0.0001, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 41.01524904214559, | |
| "grad_norm": 0.0027983556501567364, | |
| "learning_rate": 1.8305661983822906e-06, | |
| "loss": 0.0006, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 41.0160153256705, | |
| "grad_norm": 0.0025655950885266066, | |
| "learning_rate": 1.8220519369944658e-06, | |
| "loss": 0.0001, | |
| "step": 10910 | |
| }, | |
| { | |
| "epoch": 41.016781609195405, | |
| "grad_norm": 0.005393259227275848, | |
| "learning_rate": 1.8135376756066415e-06, | |
| "loss": 0.0135, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 41.01754789272031, | |
| "grad_norm": 0.002339740050956607, | |
| "learning_rate": 1.8050234142188167e-06, | |
| "loss": 0.0226, | |
| "step": 10930 | |
| }, | |
| { | |
| "epoch": 41.01831417624521, | |
| "grad_norm": 0.05366809293627739, | |
| "learning_rate": 1.7965091528309919e-06, | |
| "loss": 0.0, | |
| "step": 10940 | |
| }, | |
| { | |
| "epoch": 41.019080459770116, | |
| "grad_norm": 118.51295471191406, | |
| "learning_rate": 1.7879948914431675e-06, | |
| "loss": 0.8814, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 41.01984674329502, | |
| "grad_norm": 0.0012721455423161387, | |
| "learning_rate": 1.7794806300553427e-06, | |
| "loss": 0.0001, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 41.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.5846524238586426, | |
| "eval_runtime": 29.0995, | |
| "eval_samples_per_second": 1.546, | |
| "eval_steps_per_second": 1.546, | |
| "step": 10962 | |
| }, | |
| { | |
| "epoch": 42.00061302681992, | |
| "grad_norm": 0.0012239107163622975, | |
| "learning_rate": 1.7709663686675183e-06, | |
| "loss": 0.0001, | |
| "step": 10970 | |
| }, | |
| { | |
| "epoch": 42.00137931034483, | |
| "grad_norm": 0.12653380632400513, | |
| "learning_rate": 1.7624521072796935e-06, | |
| "loss": 0.0003, | |
| "step": 10980 | |
| }, | |
| { | |
| "epoch": 42.002145593869734, | |
| "grad_norm": 0.0005643948097713292, | |
| "learning_rate": 1.753937845891869e-06, | |
| "loss": 0.0001, | |
| "step": 10990 | |
| }, | |
| { | |
| "epoch": 42.00291187739464, | |
| "grad_norm": 0.0044510625302791595, | |
| "learning_rate": 1.7454235845040444e-06, | |
| "loss": 0.0004, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 42.00367816091954, | |
| "grad_norm": 0.0056150369346141815, | |
| "learning_rate": 1.7369093231162198e-06, | |
| "loss": 0.0001, | |
| "step": 11010 | |
| }, | |
| { | |
| "epoch": 42.004444444444445, | |
| "grad_norm": 0.0006997102755121887, | |
| "learning_rate": 1.7283950617283952e-06, | |
| "loss": 0.4059, | |
| "step": 11020 | |
| }, | |
| { | |
| "epoch": 42.00521072796935, | |
| "grad_norm": 0.0017448539147153497, | |
| "learning_rate": 1.7198808003405706e-06, | |
| "loss": 0.0, | |
| "step": 11030 | |
| }, | |
| { | |
| "epoch": 42.00597701149425, | |
| "grad_norm": 0.0004077133780810982, | |
| "learning_rate": 1.7113665389527458e-06, | |
| "loss": 0.4904, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 42.006743295019156, | |
| "grad_norm": 0.0003028385399375111, | |
| "learning_rate": 1.7028522775649215e-06, | |
| "loss": 0.0001, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 42.00750957854406, | |
| "grad_norm": 0.00037251305184327066, | |
| "learning_rate": 1.6943380161770967e-06, | |
| "loss": 0.0002, | |
| "step": 11060 | |
| }, | |
| { | |
| "epoch": 42.00827586206896, | |
| "grad_norm": 0.0021622315980494022, | |
| "learning_rate": 1.6858237547892723e-06, | |
| "loss": 0.0001, | |
| "step": 11070 | |
| }, | |
| { | |
| "epoch": 42.00904214559387, | |
| "grad_norm": 0.0006449736538343132, | |
| "learning_rate": 1.6773094934014475e-06, | |
| "loss": 0.0003, | |
| "step": 11080 | |
| }, | |
| { | |
| "epoch": 42.00980842911878, | |
| "grad_norm": 0.04130319505929947, | |
| "learning_rate": 1.6687952320136231e-06, | |
| "loss": 0.0002, | |
| "step": 11090 | |
| }, | |
| { | |
| "epoch": 42.01057471264368, | |
| "grad_norm": 0.0020695319399237633, | |
| "learning_rate": 1.6602809706257983e-06, | |
| "loss": 0.0001, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 42.011340996168585, | |
| "grad_norm": 0.0004019753250759095, | |
| "learning_rate": 1.6517667092379737e-06, | |
| "loss": 0.0003, | |
| "step": 11110 | |
| }, | |
| { | |
| "epoch": 42.01210727969349, | |
| "grad_norm": 0.00026687647914513946, | |
| "learning_rate": 1.6432524478501492e-06, | |
| "loss": 0.0001, | |
| "step": 11120 | |
| }, | |
| { | |
| "epoch": 42.01287356321839, | |
| "grad_norm": 0.0002940501144621521, | |
| "learning_rate": 1.6347381864623246e-06, | |
| "loss": 0.0, | |
| "step": 11130 | |
| }, | |
| { | |
| "epoch": 42.013639846743295, | |
| "grad_norm": 0.0004536312189884484, | |
| "learning_rate": 1.6262239250744998e-06, | |
| "loss": 0.0001, | |
| "step": 11140 | |
| }, | |
| { | |
| "epoch": 42.0144061302682, | |
| "grad_norm": 0.0009486758499406278, | |
| "learning_rate": 1.6177096636866754e-06, | |
| "loss": 0.0002, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 42.0151724137931, | |
| "grad_norm": 0.0006705402047373354, | |
| "learning_rate": 1.6091954022988506e-06, | |
| "loss": 0.5375, | |
| "step": 11160 | |
| }, | |
| { | |
| "epoch": 42.015938697318006, | |
| "grad_norm": 0.0014383239904418588, | |
| "learning_rate": 1.6006811409110262e-06, | |
| "loss": 0.0737, | |
| "step": 11170 | |
| }, | |
| { | |
| "epoch": 42.01670498084291, | |
| "grad_norm": 0.0038608924951404333, | |
| "learning_rate": 1.5921668795232015e-06, | |
| "loss": 1.4455, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 42.01747126436781, | |
| "grad_norm": 0.0008818919886834919, | |
| "learning_rate": 1.583652618135377e-06, | |
| "loss": 0.0, | |
| "step": 11190 | |
| }, | |
| { | |
| "epoch": 42.01823754789272, | |
| "grad_norm": 0.0032676539849489927, | |
| "learning_rate": 1.5751383567475523e-06, | |
| "loss": 0.0001, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 42.01900383141763, | |
| "grad_norm": 0.011811945587396622, | |
| "learning_rate": 1.5666240953597275e-06, | |
| "loss": 0.0061, | |
| "step": 11210 | |
| }, | |
| { | |
| "epoch": 42.01977011494253, | |
| "grad_norm": 0.00035030950675718486, | |
| "learning_rate": 1.5581098339719031e-06, | |
| "loss": 0.0, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 42.02, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.55254864692688, | |
| "eval_runtime": 29.1436, | |
| "eval_samples_per_second": 1.544, | |
| "eval_steps_per_second": 1.544, | |
| "step": 11223 | |
| }, | |
| { | |
| "epoch": 43.00053639846743, | |
| "grad_norm": 0.0018214564770460129, | |
| "learning_rate": 1.5495955725840783e-06, | |
| "loss": 0.0001, | |
| "step": 11230 | |
| }, | |
| { | |
| "epoch": 43.001302681992335, | |
| "grad_norm": 0.0032354502473026514, | |
| "learning_rate": 1.541081311196254e-06, | |
| "loss": 0.1826, | |
| "step": 11240 | |
| }, | |
| { | |
| "epoch": 43.00206896551724, | |
| "grad_norm": 0.0002784720272757113, | |
| "learning_rate": 1.5325670498084292e-06, | |
| "loss": 0.0001, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 43.00283524904214, | |
| "grad_norm": 0.0002746792451944202, | |
| "learning_rate": 1.5240527884206046e-06, | |
| "loss": 0.0001, | |
| "step": 11260 | |
| }, | |
| { | |
| "epoch": 43.00360153256705, | |
| "grad_norm": 0.021473722532391548, | |
| "learning_rate": 1.51553852703278e-06, | |
| "loss": 0.0, | |
| "step": 11270 | |
| }, | |
| { | |
| "epoch": 43.00436781609196, | |
| "grad_norm": 0.0003531036782078445, | |
| "learning_rate": 1.5070242656449554e-06, | |
| "loss": 0.7287, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 43.00513409961686, | |
| "grad_norm": 0.0039008602034300566, | |
| "learning_rate": 1.4985100042571306e-06, | |
| "loss": 0.0001, | |
| "step": 11290 | |
| }, | |
| { | |
| "epoch": 43.005900383141764, | |
| "grad_norm": 0.0010976424673572183, | |
| "learning_rate": 1.4899957428693062e-06, | |
| "loss": 0.0001, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 43.00666666666667, | |
| "grad_norm": 0.03243381157517433, | |
| "learning_rate": 1.4814814814814815e-06, | |
| "loss": 0.0001, | |
| "step": 11310 | |
| }, | |
| { | |
| "epoch": 43.00743295019157, | |
| "grad_norm": 0.005974556319415569, | |
| "learning_rate": 1.472967220093657e-06, | |
| "loss": 0.0001, | |
| "step": 11320 | |
| }, | |
| { | |
| "epoch": 43.008199233716475, | |
| "grad_norm": 0.0004650621267501265, | |
| "learning_rate": 1.4644529587058323e-06, | |
| "loss": 0.0001, | |
| "step": 11330 | |
| }, | |
| { | |
| "epoch": 43.00896551724138, | |
| "grad_norm": 0.00025246807490475476, | |
| "learning_rate": 1.455938697318008e-06, | |
| "loss": 0.0001, | |
| "step": 11340 | |
| }, | |
| { | |
| "epoch": 43.00973180076628, | |
| "grad_norm": 0.002089502289891243, | |
| "learning_rate": 1.4474244359301831e-06, | |
| "loss": 0.0002, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 43.010498084291186, | |
| "grad_norm": 0.0005093670915812254, | |
| "learning_rate": 1.4389101745423588e-06, | |
| "loss": 0.0001, | |
| "step": 11360 | |
| }, | |
| { | |
| "epoch": 43.01126436781609, | |
| "grad_norm": 0.005426987074315548, | |
| "learning_rate": 1.430395913154534e-06, | |
| "loss": 0.0001, | |
| "step": 11370 | |
| }, | |
| { | |
| "epoch": 43.01203065134099, | |
| "grad_norm": 0.020039698109030724, | |
| "learning_rate": 1.4218816517667094e-06, | |
| "loss": 0.0001, | |
| "step": 11380 | |
| }, | |
| { | |
| "epoch": 43.012796934865904, | |
| "grad_norm": 0.011925816535949707, | |
| "learning_rate": 1.4133673903788848e-06, | |
| "loss": 0.0001, | |
| "step": 11390 | |
| }, | |
| { | |
| "epoch": 43.01356321839081, | |
| "grad_norm": 0.0019410884706303477, | |
| "learning_rate": 1.4048531289910602e-06, | |
| "loss": 0.0001, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 43.01432950191571, | |
| "grad_norm": 0.010476148687303066, | |
| "learning_rate": 1.3963388676032354e-06, | |
| "loss": 0.0001, | |
| "step": 11410 | |
| }, | |
| { | |
| "epoch": 43.015095785440614, | |
| "grad_norm": 0.0018306553829461336, | |
| "learning_rate": 1.387824606215411e-06, | |
| "loss": 0.0002, | |
| "step": 11420 | |
| }, | |
| { | |
| "epoch": 43.01586206896552, | |
| "grad_norm": 0.00024431603378616273, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.0, | |
| "step": 11430 | |
| }, | |
| { | |
| "epoch": 43.01662835249042, | |
| "grad_norm": 0.0005057769012637436, | |
| "learning_rate": 1.3707960834397619e-06, | |
| "loss": 0.0, | |
| "step": 11440 | |
| }, | |
| { | |
| "epoch": 43.017394636015325, | |
| "grad_norm": 0.00029404531233012676, | |
| "learning_rate": 1.362281822051937e-06, | |
| "loss": 0.369, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 43.01816091954023, | |
| "grad_norm": 0.0014582271687686443, | |
| "learning_rate": 1.3537675606641127e-06, | |
| "loss": 0.0001, | |
| "step": 11460 | |
| }, | |
| { | |
| "epoch": 43.01892720306513, | |
| "grad_norm": 0.0006710296729579568, | |
| "learning_rate": 1.345253299276288e-06, | |
| "loss": 0.0, | |
| "step": 11470 | |
| }, | |
| { | |
| "epoch": 43.019693486590036, | |
| "grad_norm": 1.5178964138031006, | |
| "learning_rate": 1.3367390378884631e-06, | |
| "loss": 0.002, | |
| "step": 11480 | |
| }, | |
| { | |
| "epoch": 43.02, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.674553155899048, | |
| "eval_runtime": 27.901, | |
| "eval_samples_per_second": 1.613, | |
| "eval_steps_per_second": 1.613, | |
| "step": 11484 | |
| }, | |
| { | |
| "epoch": 44.000459770114944, | |
| "grad_norm": 0.006704441737383604, | |
| "learning_rate": 1.3282247765006387e-06, | |
| "loss": 0.0001, | |
| "step": 11490 | |
| }, | |
| { | |
| "epoch": 44.00122605363985, | |
| "grad_norm": 0.01098534557968378, | |
| "learning_rate": 1.319710515112814e-06, | |
| "loss": 0.0001, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 44.00199233716475, | |
| "grad_norm": 0.0022969290148466825, | |
| "learning_rate": 1.3111962537249896e-06, | |
| "loss": 0.6907, | |
| "step": 11510 | |
| }, | |
| { | |
| "epoch": 44.002758620689654, | |
| "grad_norm": 1385.7335205078125, | |
| "learning_rate": 1.3026819923371648e-06, | |
| "loss": 1.0054, | |
| "step": 11520 | |
| }, | |
| { | |
| "epoch": 44.00352490421456, | |
| "grad_norm": 0.00026755925500765443, | |
| "learning_rate": 1.2941677309493402e-06, | |
| "loss": 0.0009, | |
| "step": 11530 | |
| }, | |
| { | |
| "epoch": 44.00429118773946, | |
| "grad_norm": 0.00037283002166077495, | |
| "learning_rate": 1.2856534695615156e-06, | |
| "loss": 0.0001, | |
| "step": 11540 | |
| }, | |
| { | |
| "epoch": 44.005057471264365, | |
| "grad_norm": 0.001472074189223349, | |
| "learning_rate": 1.277139208173691e-06, | |
| "loss": 0.0002, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 44.00582375478927, | |
| "grad_norm": 0.01885879412293434, | |
| "learning_rate": 1.2686249467858662e-06, | |
| "loss": 0.0001, | |
| "step": 11560 | |
| }, | |
| { | |
| "epoch": 44.00659003831418, | |
| "grad_norm": 0.0011555615346878767, | |
| "learning_rate": 1.2601106853980419e-06, | |
| "loss": 0.0001, | |
| "step": 11570 | |
| }, | |
| { | |
| "epoch": 44.00735632183908, | |
| "grad_norm": 0.0003773407079279423, | |
| "learning_rate": 1.251596424010217e-06, | |
| "loss": 0.0001, | |
| "step": 11580 | |
| }, | |
| { | |
| "epoch": 44.00812260536399, | |
| "grad_norm": 0.0011189919896423817, | |
| "learning_rate": 1.2430821626223927e-06, | |
| "loss": 0.0007, | |
| "step": 11590 | |
| }, | |
| { | |
| "epoch": 44.00888888888889, | |
| "grad_norm": 0.0008237032452598214, | |
| "learning_rate": 1.234567901234568e-06, | |
| "loss": 0.0, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 44.009655172413794, | |
| "grad_norm": 0.0003113767597824335, | |
| "learning_rate": 1.2260536398467433e-06, | |
| "loss": 0.0001, | |
| "step": 11610 | |
| }, | |
| { | |
| "epoch": 44.0104214559387, | |
| "grad_norm": 0.0003217561752535403, | |
| "learning_rate": 1.2175393784589187e-06, | |
| "loss": 0.0, | |
| "step": 11620 | |
| }, | |
| { | |
| "epoch": 44.0111877394636, | |
| "grad_norm": 0.0002517100947443396, | |
| "learning_rate": 1.2090251170710942e-06, | |
| "loss": 0.0002, | |
| "step": 11630 | |
| }, | |
| { | |
| "epoch": 44.011954022988505, | |
| "grad_norm": 0.0004395510477479547, | |
| "learning_rate": 1.2005108556832696e-06, | |
| "loss": 0.0001, | |
| "step": 11640 | |
| }, | |
| { | |
| "epoch": 44.01272030651341, | |
| "grad_norm": 0.008616892620921135, | |
| "learning_rate": 1.191996594295445e-06, | |
| "loss": 0.0, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 44.01348659003831, | |
| "grad_norm": 0.0005844251718372107, | |
| "learning_rate": 1.1834823329076204e-06, | |
| "loss": 0.0005, | |
| "step": 11660 | |
| }, | |
| { | |
| "epoch": 44.014252873563215, | |
| "grad_norm": 0.00044547973084263504, | |
| "learning_rate": 1.1749680715197956e-06, | |
| "loss": 0.0001, | |
| "step": 11670 | |
| }, | |
| { | |
| "epoch": 44.01501915708812, | |
| "grad_norm": 0.002086098538711667, | |
| "learning_rate": 1.166453810131971e-06, | |
| "loss": 0.0, | |
| "step": 11680 | |
| }, | |
| { | |
| "epoch": 44.01578544061303, | |
| "grad_norm": 0.00023247070203069597, | |
| "learning_rate": 1.1579395487441465e-06, | |
| "loss": 0.0002, | |
| "step": 11690 | |
| }, | |
| { | |
| "epoch": 44.01655172413793, | |
| "grad_norm": 0.00041663722367957234, | |
| "learning_rate": 1.1494252873563219e-06, | |
| "loss": 0.0023, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 44.01731800766284, | |
| "grad_norm": 0.0008354912861250341, | |
| "learning_rate": 1.1409110259684973e-06, | |
| "loss": 0.0001, | |
| "step": 11710 | |
| }, | |
| { | |
| "epoch": 44.01808429118774, | |
| "grad_norm": 0.00040326014277525246, | |
| "learning_rate": 1.1323967645806727e-06, | |
| "loss": 0.0013, | |
| "step": 11720 | |
| }, | |
| { | |
| "epoch": 44.018850574712644, | |
| "grad_norm": 0.025961795821785927, | |
| "learning_rate": 1.1238825031928481e-06, | |
| "loss": 0.0546, | |
| "step": 11730 | |
| }, | |
| { | |
| "epoch": 44.01961685823755, | |
| "grad_norm": 0.03274637833237648, | |
| "learning_rate": 1.1153682418050235e-06, | |
| "loss": 0.0004, | |
| "step": 11740 | |
| }, | |
| { | |
| "epoch": 44.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.4888393878936768, | |
| "eval_runtime": 27.8832, | |
| "eval_samples_per_second": 1.614, | |
| "eval_steps_per_second": 1.614, | |
| "step": 11745 | |
| }, | |
| { | |
| "epoch": 45.000383141762455, | |
| "grad_norm": 0.0010363643523305655, | |
| "learning_rate": 1.106853980417199e-06, | |
| "loss": 0.0091, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 45.00114942528736, | |
| "grad_norm": 0.0014785947278141975, | |
| "learning_rate": 1.0983397190293744e-06, | |
| "loss": 0.0001, | |
| "step": 11760 | |
| }, | |
| { | |
| "epoch": 45.00191570881226, | |
| "grad_norm": 0.0003646929981186986, | |
| "learning_rate": 1.0898254576415498e-06, | |
| "loss": 0.0824, | |
| "step": 11770 | |
| }, | |
| { | |
| "epoch": 45.002681992337166, | |
| "grad_norm": 116.26171875, | |
| "learning_rate": 1.0813111962537252e-06, | |
| "loss": 0.0102, | |
| "step": 11780 | |
| }, | |
| { | |
| "epoch": 45.00344827586207, | |
| "grad_norm": 0.0005606987397186458, | |
| "learning_rate": 1.0727969348659004e-06, | |
| "loss": 0.0001, | |
| "step": 11790 | |
| }, | |
| { | |
| "epoch": 45.00421455938697, | |
| "grad_norm": 14.712228775024414, | |
| "learning_rate": 1.0642826734780758e-06, | |
| "loss": 0.0014, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 45.00498084291188, | |
| "grad_norm": 0.00022383002215065062, | |
| "learning_rate": 1.0557684120902512e-06, | |
| "loss": 0.0001, | |
| "step": 11810 | |
| }, | |
| { | |
| "epoch": 45.00574712643678, | |
| "grad_norm": 0.0012341875117272139, | |
| "learning_rate": 1.0472541507024267e-06, | |
| "loss": 0.3366, | |
| "step": 11820 | |
| }, | |
| { | |
| "epoch": 45.006513409961684, | |
| "grad_norm": 0.000215315303648822, | |
| "learning_rate": 1.038739889314602e-06, | |
| "loss": 0.0001, | |
| "step": 11830 | |
| }, | |
| { | |
| "epoch": 45.00727969348659, | |
| "grad_norm": 0.000865298556163907, | |
| "learning_rate": 1.0302256279267775e-06, | |
| "loss": 0.0001, | |
| "step": 11840 | |
| }, | |
| { | |
| "epoch": 45.00804597701149, | |
| "grad_norm": 0.0009852108778432012, | |
| "learning_rate": 1.021711366538953e-06, | |
| "loss": 0.0, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 45.008812260536395, | |
| "grad_norm": 0.001507429638877511, | |
| "learning_rate": 1.0131971051511283e-06, | |
| "loss": 0.0002, | |
| "step": 11860 | |
| }, | |
| { | |
| "epoch": 45.009578544061306, | |
| "grad_norm": 0.010270087979733944, | |
| "learning_rate": 1.0046828437633035e-06, | |
| "loss": 0.0001, | |
| "step": 11870 | |
| }, | |
| { | |
| "epoch": 45.01034482758621, | |
| "grad_norm": 0.017873380333185196, | |
| "learning_rate": 9.96168582375479e-07, | |
| "loss": 0.0001, | |
| "step": 11880 | |
| }, | |
| { | |
| "epoch": 45.01111111111111, | |
| "grad_norm": 1.1168253421783447, | |
| "learning_rate": 9.876543209876544e-07, | |
| "loss": 0.0003, | |
| "step": 11890 | |
| }, | |
| { | |
| "epoch": 45.01187739463602, | |
| "grad_norm": 0.00040387007175013423, | |
| "learning_rate": 9.791400595998298e-07, | |
| "loss": 0.0001, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 45.01264367816092, | |
| "grad_norm": 0.20708155632019043, | |
| "learning_rate": 9.706257982120052e-07, | |
| "loss": 0.0002, | |
| "step": 11910 | |
| }, | |
| { | |
| "epoch": 45.013409961685824, | |
| "grad_norm": 0.0003908909566234797, | |
| "learning_rate": 9.621115368241806e-07, | |
| "loss": 0.0, | |
| "step": 11920 | |
| }, | |
| { | |
| "epoch": 45.01417624521073, | |
| "grad_norm": 0.0021993808913975954, | |
| "learning_rate": 9.535972754363559e-07, | |
| "loss": 0.0001, | |
| "step": 11930 | |
| }, | |
| { | |
| "epoch": 45.01494252873563, | |
| "grad_norm": 0.0029007780831307173, | |
| "learning_rate": 9.450830140485314e-07, | |
| "loss": 0.7119, | |
| "step": 11940 | |
| }, | |
| { | |
| "epoch": 45.015708812260534, | |
| "grad_norm": 0.006735900416970253, | |
| "learning_rate": 9.365687526607068e-07, | |
| "loss": 0.0001, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 45.01647509578544, | |
| "grad_norm": 0.0016031097620725632, | |
| "learning_rate": 9.280544912728822e-07, | |
| "loss": 0.0, | |
| "step": 11960 | |
| }, | |
| { | |
| "epoch": 45.01724137931034, | |
| "grad_norm": 0.0003009812789969146, | |
| "learning_rate": 9.195402298850575e-07, | |
| "loss": 0.0, | |
| "step": 11970 | |
| }, | |
| { | |
| "epoch": 45.01800766283525, | |
| "grad_norm": 0.005815963726490736, | |
| "learning_rate": 9.110259684972329e-07, | |
| "loss": 0.0002, | |
| "step": 11980 | |
| }, | |
| { | |
| "epoch": 45.018773946360156, | |
| "grad_norm": 0.000267479510512203, | |
| "learning_rate": 9.025117071094083e-07, | |
| "loss": 0.4801, | |
| "step": 11990 | |
| }, | |
| { | |
| "epoch": 45.01954022988506, | |
| "grad_norm": 0.0044364649802446365, | |
| "learning_rate": 8.939974457215837e-07, | |
| "loss": 0.0001, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 45.02, | |
| "eval_accuracy": 0.6444444444444445, | |
| "eval_loss": 2.5661938190460205, | |
| "eval_runtime": 27.8145, | |
| "eval_samples_per_second": 1.618, | |
| "eval_steps_per_second": 1.618, | |
| "step": 12006 | |
| }, | |
| { | |
| "epoch": 46.00030651340996, | |
| "grad_norm": 0.0006088312948122621, | |
| "learning_rate": 8.854831843337592e-07, | |
| "loss": 0.0002, | |
| "step": 12010 | |
| }, | |
| { | |
| "epoch": 46.001072796934864, | |
| "grad_norm": 0.00147931557148695, | |
| "learning_rate": 8.769689229459345e-07, | |
| "loss": 0.0001, | |
| "step": 12020 | |
| }, | |
| { | |
| "epoch": 46.00183908045977, | |
| "grad_norm": 0.006793022621423006, | |
| "learning_rate": 8.684546615581099e-07, | |
| "loss": 0.4711, | |
| "step": 12030 | |
| }, | |
| { | |
| "epoch": 46.00260536398467, | |
| "grad_norm": 0.00036958555574528873, | |
| "learning_rate": 8.599404001702853e-07, | |
| "loss": 0.0002, | |
| "step": 12040 | |
| }, | |
| { | |
| "epoch": 46.00337164750958, | |
| "grad_norm": 0.002713366411626339, | |
| "learning_rate": 8.514261387824607e-07, | |
| "loss": 0.0, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 46.004137931034485, | |
| "grad_norm": 0.001008681021630764, | |
| "learning_rate": 8.429118773946361e-07, | |
| "loss": 0.0001, | |
| "step": 12060 | |
| }, | |
| { | |
| "epoch": 46.00490421455939, | |
| "grad_norm": 0.0019728182815015316, | |
| "learning_rate": 8.343976160068116e-07, | |
| "loss": 0.0, | |
| "step": 12070 | |
| }, | |
| { | |
| "epoch": 46.00567049808429, | |
| "grad_norm": 0.8289773464202881, | |
| "learning_rate": 8.258833546189869e-07, | |
| "loss": 0.001, | |
| "step": 12080 | |
| }, | |
| { | |
| "epoch": 46.006436781609196, | |
| "grad_norm": 0.0003419470158405602, | |
| "learning_rate": 8.173690932311623e-07, | |
| "loss": 0.0133, | |
| "step": 12090 | |
| }, | |
| { | |
| "epoch": 46.0072030651341, | |
| "grad_norm": 0.002687289845198393, | |
| "learning_rate": 8.088548318433377e-07, | |
| "loss": 0.0, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 46.007969348659, | |
| "grad_norm": 0.004384917207062244, | |
| "learning_rate": 8.003405704555131e-07, | |
| "loss": 0.0001, | |
| "step": 12110 | |
| }, | |
| { | |
| "epoch": 46.00873563218391, | |
| "grad_norm": 0.009863078594207764, | |
| "learning_rate": 7.918263090676885e-07, | |
| "loss": 0.0001, | |
| "step": 12120 | |
| }, | |
| { | |
| "epoch": 46.00950191570881, | |
| "grad_norm": 0.0006764632998965681, | |
| "learning_rate": 7.833120476798637e-07, | |
| "loss": 0.4057, | |
| "step": 12130 | |
| }, | |
| { | |
| "epoch": 46.010268199233714, | |
| "grad_norm": 0.0015612775459885597, | |
| "learning_rate": 7.747977862920392e-07, | |
| "loss": 0.0, | |
| "step": 12140 | |
| }, | |
| { | |
| "epoch": 46.01103448275862, | |
| "grad_norm": 0.0002601986925583333, | |
| "learning_rate": 7.662835249042146e-07, | |
| "loss": 0.2069, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 46.01180076628353, | |
| "grad_norm": 0.0012323512928560376, | |
| "learning_rate": 7.5776926351639e-07, | |
| "loss": 0.0001, | |
| "step": 12160 | |
| }, | |
| { | |
| "epoch": 46.01256704980843, | |
| "grad_norm": 0.00020239534205757082, | |
| "learning_rate": 7.492550021285653e-07, | |
| "loss": 0.0, | |
| "step": 12170 | |
| }, | |
| { | |
| "epoch": 46.013333333333335, | |
| "grad_norm": 0.001446614391170442, | |
| "learning_rate": 7.407407407407407e-07, | |
| "loss": 0.0001, | |
| "step": 12180 | |
| }, | |
| { | |
| "epoch": 46.01409961685824, | |
| "grad_norm": 0.0011185260955244303, | |
| "learning_rate": 7.322264793529161e-07, | |
| "loss": 0.0004, | |
| "step": 12190 | |
| }, | |
| { | |
| "epoch": 46.01486590038314, | |
| "grad_norm": 0.00250724027864635, | |
| "learning_rate": 7.237122179650916e-07, | |
| "loss": 0.0, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 46.015632183908046, | |
| "grad_norm": 0.002670790534466505, | |
| "learning_rate": 7.15197956577267e-07, | |
| "loss": 0.0312, | |
| "step": 12210 | |
| }, | |
| { | |
| "epoch": 46.01639846743295, | |
| "grad_norm": 0.004383922554552555, | |
| "learning_rate": 7.066836951894424e-07, | |
| "loss": 0.0001, | |
| "step": 12220 | |
| }, | |
| { | |
| "epoch": 46.01716475095785, | |
| "grad_norm": 0.0005220200400799513, | |
| "learning_rate": 6.981694338016177e-07, | |
| "loss": 0.0049, | |
| "step": 12230 | |
| }, | |
| { | |
| "epoch": 46.01793103448276, | |
| "grad_norm": 0.0005428689764812589, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.0001, | |
| "step": 12240 | |
| }, | |
| { | |
| "epoch": 46.01869731800766, | |
| "grad_norm": 0.4421864449977875, | |
| "learning_rate": 6.811409110259685e-07, | |
| "loss": 0.2773, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 46.019463601532564, | |
| "grad_norm": 0.0004877319443039596, | |
| "learning_rate": 6.72626649638144e-07, | |
| "loss": 0.0011, | |
| "step": 12260 | |
| }, | |
| { | |
| "epoch": 46.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.5287835597991943, | |
| "eval_runtime": 27.9127, | |
| "eval_samples_per_second": 1.612, | |
| "eval_steps_per_second": 1.612, | |
| "step": 12267 | |
| }, | |
| { | |
| "epoch": 47.00022988505747, | |
| "grad_norm": 0.015879875048995018, | |
| "learning_rate": 6.641123882503194e-07, | |
| "loss": 0.0001, | |
| "step": 12270 | |
| }, | |
| { | |
| "epoch": 47.000996168582375, | |
| "grad_norm": 0.0003854314563795924, | |
| "learning_rate": 6.555981268624948e-07, | |
| "loss": 0.0005, | |
| "step": 12280 | |
| }, | |
| { | |
| "epoch": 47.00176245210728, | |
| "grad_norm": 0.0004334681434556842, | |
| "learning_rate": 6.470838654746701e-07, | |
| "loss": 0.0001, | |
| "step": 12290 | |
| }, | |
| { | |
| "epoch": 47.00252873563218, | |
| "grad_norm": 0.0005379935028031468, | |
| "learning_rate": 6.385696040868455e-07, | |
| "loss": 0.0001, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 47.003295019157086, | |
| "grad_norm": 0.0024425890296697617, | |
| "learning_rate": 6.300553426990209e-07, | |
| "loss": 0.0001, | |
| "step": 12310 | |
| }, | |
| { | |
| "epoch": 47.00406130268199, | |
| "grad_norm": 0.0013785994378849864, | |
| "learning_rate": 6.215410813111964e-07, | |
| "loss": 0.0, | |
| "step": 12320 | |
| }, | |
| { | |
| "epoch": 47.00482758620689, | |
| "grad_norm": 0.0004363117041066289, | |
| "learning_rate": 6.130268199233717e-07, | |
| "loss": 0.0006, | |
| "step": 12330 | |
| }, | |
| { | |
| "epoch": 47.005593869731804, | |
| "grad_norm": 0.004162503406405449, | |
| "learning_rate": 6.045125585355471e-07, | |
| "loss": 0.0001, | |
| "step": 12340 | |
| }, | |
| { | |
| "epoch": 47.00636015325671, | |
| "grad_norm": 0.00028804532485082746, | |
| "learning_rate": 5.959982971477225e-07, | |
| "loss": 0.0001, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 47.00712643678161, | |
| "grad_norm": 0.00362727721221745, | |
| "learning_rate": 5.874840357598978e-07, | |
| "loss": 0.0, | |
| "step": 12360 | |
| }, | |
| { | |
| "epoch": 47.007892720306515, | |
| "grad_norm": 0.00018824009748641402, | |
| "learning_rate": 5.789697743720732e-07, | |
| "loss": 0.0001, | |
| "step": 12370 | |
| }, | |
| { | |
| "epoch": 47.00865900383142, | |
| "grad_norm": 0.0003070647071581334, | |
| "learning_rate": 5.704555129842486e-07, | |
| "loss": 0.0, | |
| "step": 12380 | |
| }, | |
| { | |
| "epoch": 47.00942528735632, | |
| "grad_norm": 0.0030327988788485527, | |
| "learning_rate": 5.619412515964241e-07, | |
| "loss": 0.0, | |
| "step": 12390 | |
| }, | |
| { | |
| "epoch": 47.010191570881226, | |
| "grad_norm": 0.0001980850356630981, | |
| "learning_rate": 5.534269902085995e-07, | |
| "loss": 0.0001, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 47.01095785440613, | |
| "grad_norm": 0.0002615888079162687, | |
| "learning_rate": 5.449127288207749e-07, | |
| "loss": 0.0, | |
| "step": 12410 | |
| }, | |
| { | |
| "epoch": 47.01172413793103, | |
| "grad_norm": 0.01400078646838665, | |
| "learning_rate": 5.363984674329502e-07, | |
| "loss": 0.0001, | |
| "step": 12420 | |
| }, | |
| { | |
| "epoch": 47.01249042145594, | |
| "grad_norm": 0.0021926367189735174, | |
| "learning_rate": 5.278842060451256e-07, | |
| "loss": 0.0, | |
| "step": 12430 | |
| }, | |
| { | |
| "epoch": 47.01325670498084, | |
| "grad_norm": 0.00019135964976157993, | |
| "learning_rate": 5.19369944657301e-07, | |
| "loss": 0.0, | |
| "step": 12440 | |
| }, | |
| { | |
| "epoch": 47.014022988505744, | |
| "grad_norm": 0.001225248328410089, | |
| "learning_rate": 5.108556832694765e-07, | |
| "loss": 0.0, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 47.014789272030654, | |
| "grad_norm": 0.0009840844431892037, | |
| "learning_rate": 5.023414218816518e-07, | |
| "loss": 0.0, | |
| "step": 12460 | |
| }, | |
| { | |
| "epoch": 47.01555555555556, | |
| "grad_norm": 466.7170715332031, | |
| "learning_rate": 4.938271604938272e-07, | |
| "loss": 0.0296, | |
| "step": 12470 | |
| }, | |
| { | |
| "epoch": 47.01632183908046, | |
| "grad_norm": 0.0032234173268079758, | |
| "learning_rate": 4.853128991060026e-07, | |
| "loss": 0.0003, | |
| "step": 12480 | |
| }, | |
| { | |
| "epoch": 47.017088122605365, | |
| "grad_norm": 0.0014763821382075548, | |
| "learning_rate": 4.7679863771817797e-07, | |
| "loss": 0.0001, | |
| "step": 12490 | |
| }, | |
| { | |
| "epoch": 47.01785440613027, | |
| "grad_norm": 0.011778397485613823, | |
| "learning_rate": 4.682843763303534e-07, | |
| "loss": 0.0001, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 47.01862068965517, | |
| "grad_norm": 0.00038351898547261953, | |
| "learning_rate": 4.5977011494252875e-07, | |
| "loss": 0.3203, | |
| "step": 12510 | |
| }, | |
| { | |
| "epoch": 47.019386973180076, | |
| "grad_norm": 0.00020511298498604447, | |
| "learning_rate": 4.5125585355470417e-07, | |
| "loss": 0.0001, | |
| "step": 12520 | |
| }, | |
| { | |
| "epoch": 47.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.561100482940674, | |
| "eval_runtime": 27.8457, | |
| "eval_samples_per_second": 1.616, | |
| "eval_steps_per_second": 1.616, | |
| "step": 12528 | |
| }, | |
| { | |
| "epoch": 48.000153256704984, | |
| "grad_norm": 0.00035865779500454664, | |
| "learning_rate": 4.427415921668796e-07, | |
| "loss": 0.0017, | |
| "step": 12530 | |
| }, | |
| { | |
| "epoch": 48.00091954022989, | |
| "grad_norm": 0.00047454124432988465, | |
| "learning_rate": 4.3422733077905495e-07, | |
| "loss": 0.0001, | |
| "step": 12540 | |
| }, | |
| { | |
| "epoch": 48.00168582375479, | |
| "grad_norm": 0.0050529418513178825, | |
| "learning_rate": 4.2571306939123036e-07, | |
| "loss": 0.6724, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 48.002452107279694, | |
| "grad_norm": 0.16734667122364044, | |
| "learning_rate": 4.171988080034058e-07, | |
| "loss": 0.0001, | |
| "step": 12560 | |
| }, | |
| { | |
| "epoch": 48.0032183908046, | |
| "grad_norm": 0.0004410755936987698, | |
| "learning_rate": 4.0868454661558115e-07, | |
| "loss": 0.0001, | |
| "step": 12570 | |
| }, | |
| { | |
| "epoch": 48.0039846743295, | |
| "grad_norm": 0.0035646662581712008, | |
| "learning_rate": 4.0017028522775656e-07, | |
| "loss": 0.0001, | |
| "step": 12580 | |
| }, | |
| { | |
| "epoch": 48.004750957854405, | |
| "grad_norm": 0.005939258728176355, | |
| "learning_rate": 3.916560238399319e-07, | |
| "loss": 0.0001, | |
| "step": 12590 | |
| }, | |
| { | |
| "epoch": 48.00551724137931, | |
| "grad_norm": 0.0007532021263614297, | |
| "learning_rate": 3.831417624521073e-07, | |
| "loss": 0.0, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 48.00628352490421, | |
| "grad_norm": 0.008377188816666603, | |
| "learning_rate": 3.7462750106428265e-07, | |
| "loss": 0.001, | |
| "step": 12610 | |
| }, | |
| { | |
| "epoch": 48.007049808429116, | |
| "grad_norm": 0.0007006659870967269, | |
| "learning_rate": 3.6611323967645807e-07, | |
| "loss": 0.0, | |
| "step": 12620 | |
| }, | |
| { | |
| "epoch": 48.00781609195402, | |
| "grad_norm": 0.005957789719104767, | |
| "learning_rate": 3.575989782886335e-07, | |
| "loss": 0.0, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 48.00858237547893, | |
| "grad_norm": 0.000713622837793082, | |
| "learning_rate": 3.4908471690080885e-07, | |
| "loss": 0.0, | |
| "step": 12640 | |
| }, | |
| { | |
| "epoch": 48.009348659003834, | |
| "grad_norm": 0.006381857208907604, | |
| "learning_rate": 3.4057045551298427e-07, | |
| "loss": 0.0, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 48.01011494252874, | |
| "grad_norm": 0.0002789432182908058, | |
| "learning_rate": 3.320561941251597e-07, | |
| "loss": 0.0, | |
| "step": 12660 | |
| }, | |
| { | |
| "epoch": 48.01088122605364, | |
| "grad_norm": 0.00030842047999612987, | |
| "learning_rate": 3.2354193273733505e-07, | |
| "loss": 0.0, | |
| "step": 12670 | |
| }, | |
| { | |
| "epoch": 48.011647509578545, | |
| "grad_norm": 0.006707367021590471, | |
| "learning_rate": 3.1502767134951047e-07, | |
| "loss": 0.0001, | |
| "step": 12680 | |
| }, | |
| { | |
| "epoch": 48.01241379310345, | |
| "grad_norm": 0.05167899280786514, | |
| "learning_rate": 3.0651340996168583e-07, | |
| "loss": 0.0, | |
| "step": 12690 | |
| }, | |
| { | |
| "epoch": 48.01318007662835, | |
| "grad_norm": 0.001056885113939643, | |
| "learning_rate": 2.9799914857386125e-07, | |
| "loss": 0.0001, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 48.013946360153255, | |
| "grad_norm": 0.00020573682559188455, | |
| "learning_rate": 2.894848871860366e-07, | |
| "loss": 0.0, | |
| "step": 12710 | |
| }, | |
| { | |
| "epoch": 48.01471264367816, | |
| "grad_norm": 0.0008276683511212468, | |
| "learning_rate": 2.8097062579821203e-07, | |
| "loss": 0.0, | |
| "step": 12720 | |
| }, | |
| { | |
| "epoch": 48.01547892720306, | |
| "grad_norm": 0.0027651754207909107, | |
| "learning_rate": 2.7245636441038745e-07, | |
| "loss": 0.0001, | |
| "step": 12730 | |
| }, | |
| { | |
| "epoch": 48.016245210727966, | |
| "grad_norm": 0.00031005279743112624, | |
| "learning_rate": 2.639421030225628e-07, | |
| "loss": 0.0001, | |
| "step": 12740 | |
| }, | |
| { | |
| "epoch": 48.01701149425288, | |
| "grad_norm": 0.0005220829043537378, | |
| "learning_rate": 2.5542784163473823e-07, | |
| "loss": 0.0001, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 48.01777777777778, | |
| "grad_norm": 0.0003989882825408131, | |
| "learning_rate": 2.469135802469136e-07, | |
| "loss": 0.0, | |
| "step": 12760 | |
| }, | |
| { | |
| "epoch": 48.018544061302684, | |
| "grad_norm": 0.012621737085282803, | |
| "learning_rate": 2.3839931885908898e-07, | |
| "loss": 0.0001, | |
| "step": 12770 | |
| }, | |
| { | |
| "epoch": 48.01931034482759, | |
| "grad_norm": 0.0004822519258596003, | |
| "learning_rate": 2.2988505747126437e-07, | |
| "loss": 0.7043, | |
| "step": 12780 | |
| }, | |
| { | |
| "epoch": 48.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.760641574859619, | |
| "eval_runtime": 27.996, | |
| "eval_samples_per_second": 1.607, | |
| "eval_steps_per_second": 1.607, | |
| "step": 12789 | |
| }, | |
| { | |
| "epoch": 49.00007662835249, | |
| "grad_norm": 0.0023209976498037577, | |
| "learning_rate": 2.213707960834398e-07, | |
| "loss": 1.3152, | |
| "step": 12790 | |
| }, | |
| { | |
| "epoch": 49.00084291187739, | |
| "grad_norm": 0.00585996313020587, | |
| "learning_rate": 2.1285653469561518e-07, | |
| "loss": 0.0, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 49.001609195402295, | |
| "grad_norm": 0.012024904601275921, | |
| "learning_rate": 2.0434227330779057e-07, | |
| "loss": 0.0001, | |
| "step": 12810 | |
| }, | |
| { | |
| "epoch": 49.002375478927206, | |
| "grad_norm": 0.004350293893367052, | |
| "learning_rate": 1.9582801191996594e-07, | |
| "loss": 0.0001, | |
| "step": 12820 | |
| }, | |
| { | |
| "epoch": 49.00314176245211, | |
| "grad_norm": 0.0006611610879190266, | |
| "learning_rate": 1.8731375053214133e-07, | |
| "loss": 0.0001, | |
| "step": 12830 | |
| }, | |
| { | |
| "epoch": 49.00390804597701, | |
| "grad_norm": 0.00027403372223488986, | |
| "learning_rate": 1.7879948914431674e-07, | |
| "loss": 0.0, | |
| "step": 12840 | |
| }, | |
| { | |
| "epoch": 49.00467432950192, | |
| "grad_norm": 0.001186919049359858, | |
| "learning_rate": 1.7028522775649214e-07, | |
| "loss": 0.0001, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 49.00544061302682, | |
| "grad_norm": 0.00033863732824102044, | |
| "learning_rate": 1.6177096636866753e-07, | |
| "loss": 0.0, | |
| "step": 12860 | |
| }, | |
| { | |
| "epoch": 49.006206896551724, | |
| "grad_norm": 0.0006599312764592469, | |
| "learning_rate": 1.5325670498084292e-07, | |
| "loss": 0.0, | |
| "step": 12870 | |
| }, | |
| { | |
| "epoch": 49.00697318007663, | |
| "grad_norm": 0.0018116752617061138, | |
| "learning_rate": 1.447424435930183e-07, | |
| "loss": 0.0001, | |
| "step": 12880 | |
| }, | |
| { | |
| "epoch": 49.00773946360153, | |
| "grad_norm": 0.0010053602745756507, | |
| "learning_rate": 1.3622818220519372e-07, | |
| "loss": 0.0, | |
| "step": 12890 | |
| }, | |
| { | |
| "epoch": 49.008505747126435, | |
| "grad_norm": 0.0005538419936783612, | |
| "learning_rate": 1.2771392081736911e-07, | |
| "loss": 0.0001, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 49.00927203065134, | |
| "grad_norm": 0.0019845098722726107, | |
| "learning_rate": 1.1919965942954449e-07, | |
| "loss": 0.0001, | |
| "step": 12910 | |
| }, | |
| { | |
| "epoch": 49.01003831417624, | |
| "grad_norm": 0.0003088304656557739, | |
| "learning_rate": 1.106853980417199e-07, | |
| "loss": 0.0001, | |
| "step": 12920 | |
| }, | |
| { | |
| "epoch": 49.01080459770115, | |
| "grad_norm": 0.012070775032043457, | |
| "learning_rate": 1.0217113665389529e-07, | |
| "loss": 0.0, | |
| "step": 12930 | |
| }, | |
| { | |
| "epoch": 49.011570881226056, | |
| "grad_norm": 0.0030147642828524113, | |
| "learning_rate": 9.365687526607066e-08, | |
| "loss": 0.0, | |
| "step": 12940 | |
| }, | |
| { | |
| "epoch": 49.01233716475096, | |
| "grad_norm": 0.00018441885185893625, | |
| "learning_rate": 8.514261387824607e-08, | |
| "loss": 0.0, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 49.013103448275864, | |
| "grad_norm": 0.0018990112002938986, | |
| "learning_rate": 7.662835249042146e-08, | |
| "loss": 0.0001, | |
| "step": 12960 | |
| }, | |
| { | |
| "epoch": 49.01386973180077, | |
| "grad_norm": 0.0002298293838975951, | |
| "learning_rate": 6.811409110259686e-08, | |
| "loss": 0.0001, | |
| "step": 12970 | |
| }, | |
| { | |
| "epoch": 49.01463601532567, | |
| "grad_norm": 24.46701431274414, | |
| "learning_rate": 5.9599829714772246e-08, | |
| "loss": 0.5053, | |
| "step": 12980 | |
| }, | |
| { | |
| "epoch": 49.015402298850574, | |
| "grad_norm": 0.0014313425635918975, | |
| "learning_rate": 5.108556832694764e-08, | |
| "loss": 0.0001, | |
| "step": 12990 | |
| }, | |
| { | |
| "epoch": 49.01616858237548, | |
| "grad_norm": 0.0029820152558386326, | |
| "learning_rate": 4.2571306939123034e-08, | |
| "loss": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 49.01693486590038, | |
| "grad_norm": 0.006488497368991375, | |
| "learning_rate": 3.405704555129843e-08, | |
| "loss": 0.0002, | |
| "step": 13010 | |
| }, | |
| { | |
| "epoch": 49.017701149425285, | |
| "grad_norm": 95.9815444946289, | |
| "learning_rate": 2.554278416347382e-08, | |
| "loss": 0.7591, | |
| "step": 13020 | |
| }, | |
| { | |
| "epoch": 49.01846743295019, | |
| "grad_norm": 0.0002131000073859468, | |
| "learning_rate": 1.7028522775649215e-08, | |
| "loss": 0.0001, | |
| "step": 13030 | |
| }, | |
| { | |
| "epoch": 49.01923371647509, | |
| "grad_norm": 0.002506563439965248, | |
| "learning_rate": 8.514261387824608e-09, | |
| "loss": 0.0, | |
| "step": 13040 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "grad_norm": 0.001619683695025742, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.796635866165161, | |
| "eval_runtime": 30.3562, | |
| "eval_samples_per_second": 1.482, | |
| "eval_steps_per_second": 1.482, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "step": 13050, | |
| "total_flos": 5.730289341462282e+19, | |
| "train_loss": 0.5461062897654483, | |
| "train_runtime": 22302.3156, | |
| "train_samples_per_second": 0.585, | |
| "train_steps_per_second": 0.585 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.1094002723693848, | |
| "eval_runtime": 27.4144, | |
| "eval_samples_per_second": 1.641, | |
| "eval_steps_per_second": 1.641, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.109400510787964, | |
| "eval_runtime": 27.2642, | |
| "eval_samples_per_second": 1.651, | |
| "eval_steps_per_second": 1.651, | |
| "step": 13050 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 13050, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.730289341462282e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |