| { | |
| "best_metric": 0.9882, | |
| "best_model_checkpoint": "beit-base-patch16-224-finetuned-eurosat/checkpoint-2109", | |
| "epoch": 2.998933522929257, | |
| "eval_steps": 500, | |
| "global_step": 2109, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 42.045841217041016, | |
| "learning_rate": 2.3696682464454976e-06, | |
| "loss": 2.6733, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 19.79863166809082, | |
| "learning_rate": 4.739336492890995e-06, | |
| "loss": 2.5329, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 21.456212997436523, | |
| "learning_rate": 7.109004739336493e-06, | |
| "loss": 2.3177, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 24.893077850341797, | |
| "learning_rate": 9.47867298578199e-06, | |
| "loss": 2.0632, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 25.683191299438477, | |
| "learning_rate": 1.184834123222749e-05, | |
| "loss": 1.7585, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 27.140174865722656, | |
| "learning_rate": 1.4218009478672985e-05, | |
| "loss": 1.4911, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 17.97708511352539, | |
| "learning_rate": 1.6587677725118483e-05, | |
| "loss": 1.2294, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 36.129730224609375, | |
| "learning_rate": 1.895734597156398e-05, | |
| "loss": 1.0185, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 22.884685516357422, | |
| "learning_rate": 2.132701421800948e-05, | |
| "loss": 0.8923, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 19.778934478759766, | |
| "learning_rate": 2.369668246445498e-05, | |
| "loss": 0.7784, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 20.47595977783203, | |
| "learning_rate": 2.6066350710900477e-05, | |
| "loss": 0.6687, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 17.096622467041016, | |
| "learning_rate": 2.843601895734597e-05, | |
| "loss": 0.6549, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 18.601505279541016, | |
| "learning_rate": 3.080568720379147e-05, | |
| "loss": 0.6253, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 15.65523624420166, | |
| "learning_rate": 3.3175355450236966e-05, | |
| "loss": 0.6118, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 15.892389297485352, | |
| "learning_rate": 3.554502369668247e-05, | |
| "loss": 0.6227, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 12.784531593322754, | |
| "learning_rate": 3.791469194312796e-05, | |
| "loss": 0.4946, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 13.44602108001709, | |
| "learning_rate": 4.028436018957346e-05, | |
| "loss": 0.6287, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 17.538671493530273, | |
| "learning_rate": 4.265402843601896e-05, | |
| "loss": 0.5645, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 11.633671760559082, | |
| "learning_rate": 4.502369668246446e-05, | |
| "loss": 0.5169, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 13.391677856445312, | |
| "learning_rate": 4.739336492890996e-05, | |
| "loss": 0.6072, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 8.77043342590332, | |
| "learning_rate": 4.976303317535545e-05, | |
| "loss": 0.4768, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 11.799147605895996, | |
| "learning_rate": 4.976290832455216e-05, | |
| "loss": 0.4924, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.894740104675293, | |
| "learning_rate": 4.949947312961012e-05, | |
| "loss": 0.4062, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 11.966766357421875, | |
| "learning_rate": 4.923603793466807e-05, | |
| "loss": 0.5118, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 12.86320686340332, | |
| "learning_rate": 4.8972602739726034e-05, | |
| "loss": 0.5241, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.136141777038574, | |
| "learning_rate": 4.8709167544783986e-05, | |
| "loss": 0.4875, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 12.6289644241333, | |
| "learning_rate": 4.8445732349841945e-05, | |
| "loss": 0.482, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 11.868968963623047, | |
| "learning_rate": 4.8182297154899896e-05, | |
| "loss": 0.5488, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 11.955994606018066, | |
| "learning_rate": 4.791886195995785e-05, | |
| "loss": 0.4242, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 10.719066619873047, | |
| "learning_rate": 4.7655426765015806e-05, | |
| "loss": 0.4832, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 9.737288475036621, | |
| "learning_rate": 4.7391991570073765e-05, | |
| "loss": 0.4619, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.085290908813477, | |
| "learning_rate": 4.712855637513172e-05, | |
| "loss": 0.4583, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.268682956695557, | |
| "learning_rate": 4.6865121180189675e-05, | |
| "loss": 0.4926, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 8.957120895385742, | |
| "learning_rate": 4.6601685985247633e-05, | |
| "loss": 0.4615, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.488365650177002, | |
| "learning_rate": 4.6338250790305585e-05, | |
| "loss": 0.4487, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 9.455901145935059, | |
| "learning_rate": 4.6074815595363544e-05, | |
| "loss": 0.4429, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 10.715672492980957, | |
| "learning_rate": 4.58113804004215e-05, | |
| "loss": 0.4603, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 10.67427921295166, | |
| "learning_rate": 4.5547945205479454e-05, | |
| "loss": 0.3811, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 7.144228935241699, | |
| "learning_rate": 4.528451001053741e-05, | |
| "loss": 0.4789, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 10.277647972106934, | |
| "learning_rate": 4.5021074815595364e-05, | |
| "loss": 0.4265, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 9.004936218261719, | |
| "learning_rate": 4.4757639620653316e-05, | |
| "loss": 0.468, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 8.17403507232666, | |
| "learning_rate": 4.449420442571128e-05, | |
| "loss": 0.3897, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 7.23596715927124, | |
| "learning_rate": 4.423076923076923e-05, | |
| "loss": 0.3734, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 8.713665962219238, | |
| "learning_rate": 4.396733403582719e-05, | |
| "loss": 0.4557, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 9.222480773925781, | |
| "learning_rate": 4.370389884088514e-05, | |
| "loss": 0.3975, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 8.174620628356934, | |
| "learning_rate": 4.34404636459431e-05, | |
| "loss": 0.4742, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 10.548660278320312, | |
| "learning_rate": 4.317702845100105e-05, | |
| "loss": 0.4072, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 10.924005508422852, | |
| "learning_rate": 4.291359325605901e-05, | |
| "loss": 0.3547, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 7.682478427886963, | |
| "learning_rate": 4.265015806111697e-05, | |
| "loss": 0.3249, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 9.409101486206055, | |
| "learning_rate": 4.238672286617492e-05, | |
| "loss": 0.3852, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 8.597813606262207, | |
| "learning_rate": 4.212328767123288e-05, | |
| "loss": 0.3998, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 8.316597938537598, | |
| "learning_rate": 4.185985247629083e-05, | |
| "loss": 0.4554, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 7.785365581512451, | |
| "learning_rate": 4.159641728134879e-05, | |
| "loss": 0.4193, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 9.083528518676758, | |
| "learning_rate": 4.133298208640675e-05, | |
| "loss": 0.3247, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 6.537896633148193, | |
| "learning_rate": 4.10695468914647e-05, | |
| "loss": 0.3701, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.992751598358154, | |
| "learning_rate": 4.080611169652266e-05, | |
| "loss": 0.3539, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 9.362491607666016, | |
| "learning_rate": 4.054267650158061e-05, | |
| "loss": 0.3521, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 9.023208618164062, | |
| "learning_rate": 4.027924130663857e-05, | |
| "loss": 0.3028, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 11.851508140563965, | |
| "learning_rate": 4.001580611169653e-05, | |
| "loss": 0.3854, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 11.16539192199707, | |
| "learning_rate": 3.975237091675448e-05, | |
| "loss": 0.4021, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 7.1588311195373535, | |
| "learning_rate": 3.948893572181244e-05, | |
| "loss": 0.3865, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 8.23454761505127, | |
| "learning_rate": 3.922550052687039e-05, | |
| "loss": 0.3941, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 9.590554237365723, | |
| "learning_rate": 3.896206533192835e-05, | |
| "loss": 0.3831, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 7.980281829833984, | |
| "learning_rate": 3.86986301369863e-05, | |
| "loss": 0.4134, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 6.231551647186279, | |
| "learning_rate": 3.843519494204426e-05, | |
| "loss": 0.3746, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 8.723283767700195, | |
| "learning_rate": 3.8171759747102217e-05, | |
| "loss": 0.3349, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 7.533834934234619, | |
| "learning_rate": 3.790832455216017e-05, | |
| "loss": 0.3229, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 6.750338554382324, | |
| "learning_rate": 3.764488935721813e-05, | |
| "loss": 0.3163, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 11.813076972961426, | |
| "learning_rate": 3.738145416227608e-05, | |
| "loss": 0.4089, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.000370979309082, | |
| "learning_rate": 3.711801896733404e-05, | |
| "loss": 0.4327, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9774, | |
| "eval_loss": 0.06401650607585907, | |
| "eval_runtime": 65.8888, | |
| "eval_samples_per_second": 75.885, | |
| "eval_steps_per_second": 4.75, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 7.698484420776367, | |
| "learning_rate": 3.6854583772391995e-05, | |
| "loss": 0.359, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 10.548850059509277, | |
| "learning_rate": 3.6591148577449954e-05, | |
| "loss": 0.3162, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 10.149295806884766, | |
| "learning_rate": 3.6327713382507905e-05, | |
| "loss": 0.3265, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 6.814464092254639, | |
| "learning_rate": 3.606427818756586e-05, | |
| "loss": 0.3536, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 8.711482048034668, | |
| "learning_rate": 3.5800842992623816e-05, | |
| "loss": 0.379, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 8.469487190246582, | |
| "learning_rate": 3.5537407797681774e-05, | |
| "loss": 0.3117, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 9.90764045715332, | |
| "learning_rate": 3.527397260273973e-05, | |
| "loss": 0.3177, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 7.835967063903809, | |
| "learning_rate": 3.5010537407797684e-05, | |
| "loss": 0.347, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 5.563517093658447, | |
| "learning_rate": 3.4747102212855636e-05, | |
| "loss": 0.3244, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 10.490474700927734, | |
| "learning_rate": 3.4483667017913594e-05, | |
| "loss": 0.3159, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 7.063011169433594, | |
| "learning_rate": 3.4220231822971546e-05, | |
| "loss": 0.2711, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 9.191926002502441, | |
| "learning_rate": 3.3956796628029505e-05, | |
| "loss": 0.3441, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 5.757563591003418, | |
| "learning_rate": 3.369336143308746e-05, | |
| "loss": 0.3562, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 5.849490642547607, | |
| "learning_rate": 3.342992623814542e-05, | |
| "loss": 0.2797, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 5.61791467666626, | |
| "learning_rate": 3.316649104320337e-05, | |
| "loss": 0.3362, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 7.7571916580200195, | |
| "learning_rate": 3.2903055848261325e-05, | |
| "loss": 0.2682, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 9.108525276184082, | |
| "learning_rate": 3.2639620653319283e-05, | |
| "loss": 0.3755, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 5.915694713592529, | |
| "learning_rate": 3.237618545837724e-05, | |
| "loss": 0.3061, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 7.744335651397705, | |
| "learning_rate": 3.21127502634352e-05, | |
| "loss": 0.367, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 9.851384162902832, | |
| "learning_rate": 3.184931506849315e-05, | |
| "loss": 0.3318, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 6.104685306549072, | |
| "learning_rate": 3.1585879873551104e-05, | |
| "loss": 0.3477, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 7.819393157958984, | |
| "learning_rate": 3.132244467860906e-05, | |
| "loss": 0.2994, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 9.686634063720703, | |
| "learning_rate": 3.105900948366702e-05, | |
| "loss": 0.2885, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 4.374415397644043, | |
| "learning_rate": 3.079557428872498e-05, | |
| "loss": 0.3053, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 9.000449180603027, | |
| "learning_rate": 3.053213909378293e-05, | |
| "loss": 0.2896, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 6.736356258392334, | |
| "learning_rate": 3.026870389884089e-05, | |
| "loss": 0.2897, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 6.521255016326904, | |
| "learning_rate": 3.000526870389884e-05, | |
| "loss": 0.3365, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 7.526865482330322, | |
| "learning_rate": 2.9741833508956796e-05, | |
| "loss": 0.2817, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 3.9108359813690186, | |
| "learning_rate": 2.9478398314014755e-05, | |
| "loss": 0.3002, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 9.110422134399414, | |
| "learning_rate": 2.921496311907271e-05, | |
| "loss": 0.3055, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 5.4320573806762695, | |
| "learning_rate": 2.8951527924130668e-05, | |
| "loss": 0.3126, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 8.311667442321777, | |
| "learning_rate": 2.868809272918862e-05, | |
| "loss": 0.3177, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 8.694578170776367, | |
| "learning_rate": 2.842465753424658e-05, | |
| "loss": 0.2458, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 6.780071258544922, | |
| "learning_rate": 2.8161222339304533e-05, | |
| "loss": 0.2171, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 8.7374906539917, | |
| "learning_rate": 2.7897787144362485e-05, | |
| "loss": 0.2914, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 8.430846214294434, | |
| "learning_rate": 2.7634351949420444e-05, | |
| "loss": 0.2981, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 6.208935737609863, | |
| "learning_rate": 2.73709167544784e-05, | |
| "loss": 0.2859, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 6.287723541259766, | |
| "learning_rate": 2.7107481559536357e-05, | |
| "loss": 0.2559, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 9.158524513244629, | |
| "learning_rate": 2.6844046364594312e-05, | |
| "loss": 0.2544, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 8.676222801208496, | |
| "learning_rate": 2.6580611169652264e-05, | |
| "loss": 0.2648, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 5.815190315246582, | |
| "learning_rate": 2.6317175974710222e-05, | |
| "loss": 0.2794, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 7.133479595184326, | |
| "learning_rate": 2.6053740779768177e-05, | |
| "loss": 0.3008, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 9.465476989746094, | |
| "learning_rate": 2.5790305584826136e-05, | |
| "loss": 0.3032, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 5.298926830291748, | |
| "learning_rate": 2.5526870389884088e-05, | |
| "loss": 0.2616, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 3.1998720169067383, | |
| "learning_rate": 2.5263435194942046e-05, | |
| "loss": 0.3016, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 9.716959953308105, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2625, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 9.040936470031738, | |
| "learning_rate": 2.4736564805057956e-05, | |
| "loss": 0.2634, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 7.872269630432129, | |
| "learning_rate": 2.4473129610115915e-05, | |
| "loss": 0.2744, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 4.666429042816162, | |
| "learning_rate": 2.420969441517387e-05, | |
| "loss": 0.2552, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 5.423487663269043, | |
| "learning_rate": 2.394625922023182e-05, | |
| "loss": 0.3138, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 5.613728046417236, | |
| "learning_rate": 2.368282402528978e-05, | |
| "loss": 0.3048, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 5.959019660949707, | |
| "learning_rate": 2.3419388830347735e-05, | |
| "loss": 0.2784, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 5.206528663635254, | |
| "learning_rate": 2.315595363540569e-05, | |
| "loss": 0.3245, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 6.318498611450195, | |
| "learning_rate": 2.289251844046365e-05, | |
| "loss": 0.299, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 5.1116437911987305, | |
| "learning_rate": 2.2629083245521604e-05, | |
| "loss": 0.2706, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 9.792173385620117, | |
| "learning_rate": 2.236564805057956e-05, | |
| "loss": 0.2698, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 5.214483737945557, | |
| "learning_rate": 2.2102212855637514e-05, | |
| "loss": 0.2968, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 5.681826114654541, | |
| "learning_rate": 2.183877766069547e-05, | |
| "loss": 0.2893, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 4.573737621307373, | |
| "learning_rate": 2.1575342465753427e-05, | |
| "loss": 0.2424, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 5.404233932495117, | |
| "learning_rate": 2.1311907270811383e-05, | |
| "loss": 0.2956, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 7.8233160972595215, | |
| "learning_rate": 2.1048472075869338e-05, | |
| "loss": 0.2992, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 5.9538373947143555, | |
| "learning_rate": 2.0785036880927293e-05, | |
| "loss": 0.2843, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 6.776533603668213, | |
| "learning_rate": 2.0521601685985248e-05, | |
| "loss": 0.2255, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 4.279127597808838, | |
| "learning_rate": 2.0258166491043203e-05, | |
| "loss": 0.2791, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 9.389575958251953, | |
| "learning_rate": 1.999473129610116e-05, | |
| "loss": 0.3064, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 6.343210697174072, | |
| "learning_rate": 1.9731296101159116e-05, | |
| "loss": 0.215, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 6.596449375152588, | |
| "learning_rate": 1.946786090621707e-05, | |
| "loss": 0.2429, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 6.299315929412842, | |
| "learning_rate": 1.9204425711275027e-05, | |
| "loss": 0.3099, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 7.726513385772705, | |
| "learning_rate": 1.894099051633298e-05, | |
| "loss": 0.2435, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 7.626183986663818, | |
| "learning_rate": 1.8677555321390937e-05, | |
| "loss": 0.2814, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9834, | |
| "eval_loss": 0.048424676060676575, | |
| "eval_runtime": 66.0101, | |
| "eval_samples_per_second": 75.746, | |
| "eval_steps_per_second": 4.742, | |
| "step": 1406 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.147545337677002, | |
| "learning_rate": 1.8414120126448895e-05, | |
| "loss": 0.2343, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.9744009971618652, | |
| "learning_rate": 1.815068493150685e-05, | |
| "loss": 0.2542, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 7.23543643951416, | |
| "learning_rate": 1.7887249736564805e-05, | |
| "loss": 0.2568, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 4.5337395668029785, | |
| "learning_rate": 1.7623814541622764e-05, | |
| "loss": 0.2348, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 7.225223541259766, | |
| "learning_rate": 1.7360379346680716e-05, | |
| "loss": 0.2861, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 6.333385467529297, | |
| "learning_rate": 1.7096944151738674e-05, | |
| "loss": 0.2237, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 9.622603416442871, | |
| "learning_rate": 1.683350895679663e-05, | |
| "loss": 0.2707, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.6370978355407715, | |
| "learning_rate": 1.6570073761854584e-05, | |
| "loss": 0.2177, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 5.092869758605957, | |
| "learning_rate": 1.630663856691254e-05, | |
| "loss": 0.174, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 7.764402389526367, | |
| "learning_rate": 1.6043203371970498e-05, | |
| "loss": 0.2166, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 5.907904624938965, | |
| "learning_rate": 1.577976817702845e-05, | |
| "loss": 0.219, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 5.130086898803711, | |
| "learning_rate": 1.5516332982086408e-05, | |
| "loss": 0.1956, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 7.971607685089111, | |
| "learning_rate": 1.5252897787144363e-05, | |
| "loss": 0.1871, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 10.202670097351074, | |
| "learning_rate": 1.498946259220232e-05, | |
| "loss": 0.2444, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 5.461312770843506, | |
| "learning_rate": 1.4726027397260275e-05, | |
| "loss": 0.2538, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 8.133660316467285, | |
| "learning_rate": 1.4462592202318232e-05, | |
| "loss": 0.2324, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 7.940431594848633, | |
| "learning_rate": 1.4199157007376185e-05, | |
| "loss": 0.2264, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 6.688675403594971, | |
| "learning_rate": 1.3935721812434142e-05, | |
| "loss": 0.199, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 7.409407138824463, | |
| "learning_rate": 1.3672286617492097e-05, | |
| "loss": 0.2311, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 9.85133171081543, | |
| "learning_rate": 1.3408851422550054e-05, | |
| "loss": 0.2153, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 6.340139865875244, | |
| "learning_rate": 1.3145416227608009e-05, | |
| "loss": 0.2717, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 5.589156627655029, | |
| "learning_rate": 1.2881981032665966e-05, | |
| "loss": 0.2068, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 8.83266830444336, | |
| "learning_rate": 1.2618545837723922e-05, | |
| "loss": 0.2459, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 6.671909809112549, | |
| "learning_rate": 1.2355110642781877e-05, | |
| "loss": 0.2332, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 8.771206855773926, | |
| "learning_rate": 1.209167544783983e-05, | |
| "loss": 0.2526, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 9.015884399414062, | |
| "learning_rate": 1.1828240252897788e-05, | |
| "loss": 0.2239, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 9.24109172821045, | |
| "learning_rate": 1.1564805057955744e-05, | |
| "loss": 0.2304, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 6.8663330078125, | |
| "learning_rate": 1.1301369863013698e-05, | |
| "loss": 0.2107, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 5.675321578979492, | |
| "learning_rate": 1.1037934668071655e-05, | |
| "loss": 0.2607, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 6.900195598602295, | |
| "learning_rate": 1.0774499473129611e-05, | |
| "loss": 0.2323, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 11.515998840332031, | |
| "learning_rate": 1.0511064278187566e-05, | |
| "loss": 0.2704, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 7.85768461227417, | |
| "learning_rate": 1.0247629083245521e-05, | |
| "loss": 0.2175, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 4.934202194213867, | |
| "learning_rate": 9.984193888303478e-06, | |
| "loss": 0.2069, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 5.035346031188965, | |
| "learning_rate": 9.720758693361433e-06, | |
| "loss": 0.2145, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 7.098711967468262, | |
| "learning_rate": 9.457323498419388e-06, | |
| "loss": 0.2642, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 7.717360019683838, | |
| "learning_rate": 9.193888303477345e-06, | |
| "loss": 0.1801, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 7.103116035461426, | |
| "learning_rate": 8.930453108535302e-06, | |
| "loss": 0.212, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 4.945562839508057, | |
| "learning_rate": 8.667017913593255e-06, | |
| "loss": 0.2158, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 8.597467422485352, | |
| "learning_rate": 8.403582718651212e-06, | |
| "loss": 0.233, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 3.4780356884002686, | |
| "learning_rate": 8.140147523709169e-06, | |
| "loss": 0.2139, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 8.698033332824707, | |
| "learning_rate": 7.876712328767124e-06, | |
| "loss": 0.2285, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 8.690217971801758, | |
| "learning_rate": 7.613277133825079e-06, | |
| "loss": 0.221, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 6.253328800201416, | |
| "learning_rate": 7.349841938883036e-06, | |
| "loss": 0.2156, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 6.390902042388916, | |
| "learning_rate": 7.08640674394099e-06, | |
| "loss": 0.2134, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 3.520571231842041, | |
| "learning_rate": 6.822971548998947e-06, | |
| "loss": 0.2177, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 4.312754154205322, | |
| "learning_rate": 6.559536354056903e-06, | |
| "loss": 0.1913, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 4.5632758140563965, | |
| "learning_rate": 6.296101159114858e-06, | |
| "loss": 0.2014, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 8.447988510131836, | |
| "learning_rate": 6.032665964172814e-06, | |
| "loss": 0.2445, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 4.897846698760986, | |
| "learning_rate": 5.76923076923077e-06, | |
| "loss": 0.1689, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 6.561771392822266, | |
| "learning_rate": 5.505795574288726e-06, | |
| "loss": 0.2089, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 8.186767578125, | |
| "learning_rate": 5.242360379346681e-06, | |
| "loss": 0.2337, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 3.8493165969848633, | |
| "learning_rate": 4.978925184404637e-06, | |
| "loss": 0.2202, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 4.461836814880371, | |
| "learning_rate": 4.715489989462593e-06, | |
| "loss": 0.2052, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 6.122828483581543, | |
| "learning_rate": 4.452054794520548e-06, | |
| "loss": 0.2447, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 7.432437419891357, | |
| "learning_rate": 4.188619599578504e-06, | |
| "loss": 0.2148, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 5.491920471191406, | |
| "learning_rate": 3.92518440463646e-06, | |
| "loss": 0.2425, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 5.547910690307617, | |
| "learning_rate": 3.661749209694415e-06, | |
| "loss": 0.2409, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 7.146701335906982, | |
| "learning_rate": 3.398314014752371e-06, | |
| "loss": 0.2349, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 9.67960262298584, | |
| "learning_rate": 3.1348788198103265e-06, | |
| "loss": 0.2119, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 7.780117511749268, | |
| "learning_rate": 2.8714436248682825e-06, | |
| "loss": 0.225, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 7.887894153594971, | |
| "learning_rate": 2.6080084299262384e-06, | |
| "loss": 0.1839, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 3.4736483097076416, | |
| "learning_rate": 2.3445732349841943e-06, | |
| "loss": 0.1897, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 3.9504001140594482, | |
| "learning_rate": 2.08113804004215e-06, | |
| "loss": 0.1665, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 3.5273189544677734, | |
| "learning_rate": 1.8177028451001056e-06, | |
| "loss": 0.1694, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 6.938018321990967, | |
| "learning_rate": 1.554267650158061e-06, | |
| "loss": 0.2071, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 3.9890809059143066, | |
| "learning_rate": 1.290832455216017e-06, | |
| "loss": 0.1837, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 5.470897197723389, | |
| "learning_rate": 1.0273972602739725e-06, | |
| "loss": 0.1977, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 7.521297931671143, | |
| "learning_rate": 7.639620653319284e-07, | |
| "loss": 0.2391, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 5.917148590087891, | |
| "learning_rate": 5.00526870389884e-07, | |
| "loss": 0.1858, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 5.207727909088135, | |
| "learning_rate": 2.3709167544783985e-07, | |
| "loss": 0.1655, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9882, | |
| "eval_loss": 0.03509841486811638, | |
| "eval_runtime": 65.8746, | |
| "eval_samples_per_second": 75.902, | |
| "eval_steps_per_second": 4.751, | |
| "step": 2109 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2109, | |
| "total_flos": 1.045459878817923e+19, | |
| "train_loss": 0.37951253058504525, | |
| "train_runtime": 4954.8349, | |
| "train_samples_per_second": 27.246, | |
| "train_steps_per_second": 0.426 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2109, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 1.045459878817923e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |