| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.476275738585497, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008952551477170993, | |
| "grad_norm": 6.379906177520752, | |
| "learning_rate": 9e-06, | |
| "loss": 0.9258, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.017905102954341987, | |
| "grad_norm": 3.932647943496704, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.3837, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02685765443151298, | |
| "grad_norm": 2.0510828495025635, | |
| "learning_rate": 2.9e-05, | |
| "loss": 0.293, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03581020590868397, | |
| "grad_norm": 2.784146785736084, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 0.2134, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.044762757385854966, | |
| "grad_norm": 1.2497658729553223, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.2007, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05371530886302596, | |
| "grad_norm": 1.8504348993301392, | |
| "learning_rate": 5.9e-05, | |
| "loss": 0.1805, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06266786034019696, | |
| "grad_norm": 0.86327064037323, | |
| "learning_rate": 6.9e-05, | |
| "loss": 0.1332, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07162041181736795, | |
| "grad_norm": 0.8830410242080688, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 0.1299, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08057296329453895, | |
| "grad_norm": 1.671918511390686, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 0.1184, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08952551477170993, | |
| "grad_norm": 0.9165273308753967, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 0.116, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09847806624888093, | |
| "grad_norm": 0.9304006099700928, | |
| "learning_rate": 9.999446382823013e-05, | |
| "loss": 0.1115, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10743061772605192, | |
| "grad_norm": 1.2356555461883545, | |
| "learning_rate": 9.997532801828658e-05, | |
| "loss": 0.0999, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11638316920322292, | |
| "grad_norm": 1.0443627834320068, | |
| "learning_rate": 9.99425294526634e-05, | |
| "loss": 0.1, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.12533572068039392, | |
| "grad_norm": 0.5913493633270264, | |
| "learning_rate": 9.989607709816091e-05, | |
| "loss": 0.099, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.13428827215756492, | |
| "grad_norm": 0.9996615052223206, | |
| "learning_rate": 9.983598365438902e-05, | |
| "loss": 0.0886, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1432408236347359, | |
| "grad_norm": 0.839759111404419, | |
| "learning_rate": 9.976226555029522e-05, | |
| "loss": 0.0909, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1521933751119069, | |
| "grad_norm": 1.140505313873291, | |
| "learning_rate": 9.967494293967312e-05, | |
| "loss": 0.0901, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1611459265890779, | |
| "grad_norm": 0.7392069101333618, | |
| "learning_rate": 9.95740396956525e-05, | |
| "loss": 0.0861, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1700984780662489, | |
| "grad_norm": 0.45939213037490845, | |
| "learning_rate": 9.945958340417283e-05, | |
| "loss": 0.0765, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.17905102954341987, | |
| "grad_norm": 0.706051766872406, | |
| "learning_rate": 9.93316053564413e-05, | |
| "loss": 0.0777, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18800358102059087, | |
| "grad_norm": 1.0259908437728882, | |
| "learning_rate": 9.919014054037836e-05, | |
| "loss": 0.0806, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.19695613249776187, | |
| "grad_norm": 0.5664586424827576, | |
| "learning_rate": 9.903522763105218e-05, | |
| "loss": 0.0721, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.20590868397493287, | |
| "grad_norm": 0.6686021685600281, | |
| "learning_rate": 9.886690898010535e-05, | |
| "loss": 0.0714, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21486123545210384, | |
| "grad_norm": 0.6658005118370056, | |
| "learning_rate": 9.868523060417646e-05, | |
| "loss": 0.0748, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.22381378692927484, | |
| "grad_norm": 0.6819655299186707, | |
| "learning_rate": 9.849024217231935e-05, | |
| "loss": 0.0666, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.23276633840644584, | |
| "grad_norm": 0.8272667527198792, | |
| "learning_rate": 9.82819969924244e-05, | |
| "loss": 0.0673, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.24171888988361684, | |
| "grad_norm": 1.0225334167480469, | |
| "learning_rate": 9.806055199664446e-05, | |
| "loss": 0.0721, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.25067144136078784, | |
| "grad_norm": 0.8607687950134277, | |
| "learning_rate": 9.782596772583026e-05, | |
| "loss": 0.0693, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.25962399283795884, | |
| "grad_norm": 0.7404839992523193, | |
| "learning_rate": 9.757830831297914e-05, | |
| "loss": 0.071, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.26857654431512984, | |
| "grad_norm": 0.7764973640441895, | |
| "learning_rate": 9.731764146570173e-05, | |
| "loss": 0.0652, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2775290957923008, | |
| "grad_norm": 0.5465092062950134, | |
| "learning_rate": 9.704403844771128e-05, | |
| "loss": 0.0759, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2864816472694718, | |
| "grad_norm": 0.6784430146217346, | |
| "learning_rate": 9.675757405934103e-05, | |
| "loss": 0.0638, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2954341987466428, | |
| "grad_norm": 0.7917226552963257, | |
| "learning_rate": 9.645832661709444e-05, | |
| "loss": 0.0647, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3043867502238138, | |
| "grad_norm": 1.0117071866989136, | |
| "learning_rate": 9.614637793223425e-05, | |
| "loss": 0.0676, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3133393017009848, | |
| "grad_norm": 0.8718525767326355, | |
| "learning_rate": 9.582181328841611e-05, | |
| "loss": 0.0632, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3222918531781558, | |
| "grad_norm": 0.7445501089096069, | |
| "learning_rate": 9.548472141837286e-05, | |
| "loss": 0.0697, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3312444046553268, | |
| "grad_norm": 0.6769773960113525, | |
| "learning_rate": 9.513519447965595e-05, | |
| "loss": 0.0698, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3401969561324978, | |
| "grad_norm": 0.5535451769828796, | |
| "learning_rate": 9.477332802944044e-05, | |
| "loss": 0.0634, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.34914950760966873, | |
| "grad_norm": 0.5629348158836365, | |
| "learning_rate": 9.439922099840054e-05, | |
| "loss": 0.0608, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.35810205908683973, | |
| "grad_norm": 0.585233211517334, | |
| "learning_rate": 9.401297566366318e-05, | |
| "loss": 0.0543, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36705461056401073, | |
| "grad_norm": 0.7889137268066406, | |
| "learning_rate": 9.36146976208462e-05, | |
| "loss": 0.0629, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.37600716204118173, | |
| "grad_norm": 0.3033806085586548, | |
| "learning_rate": 9.320449575518972e-05, | |
| "loss": 0.0557, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.38495971351835273, | |
| "grad_norm": 0.47413498163223267, | |
| "learning_rate": 9.278248221178798e-05, | |
| "loss": 0.06, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.39391226499552373, | |
| "grad_norm": 0.444699764251709, | |
| "learning_rate": 9.234877236492997e-05, | |
| "loss": 0.0546, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.40286481647269473, | |
| "grad_norm": 0.5341030955314636, | |
| "learning_rate": 9.190348478655724e-05, | |
| "loss": 0.052, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.41181736794986573, | |
| "grad_norm": 0.7423449754714966, | |
| "learning_rate": 9.144674121384757e-05, | |
| "loss": 0.0555, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.42076991942703673, | |
| "grad_norm": 0.7645367383956909, | |
| "learning_rate": 9.097866651593317e-05, | |
| "loss": 0.063, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4297224709042077, | |
| "grad_norm": 0.5560976266860962, | |
| "learning_rate": 9.049938865976275e-05, | |
| "loss": 0.0621, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4386750223813787, | |
| "grad_norm": 0.3317752182483673, | |
| "learning_rate": 9.000903867511666e-05, | |
| "loss": 0.0527, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4476275738585497, | |
| "grad_norm": 0.5263016819953918, | |
| "learning_rate": 8.950775061878453e-05, | |
| "loss": 0.0521, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4565801253357207, | |
| "grad_norm": 0.7472353577613831, | |
| "learning_rate": 8.899566153791566e-05, | |
| "loss": 0.0582, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4655326768128917, | |
| "grad_norm": 0.5456057786941528, | |
| "learning_rate": 8.84729114325516e-05, | |
| "loss": 0.0516, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4744852282900627, | |
| "grad_norm": 0.41416245698928833, | |
| "learning_rate": 8.79396432173515e-05, | |
| "loss": 0.0538, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4834377797672337, | |
| "grad_norm": 0.5883777141571045, | |
| "learning_rate": 8.739600268252078e-05, | |
| "loss": 0.0485, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4923903312444047, | |
| "grad_norm": 0.5608022809028625, | |
| "learning_rate": 8.684213845395339e-05, | |
| "loss": 0.0447, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5013428827215757, | |
| "grad_norm": 0.447710245847702, | |
| "learning_rate": 8.627820195259918e-05, | |
| "loss": 0.0471, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5102954341987467, | |
| "grad_norm": 0.6973375082015991, | |
| "learning_rate": 8.570434735306671e-05, | |
| "loss": 0.047, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5192479856759177, | |
| "grad_norm": 0.44058454036712646, | |
| "learning_rate": 8.512073154147362e-05, | |
| "loss": 0.0453, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5282005371530887, | |
| "grad_norm": 0.40700763463974, | |
| "learning_rate": 8.452751407255541e-05, | |
| "loss": 0.0471, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5371530886302597, | |
| "grad_norm": 0.6121862530708313, | |
| "learning_rate": 8.392485712604483e-05, | |
| "loss": 0.0516, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5461056401074306, | |
| "grad_norm": 0.5500701665878296, | |
| "learning_rate": 8.331292546233362e-05, | |
| "loss": 0.0558, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5550581915846016, | |
| "grad_norm": 0.6602392196655273, | |
| "learning_rate": 8.269188637742846e-05, | |
| "loss": 0.0447, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5640107430617726, | |
| "grad_norm": 0.40000760555267334, | |
| "learning_rate": 8.206190965721419e-05, | |
| "loss": 0.0438, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5729632945389436, | |
| "grad_norm": 0.4768178164958954, | |
| "learning_rate": 8.14231675310358e-05, | |
| "loss": 0.0422, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5819158460161146, | |
| "grad_norm": 0.3286081552505493, | |
| "learning_rate": 8.077583462461283e-05, | |
| "loss": 0.0463, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5908683974932856, | |
| "grad_norm": 0.45187002420425415, | |
| "learning_rate": 8.012008791229826e-05, | |
| "loss": 0.0506, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5998209489704566, | |
| "grad_norm": 0.7616209983825684, | |
| "learning_rate": 7.945610666869568e-05, | |
| "loss": 0.0578, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6087735004476276, | |
| "grad_norm": 0.5608704686164856, | |
| "learning_rate": 7.878407241964729e-05, | |
| "loss": 0.0504, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6177260519247986, | |
| "grad_norm": 0.6086702346801758, | |
| "learning_rate": 7.810416889260653e-05, | |
| "loss": 0.0487, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6266786034019696, | |
| "grad_norm": 0.48586443066596985, | |
| "learning_rate": 7.741658196640892e-05, | |
| "loss": 0.0518, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6356311548791406, | |
| "grad_norm": 0.9479012489318848, | |
| "learning_rate": 7.672149962045457e-05, | |
| "loss": 0.056, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6445837063563116, | |
| "grad_norm": 0.5251392722129822, | |
| "learning_rate": 7.60191118833165e-05, | |
| "loss": 0.05, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6535362578334826, | |
| "grad_norm": 0.716193437576294, | |
| "learning_rate": 7.530961078078873e-05, | |
| "loss": 0.0479, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6624888093106536, | |
| "grad_norm": 0.7000860571861267, | |
| "learning_rate": 7.45931902833884e-05, | |
| "loss": 0.0374, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6714413607878246, | |
| "grad_norm": 0.46128469705581665, | |
| "learning_rate": 7.387004625332608e-05, | |
| "loss": 0.0441, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6803939122649956, | |
| "grad_norm": 0.5693628191947937, | |
| "learning_rate": 7.3140376390959e-05, | |
| "loss": 0.0391, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6893464637421666, | |
| "grad_norm": 0.415169894695282, | |
| "learning_rate": 7.240438018074189e-05, | |
| "loss": 0.0378, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6982990152193375, | |
| "grad_norm": 0.5737828016281128, | |
| "learning_rate": 7.166225883668969e-05, | |
| "loss": 0.0401, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7072515666965085, | |
| "grad_norm": 0.43452468514442444, | |
| "learning_rate": 7.091421524736784e-05, | |
| "loss": 0.0364, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7162041181736795, | |
| "grad_norm": 0.4742269515991211, | |
| "learning_rate": 7.016045392042452e-05, | |
| "loss": 0.0388, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7251566696508505, | |
| "grad_norm": 0.4893990159034729, | |
| "learning_rate": 6.940118092668022e-05, | |
| "loss": 0.0392, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7341092211280215, | |
| "grad_norm": 0.6735272407531738, | |
| "learning_rate": 6.863660384379017e-05, | |
| "loss": 0.0457, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7430617726051925, | |
| "grad_norm": 0.36594638228416443, | |
| "learning_rate": 6.786693169949455e-05, | |
| "loss": 0.0423, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7520143240823635, | |
| "grad_norm": 0.46647775173187256, | |
| "learning_rate": 6.709237491447249e-05, | |
| "loss": 0.0404, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7609668755595345, | |
| "grad_norm": 0.49492594599723816, | |
| "learning_rate": 6.631314524481513e-05, | |
| "loss": 0.0442, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7699194270367055, | |
| "grad_norm": 0.4176277816295624, | |
| "learning_rate": 6.552945572413358e-05, | |
| "loss": 0.0422, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7788719785138765, | |
| "grad_norm": 0.4090403616428375, | |
| "learning_rate": 6.474152060531768e-05, | |
| "loss": 0.041, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7878245299910475, | |
| "grad_norm": 0.3255680203437805, | |
| "learning_rate": 6.394955530196147e-05, | |
| "loss": 0.0392, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7967770814682185, | |
| "grad_norm": 0.33741018176078796, | |
| "learning_rate": 6.315377632947115e-05, | |
| "loss": 0.035, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8057296329453895, | |
| "grad_norm": 0.35257619619369507, | |
| "learning_rate": 6.235440124587198e-05, | |
| "loss": 0.0429, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8146821844225605, | |
| "grad_norm": 0.35682690143585205, | |
| "learning_rate": 6.155164859233012e-05, | |
| "loss": 0.0332, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8236347358997315, | |
| "grad_norm": 0.49269065260887146, | |
| "learning_rate": 6.074573783340562e-05, | |
| "loss": 0.039, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8325872873769025, | |
| "grad_norm": 0.35709279775619507, | |
| "learning_rate": 5.9936889297052986e-05, | |
| "loss": 0.0373, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8415398388540735, | |
| "grad_norm": 0.38914117217063904, | |
| "learning_rate": 5.912532411438576e-05, | |
| "loss": 0.033, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8504923903312444, | |
| "grad_norm": 0.5345783829689026, | |
| "learning_rate": 5.831126415922148e-05, | |
| "loss": 0.0464, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8594449418084154, | |
| "grad_norm": 0.5637487769126892, | |
| "learning_rate": 5.74949319874235e-05, | |
| "loss": 0.0353, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8683974932855864, | |
| "grad_norm": 0.3370537757873535, | |
| "learning_rate": 5.667655077605659e-05, | |
| "loss": 0.035, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8773500447627574, | |
| "grad_norm": 0.430148184299469, | |
| "learning_rate": 5.585634426237246e-05, | |
| "loss": 0.0379, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8863025962399284, | |
| "grad_norm": 0.19934052228927612, | |
| "learning_rate": 5.5034536682642224e-05, | |
| "loss": 0.0344, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8952551477170994, | |
| "grad_norm": 0.3409677743911743, | |
| "learning_rate": 5.4211352710852495e-05, | |
| "loss": 0.034, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9042076991942704, | |
| "grad_norm": 0.6933218836784363, | |
| "learning_rate": 9.383125572106752e-05, | |
| "loss": 0.0337, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9131602506714414, | |
| "grad_norm": 0.5733363628387451, | |
| "learning_rate": 9.367117688664791e-05, | |
| "loss": 0.0377, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9221128021486124, | |
| "grad_norm": 0.6528266668319702, | |
| "learning_rate": 9.35091877334763e-05, | |
| "loss": 0.0437, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9310653536257834, | |
| "grad_norm": 0.3507431149482727, | |
| "learning_rate": 9.334529534748297e-05, | |
| "loss": 0.0423, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9400179051029544, | |
| "grad_norm": 0.5123169422149658, | |
| "learning_rate": 9.317950689785188e-05, | |
| "loss": 0.0349, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9489704565801254, | |
| "grad_norm": 0.551750123500824, | |
| "learning_rate": 9.301182963670688e-05, | |
| "loss": 0.0463, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9579230080572964, | |
| "grad_norm": 0.3931156396865845, | |
| "learning_rate": 9.284227089879456e-05, | |
| "loss": 0.0406, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9668755595344674, | |
| "grad_norm": 0.5835707783699036, | |
| "learning_rate": 9.26708381011634e-05, | |
| "loss": 0.0426, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9758281110116384, | |
| "grad_norm": 0.3757207989692688, | |
| "learning_rate": 9.249753874283937e-05, | |
| "loss": 0.0315, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9847806624888094, | |
| "grad_norm": 0.6495594382286072, | |
| "learning_rate": 9.232238040449779e-05, | |
| "loss": 0.0419, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9937332139659804, | |
| "grad_norm": 0.6033176779747009, | |
| "learning_rate": 9.214537074813181e-05, | |
| "loss": 0.0414, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.0026857654431514, | |
| "grad_norm": 0.599250316619873, | |
| "learning_rate": 9.196651751671724e-05, | |
| "loss": 0.0421, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.0116383169203222, | |
| "grad_norm": 0.4811965227127075, | |
| "learning_rate": 9.178582853387384e-05, | |
| "loss": 0.0477, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.0205908683974934, | |
| "grad_norm": 0.6312634944915771, | |
| "learning_rate": 9.160331170352304e-05, | |
| "loss": 0.0385, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.0295434198746642, | |
| "grad_norm": 0.4195096492767334, | |
| "learning_rate": 9.141897500954229e-05, | |
| "loss": 0.0425, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.0384959713518354, | |
| "grad_norm": 0.4496900141239166, | |
| "learning_rate": 9.123282651541576e-05, | |
| "loss": 0.039, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.0474485228290062, | |
| "grad_norm": 0.5114062428474426, | |
| "learning_rate": 9.104487436388161e-05, | |
| "loss": 0.0362, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.0564010743061774, | |
| "grad_norm": 0.46173688769340515, | |
| "learning_rate": 9.085512677657582e-05, | |
| "loss": 0.04, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.0653536257833482, | |
| "grad_norm": 0.5690197348594666, | |
| "learning_rate": 9.066359205367258e-05, | |
| "loss": 0.0424, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.0743061772605191, | |
| "grad_norm": 0.33746033906936646, | |
| "learning_rate": 9.047027857352112e-05, | |
| "loss": 0.0468, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0832587287376902, | |
| "grad_norm": 0.6568852663040161, | |
| "learning_rate": 9.027519479227935e-05, | |
| "loss": 0.0406, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.0922112802148611, | |
| "grad_norm": 0.4321780502796173, | |
| "learning_rate": 9.007834924354383e-05, | |
| "loss": 0.0366, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.1011638316920322, | |
| "grad_norm": 0.5011163949966431, | |
| "learning_rate": 8.987975053797655e-05, | |
| "loss": 0.04, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.1101163831692031, | |
| "grad_norm": 0.4079154133796692, | |
| "learning_rate": 8.967940736292825e-05, | |
| "loss": 0.0389, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.1190689346463742, | |
| "grad_norm": 0.7240653038024902, | |
| "learning_rate": 8.947732848205846e-05, | |
| "loss": 0.0448, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.1280214861235451, | |
| "grad_norm": 0.6067286729812622, | |
| "learning_rate": 8.927352273495204e-05, | |
| "loss": 0.0366, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.1369740376007162, | |
| "grad_norm": 0.5315434336662292, | |
| "learning_rate": 8.906799903673265e-05, | |
| "loss": 0.0406, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.1459265890778871, | |
| "grad_norm": 0.3086467385292053, | |
| "learning_rate": 8.88607663776726e-05, | |
| "loss": 0.0386, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.1548791405550582, | |
| "grad_norm": 0.3972776234149933, | |
| "learning_rate": 8.865183382279978e-05, | |
| "loss": 0.0414, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.1638316920322291, | |
| "grad_norm": 0.28744080662727356, | |
| "learning_rate": 8.844121051150096e-05, | |
| "loss": 0.0355, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.1727842435094002, | |
| "grad_norm": 0.4855988919734955, | |
| "learning_rate": 8.822890565712211e-05, | |
| "loss": 0.036, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.1817367949865711, | |
| "grad_norm": 0.5339410305023193, | |
| "learning_rate": 8.801492854656536e-05, | |
| "loss": 0.0401, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.1906893464637422, | |
| "grad_norm": 0.3985954225063324, | |
| "learning_rate": 8.779928853988268e-05, | |
| "loss": 0.0335, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.1996418979409131, | |
| "grad_norm": 0.40003374218940735, | |
| "learning_rate": 8.758199506986655e-05, | |
| "loss": 0.0376, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.2085944494180842, | |
| "grad_norm": 0.6960264444351196, | |
| "learning_rate": 8.73630576416373e-05, | |
| "loss": 0.0342, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.2175470008952551, | |
| "grad_norm": 0.49340131878852844, | |
| "learning_rate": 8.714248583222726e-05, | |
| "loss": 0.0388, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.226499552372426, | |
| "grad_norm": 0.43804675340652466, | |
| "learning_rate": 8.692028929016196e-05, | |
| "loss": 0.0292, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.2354521038495971, | |
| "grad_norm": 0.37847551703453064, | |
| "learning_rate": 8.669647773503797e-05, | |
| "loss": 0.036, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.2444046553267682, | |
| "grad_norm": 0.35698914527893066, | |
| "learning_rate": 8.647106095709773e-05, | |
| "loss": 0.033, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.2533572068039391, | |
| "grad_norm": 0.4798845648765564, | |
| "learning_rate": 8.624404881680139e-05, | |
| "loss": 0.0357, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.26230975828111, | |
| "grad_norm": 0.5304312109947205, | |
| "learning_rate": 8.601545124439535e-05, | |
| "loss": 0.0326, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.2712623097582811, | |
| "grad_norm": 0.4820861220359802, | |
| "learning_rate": 8.5785278239478e-05, | |
| "loss": 0.0392, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.280214861235452, | |
| "grad_norm": 0.34791550040245056, | |
| "learning_rate": 8.555353987056224e-05, | |
| "loss": 0.0333, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.2891674127126231, | |
| "grad_norm": 0.47185078263282776, | |
| "learning_rate": 8.532024627463505e-05, | |
| "loss": 0.0357, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.298119964189794, | |
| "grad_norm": 0.44588857889175415, | |
| "learning_rate": 8.508540765671407e-05, | |
| "loss": 0.0352, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.3070725156669651, | |
| "grad_norm": 0.4541409909725189, | |
| "learning_rate": 8.484903428940121e-05, | |
| "loss": 0.0376, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.316025067144136, | |
| "grad_norm": 0.45099449157714844, | |
| "learning_rate": 8.461113651243334e-05, | |
| "loss": 0.0383, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.3249776186213071, | |
| "grad_norm": 0.5666273832321167, | |
| "learning_rate": 8.437172473222987e-05, | |
| "loss": 0.0372, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.333930170098478, | |
| "grad_norm": 0.35443374514579773, | |
| "learning_rate": 8.413080942143767e-05, | |
| "loss": 0.0321, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.3428827215756491, | |
| "grad_norm": 0.4038570821285248, | |
| "learning_rate": 8.388840111847288e-05, | |
| "loss": 0.0391, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.35183527305282, | |
| "grad_norm": 0.44889935851097107, | |
| "learning_rate": 8.364451042705998e-05, | |
| "loss": 0.0357, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.3607878245299911, | |
| "grad_norm": 0.24770434200763702, | |
| "learning_rate": 8.33991480157679e-05, | |
| "loss": 0.0348, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.369740376007162, | |
| "grad_norm": 0.2446165382862091, | |
| "learning_rate": 8.315232461754338e-05, | |
| "loss": 0.0351, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.378692927484333, | |
| "grad_norm": 0.36377421021461487, | |
| "learning_rate": 8.290405102924144e-05, | |
| "loss": 0.0389, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.387645478961504, | |
| "grad_norm": 0.38384515047073364, | |
| "learning_rate": 8.265433811115316e-05, | |
| "loss": 0.038, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.3965980304386751, | |
| "grad_norm": 0.5154732465744019, | |
| "learning_rate": 8.240319678653049e-05, | |
| "loss": 0.037, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.405550581915846, | |
| "grad_norm": 0.39229637384414673, | |
| "learning_rate": 8.215063804110857e-05, | |
| "loss": 0.0358, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.414503133393017, | |
| "grad_norm": 0.30531948804855347, | |
| "learning_rate": 8.189667292262512e-05, | |
| "loss": 0.0321, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.423455684870188, | |
| "grad_norm": 0.3962353765964508, | |
| "learning_rate": 8.164131254033716e-05, | |
| "loss": 0.036, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.432408236347359, | |
| "grad_norm": 0.24542206525802612, | |
| "learning_rate": 8.138456806453503e-05, | |
| "loss": 0.0346, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.44136078782453, | |
| "grad_norm": 0.39958980679512024, | |
| "learning_rate": 8.112645072605386e-05, | |
| "loss": 0.0282, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.450313339301701, | |
| "grad_norm": 0.40440255403518677, | |
| "learning_rate": 8.086697181578222e-05, | |
| "loss": 0.0316, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.459265890778872, | |
| "grad_norm": 0.4727575182914734, | |
| "learning_rate": 8.060614268416823e-05, | |
| "loss": 0.0331, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.468218442256043, | |
| "grad_norm": 0.5293134450912476, | |
| "learning_rate": 8.034397474072309e-05, | |
| "loss": 0.0385, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.477170993733214, | |
| "grad_norm": 0.757794976234436, | |
| "learning_rate": 8.008047945352193e-05, | |
| "loss": 0.0384, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.486123545210385, | |
| "grad_norm": 0.5270459055900574, | |
| "learning_rate": 7.981566834870225e-05, | |
| "loss": 0.0278, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.495076096687556, | |
| "grad_norm": 0.44656965136528015, | |
| "learning_rate": 7.954955300995961e-05, | |
| "loss": 0.038, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.504028648164727, | |
| "grad_norm": 0.3250710070133209, | |
| "learning_rate": 7.928214507804104e-05, | |
| "loss": 0.0318, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.512981199641898, | |
| "grad_norm": 0.26046040654182434, | |
| "learning_rate": 7.901345625023576e-05, | |
| "loss": 0.0322, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.521933751119069, | |
| "grad_norm": 0.43262678384780884, | |
| "learning_rate": 7.874349827986354e-05, | |
| "loss": 0.0293, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.5308863025962398, | |
| "grad_norm": 0.44547683000564575, | |
| "learning_rate": 7.847228297576053e-05, | |
| "loss": 0.0369, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.539838854073411, | |
| "grad_norm": 0.39528220891952515, | |
| "learning_rate": 7.819982220176276e-05, | |
| "loss": 0.0331, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.548791405550582, | |
| "grad_norm": 0.522739827632904, | |
| "learning_rate": 7.792612787618714e-05, | |
| "loss": 0.0327, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.557743957027753, | |
| "grad_norm": 0.3753514289855957, | |
| "learning_rate": 7.765121197131009e-05, | |
| "loss": 0.0303, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.5666965085049238, | |
| "grad_norm": 0.41809502243995667, | |
| "learning_rate": 7.737508651284391e-05, | |
| "loss": 0.0334, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.575649059982095, | |
| "grad_norm": 0.43867743015289307, | |
| "learning_rate": 7.709776357941069e-05, | |
| "loss": 0.0351, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.584601611459266, | |
| "grad_norm": 0.4181774854660034, | |
| "learning_rate": 7.681925530201392e-05, | |
| "loss": 0.0337, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.593554162936437, | |
| "grad_norm": 0.40942829847335815, | |
| "learning_rate": 7.65395738635079e-05, | |
| "loss": 0.0322, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.6025067144136078, | |
| "grad_norm": 0.4726716876029968, | |
| "learning_rate": 7.62587314980648e-05, | |
| "loss": 0.0392, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.611459265890779, | |
| "grad_norm": 0.43389004468917847, | |
| "learning_rate": 7.597674049063947e-05, | |
| "loss": 0.0342, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.6204118173679498, | |
| "grad_norm": 0.60209721326828, | |
| "learning_rate": 7.569361317643211e-05, | |
| "loss": 0.0307, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.629364368845121, | |
| "grad_norm": 0.27818477153778076, | |
| "learning_rate": 7.540936194034865e-05, | |
| "loss": 0.0298, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.6383169203222918, | |
| "grad_norm": 0.3243078291416168, | |
| "learning_rate": 7.512399921645901e-05, | |
| "loss": 0.0305, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.6472694717994627, | |
| "grad_norm": 0.5586011409759521, | |
| "learning_rate": 7.483753748745317e-05, | |
| "loss": 0.0316, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.6562220232766338, | |
| "grad_norm": 0.3181556165218353, | |
| "learning_rate": 7.454998928409516e-05, | |
| "loss": 0.0342, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.665174574753805, | |
| "grad_norm": 0.3369368016719818, | |
| "learning_rate": 7.426136718467493e-05, | |
| "loss": 0.0348, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.6741271262309758, | |
| "grad_norm": 0.42150554060935974, | |
| "learning_rate": 7.397168381445812e-05, | |
| "loss": 0.032, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.6830796777081467, | |
| "grad_norm": 0.4693697392940521, | |
| "learning_rate": 7.368095184513377e-05, | |
| "loss": 0.0282, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.6920322291853178, | |
| "grad_norm": 0.33959540724754333, | |
| "learning_rate": 7.338918399426005e-05, | |
| "loss": 0.0253, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.700984780662489, | |
| "grad_norm": 0.3847089409828186, | |
| "learning_rate": 7.309639302470801e-05, | |
| "loss": 0.0342, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.7099373321396598, | |
| "grad_norm": 0.49096646904945374, | |
| "learning_rate": 7.280259174410312e-05, | |
| "loss": 0.0326, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.7188898836168307, | |
| "grad_norm": 0.36844900250434875, | |
| "learning_rate": 7.250779300426517e-05, | |
| "loss": 0.0289, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.7278424350940018, | |
| "grad_norm": 0.3279619514942169, | |
| "learning_rate": 7.22120097006461e-05, | |
| "loss": 0.0335, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.736794986571173, | |
| "grad_norm": 0.4344680905342102, | |
| "learning_rate": 7.191525477176577e-05, | |
| "loss": 0.0305, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.7457475380483438, | |
| "grad_norm": 0.2836930453777313, | |
| "learning_rate": 7.161754119864616e-05, | |
| "loss": 0.0278, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.7547000895255147, | |
| "grad_norm": 0.49985265731811523, | |
| "learning_rate": 7.131888200424339e-05, | |
| "loss": 0.0261, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.7636526410026858, | |
| "grad_norm": 0.40372687578201294, | |
| "learning_rate": 7.101929025287816e-05, | |
| "loss": 0.0309, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.7726051924798567, | |
| "grad_norm": 0.34391847252845764, | |
| "learning_rate": 7.071877904966423e-05, | |
| "loss": 0.0352, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.7815577439570278, | |
| "grad_norm": 0.3783436417579651, | |
| "learning_rate": 7.04173615399351e-05, | |
| "loss": 0.0275, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.7905102954341987, | |
| "grad_norm": 0.35148048400878906, | |
| "learning_rate": 7.011505090866913e-05, | |
| "loss": 0.0278, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.7994628469113696, | |
| "grad_norm": 0.4762101471424103, | |
| "learning_rate": 6.981186037991271e-05, | |
| "loss": 0.0271, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.8084153983885407, | |
| "grad_norm": 0.28804442286491394, | |
| "learning_rate": 6.950780321620174e-05, | |
| "loss": 0.0299, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.8173679498657118, | |
| "grad_norm": 0.4394608438014984, | |
| "learning_rate": 6.920289271798157e-05, | |
| "loss": 0.0316, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.8263205013428827, | |
| "grad_norm": 0.3154217302799225, | |
| "learning_rate": 6.889714222302517e-05, | |
| "loss": 0.0352, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.8352730528200536, | |
| "grad_norm": 0.390591561794281, | |
| "learning_rate": 6.85905651058497e-05, | |
| "loss": 0.0327, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.8442256042972247, | |
| "grad_norm": 0.42770206928253174, | |
| "learning_rate": 6.82831747771314e-05, | |
| "loss": 0.0259, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.8531781557743958, | |
| "grad_norm": 0.28651660680770874, | |
| "learning_rate": 6.797498468311907e-05, | |
| "loss": 0.0258, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.8621307072515667, | |
| "grad_norm": 0.41473740339279175, | |
| "learning_rate": 6.766600830504585e-05, | |
| "loss": 0.0241, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.8710832587287376, | |
| "grad_norm": 0.31760624051094055, | |
| "learning_rate": 6.735625915853942e-05, | |
| "loss": 0.0233, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.8800358102059087, | |
| "grad_norm": 0.41052040457725525, | |
| "learning_rate": 6.70457507930309e-05, | |
| "loss": 0.0253, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.8889883616830798, | |
| "grad_norm": 0.37952539324760437, | |
| "learning_rate": 6.673449679116215e-05, | |
| "loss": 0.025, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.8979409131602507, | |
| "grad_norm": 0.33014625310897827, | |
| "learning_rate": 6.642251076819148e-05, | |
| "loss": 0.023, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.9068934646374216, | |
| "grad_norm": 0.3890449106693268, | |
| "learning_rate": 6.610980637139827e-05, | |
| "loss": 0.0278, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.9158460161145927, | |
| "grad_norm": 0.34647008776664734, | |
| "learning_rate": 6.579639727948583e-05, | |
| "loss": 0.0289, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.9247985675917636, | |
| "grad_norm": 0.21908289194107056, | |
| "learning_rate": 6.548229720198315e-05, | |
| "loss": 0.0349, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.9337511190689347, | |
| "grad_norm": 0.3591015636920929, | |
| "learning_rate": 6.516751987864517e-05, | |
| "loss": 0.0271, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.9427036705461056, | |
| "grad_norm": 0.4022800922393799, | |
| "learning_rate": 6.485207907885175e-05, | |
| "loss": 0.0233, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.9516562220232765, | |
| "grad_norm": 0.6193303465843201, | |
| "learning_rate": 6.453598860100536e-05, | |
| "loss": 0.0304, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.9606087735004476, | |
| "grad_norm": 0.2870684862136841, | |
| "learning_rate": 6.421926227192749e-05, | |
| "loss": 0.0239, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.9695613249776187, | |
| "grad_norm": 0.4261172413825989, | |
| "learning_rate": 6.390191394625381e-05, | |
| "loss": 0.0294, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.9785138764547896, | |
| "grad_norm": 0.17389577627182007, | |
| "learning_rate": 6.358395750582817e-05, | |
| "loss": 0.0242, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.9874664279319605, | |
| "grad_norm": 0.2946527302265167, | |
| "learning_rate": 6.326540685909532e-05, | |
| "loss": 0.0219, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.9964189794091316, | |
| "grad_norm": 0.35331255197525024, | |
| "learning_rate": 6.294627594049249e-05, | |
| "loss": 0.0291, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 2.0053715308863027, | |
| "grad_norm": 0.3004955053329468, | |
| "learning_rate": 6.262657870983989e-05, | |
| "loss": 0.0244, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.0143240823634736, | |
| "grad_norm": 0.28008395433425903, | |
| "learning_rate": 6.230632915173009e-05, | |
| "loss": 0.0272, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.0232766338406445, | |
| "grad_norm": 0.2564995586872101, | |
| "learning_rate": 6.198554127491622e-05, | |
| "loss": 0.0264, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.0322291853178154, | |
| "grad_norm": 0.3245619833469391, | |
| "learning_rate": 6.166422911169923e-05, | |
| "loss": 0.0223, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 2.0411817367949867, | |
| "grad_norm": 0.19854438304901123, | |
| "learning_rate": 6.1342406717314e-05, | |
| "loss": 0.0277, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.0501342882721576, | |
| "grad_norm": 0.1700948178768158, | |
| "learning_rate": 6.102008816931466e-05, | |
| "loss": 0.0248, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 2.0590868397493285, | |
| "grad_norm": 0.2796550989151001, | |
| "learning_rate": 6.069728756695866e-05, | |
| "loss": 0.0247, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.0680393912264994, | |
| "grad_norm": 0.21811795234680176, | |
| "learning_rate": 6.037401903059008e-05, | |
| "loss": 0.0284, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 2.0769919427036707, | |
| "grad_norm": 0.24635659158229828, | |
| "learning_rate": 6.005029670102195e-05, | |
| "loss": 0.0218, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.0859444941808416, | |
| "grad_norm": 0.3100559115409851, | |
| "learning_rate": 5.972613473891766e-05, | |
| "loss": 0.0312, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 2.0948970456580125, | |
| "grad_norm": 0.3434387743473053, | |
| "learning_rate": 5.940154732417158e-05, | |
| "loss": 0.0234, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.1038495971351834, | |
| "grad_norm": 0.4951243996620178, | |
| "learning_rate": 5.907654865528876e-05, | |
| "loss": 0.0305, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.1128021486123547, | |
| "grad_norm": 0.3198152184486389, | |
| "learning_rate": 5.875115294876381e-05, | |
| "loss": 0.0253, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.1217547000895256, | |
| "grad_norm": 0.6087833046913147, | |
| "learning_rate": 5.842537443845908e-05, | |
| "loss": 0.0241, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 2.1307072515666965, | |
| "grad_norm": 0.30654656887054443, | |
| "learning_rate": 5.809922737498198e-05, | |
| "loss": 0.0211, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.1396598030438674, | |
| "grad_norm": 0.4575793445110321, | |
| "learning_rate": 5.777272602506165e-05, | |
| "loss": 0.0286, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.1486123545210383, | |
| "grad_norm": 0.6392210721969604, | |
| "learning_rate": 5.744588467092483e-05, | |
| "loss": 0.0269, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.1575649059982096, | |
| "grad_norm": 0.2926782965660095, | |
| "learning_rate": 5.7118717609671194e-05, | |
| "loss": 0.0225, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.1665174574753805, | |
| "grad_norm": 0.3610515296459198, | |
| "learning_rate": 5.679123915264786e-05, | |
| "loss": 0.0249, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.1754700089525514, | |
| "grad_norm": 0.31393954157829285, | |
| "learning_rate": 5.646346362482342e-05, | |
| "loss": 0.0225, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.1844225604297223, | |
| "grad_norm": 0.41921770572662354, | |
| "learning_rate": 5.613540536416132e-05, | |
| "loss": 0.0261, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.1933751119068936, | |
| "grad_norm": 0.42108452320098877, | |
| "learning_rate": 5.5807078720992645e-05, | |
| "loss": 0.0259, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.2023276633840645, | |
| "grad_norm": 0.43240779638290405, | |
| "learning_rate": 5.547849805738836e-05, | |
| "loss": 0.0274, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.2112802148612354, | |
| "grad_norm": 0.285217821598053, | |
| "learning_rate": 5.514967774653118e-05, | |
| "loss": 0.0217, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.2202327663384063, | |
| "grad_norm": 0.3506968021392822, | |
| "learning_rate": 5.482063217208674e-05, | |
| "loss": 0.0224, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.2291853178155776, | |
| "grad_norm": 0.2749573886394501, | |
| "learning_rate": 5.449137572757439e-05, | |
| "loss": 0.0232, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.2381378692927485, | |
| "grad_norm": 0.3615213930606842, | |
| "learning_rate": 5.4161922815737696e-05, | |
| "loss": 0.0277, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.2470904207699194, | |
| "grad_norm": 0.26421698927879333, | |
| "learning_rate": 5.3832287847914276e-05, | |
| "loss": 0.0241, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 2.2560429722470903, | |
| "grad_norm": 0.20106279850006104, | |
| "learning_rate": 5.35024852434055e-05, | |
| "loss": 0.0206, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.2649955237242616, | |
| "grad_norm": 0.2898752987384796, | |
| "learning_rate": 5.317252942884567e-05, | |
| "loss": 0.0204, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 2.2739480752014325, | |
| "grad_norm": 0.261635959148407, | |
| "learning_rate": 5.284243483757109e-05, | |
| "loss": 0.0265, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.2829006266786034, | |
| "grad_norm": 0.3588390052318573, | |
| "learning_rate": 5.2512215908988484e-05, | |
| "loss": 0.0269, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.2918531781557743, | |
| "grad_norm": 0.36631637811660767, | |
| "learning_rate": 5.218188708794357e-05, | |
| "loss": 0.0241, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.3008057296329456, | |
| "grad_norm": 0.24271060526371002, | |
| "learning_rate": 5.18514628240891e-05, | |
| "loss": 0.0208, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 2.3097582811101165, | |
| "grad_norm": 0.282932311296463, | |
| "learning_rate": 5.1520957571252795e-05, | |
| "loss": 0.0219, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.3187108325872874, | |
| "grad_norm": 0.39644038677215576, | |
| "learning_rate": 5.1190385786805106e-05, | |
| "loss": 0.0232, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 2.3276633840644583, | |
| "grad_norm": 0.3940184712409973, | |
| "learning_rate": 5.085976193102677e-05, | |
| "loss": 0.0261, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.336615935541629, | |
| "grad_norm": 0.3295639753341675, | |
| "learning_rate": 5.052910046647634e-05, | |
| "loss": 0.023, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 2.3455684870188005, | |
| "grad_norm": 0.20834366977214813, | |
| "learning_rate": 5.0198415857357464e-05, | |
| "loss": 0.0232, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.3545210384959714, | |
| "grad_norm": 0.20106403529644012, | |
| "learning_rate": 4.9867722568886223e-05, | |
| "loss": 0.0205, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 2.3634735899731423, | |
| "grad_norm": 0.314100056886673, | |
| "learning_rate": 4.9537035066658314e-05, | |
| "loss": 0.019, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.372426141450313, | |
| "grad_norm": 0.3410167694091797, | |
| "learning_rate": 4.920636781601638e-05, | |
| "loss": 0.0237, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.3813786929274845, | |
| "grad_norm": 0.29781830310821533, | |
| "learning_rate": 4.88757352814172e-05, | |
| "loss": 0.0267, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.3903312444046554, | |
| "grad_norm": 0.26218900084495544, | |
| "learning_rate": 4.8545151925798924e-05, | |
| "loss": 0.0214, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 2.3992837958818263, | |
| "grad_norm": 0.24904198944568634, | |
| "learning_rate": 4.821463220994848e-05, | |
| "loss": 0.0217, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.408236347358997, | |
| "grad_norm": 0.3924921751022339, | |
| "learning_rate": 4.788419059186895e-05, | |
| "loss": 0.0225, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.4171888988361685, | |
| "grad_norm": 0.3247718811035156, | |
| "learning_rate": 4.7553841526147205e-05, | |
| "loss": 0.0238, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.4261414503133394, | |
| "grad_norm": 0.2955084443092346, | |
| "learning_rate": 4.722359946332156e-05, | |
| "loss": 0.0227, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.4350940017905103, | |
| "grad_norm": 0.25664591789245605, | |
| "learning_rate": 4.6893478849249654e-05, | |
| "loss": 0.0217, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.444046553267681, | |
| "grad_norm": 0.2787065804004669, | |
| "learning_rate": 4.656349412447664e-05, | |
| "loss": 0.0242, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.452999104744852, | |
| "grad_norm": 0.2122863531112671, | |
| "learning_rate": 4.623365972360337e-05, | |
| "loss": 0.0207, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.4619516562220234, | |
| "grad_norm": 0.34877288341522217, | |
| "learning_rate": 4.590399007465503e-05, | |
| "loss": 0.0187, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.4709042076991943, | |
| "grad_norm": 0.31759724020957947, | |
| "learning_rate": 4.557449959845005e-05, | |
| "loss": 0.0215, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.479856759176365, | |
| "grad_norm": 0.30043768882751465, | |
| "learning_rate": 4.524520270796927e-05, | |
| "loss": 0.0185, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.4888093106535365, | |
| "grad_norm": 0.20219744741916656, | |
| "learning_rate": 4.491611380772545e-05, | |
| "loss": 0.019, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.4977618621307074, | |
| "grad_norm": 0.3866671621799469, | |
| "learning_rate": 4.458724729313318e-05, | |
| "loss": 0.0175, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.5067144136078783, | |
| "grad_norm": 0.23905040323734283, | |
| "learning_rate": 4.42586175498792e-05, | |
| "loss": 0.0215, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.515666965085049, | |
| "grad_norm": 0.3837669789791107, | |
| "learning_rate": 4.3930238953293094e-05, | |
| "loss": 0.0208, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.52461951656222, | |
| "grad_norm": 0.3023552894592285, | |
| "learning_rate": 4.360212586771847e-05, | |
| "loss": 0.0163, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.5335720680393914, | |
| "grad_norm": 0.25578853487968445, | |
| "learning_rate": 4.327429264588463e-05, | |
| "loss": 0.0231, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.5425246195165623, | |
| "grad_norm": 0.2065793126821518, | |
| "learning_rate": 4.2946753628278725e-05, | |
| "loss": 0.0203, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.551477170993733, | |
| "grad_norm": 0.30230146646499634, | |
| "learning_rate": 4.2619523142518474e-05, | |
| "loss": 0.0196, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.560429722470904, | |
| "grad_norm": 0.336173415184021, | |
| "learning_rate": 4.229261550272539e-05, | |
| "loss": 0.0191, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.569382273948075, | |
| "grad_norm": 0.3281089663505554, | |
| "learning_rate": 4.196604500889868e-05, | |
| "loss": 0.0218, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.5783348254252463, | |
| "grad_norm": 0.15925340354442596, | |
| "learning_rate": 4.163982594628969e-05, | |
| "loss": 0.0185, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.587287376902417, | |
| "grad_norm": 0.12922883033752441, | |
| "learning_rate": 4.131397258477702e-05, | |
| "loss": 0.0192, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.596239928379588, | |
| "grad_norm": 0.25064435601234436, | |
| "learning_rate": 4.0988499178242315e-05, | |
| "loss": 0.0203, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.6051924798567594, | |
| "grad_norm": 0.42107462882995605, | |
| "learning_rate": 4.066341996394678e-05, | |
| "loss": 0.0281, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.6141450313339303, | |
| "grad_norm": 0.3484732210636139, | |
| "learning_rate": 4.033874916190833e-05, | |
| "loss": 0.0203, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.623097582811101, | |
| "grad_norm": 0.3492477834224701, | |
| "learning_rate": 4.001450097427966e-05, | |
| "loss": 0.0196, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 2.632050134288272, | |
| "grad_norm": 0.2821838855743408, | |
| "learning_rate": 3.9690689584726894e-05, | |
| "loss": 0.018, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.641002685765443, | |
| "grad_norm": 0.3413795232772827, | |
| "learning_rate": 3.936732915780923e-05, | |
| "loss": 0.019, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.6499552372426143, | |
| "grad_norm": 0.2790643572807312, | |
| "learning_rate": 3.904443383835929e-05, | |
| "loss": 0.0243, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.658907788719785, | |
| "grad_norm": 0.20392818748950958, | |
| "learning_rate": 3.872201775086437e-05, | |
| "loss": 0.0225, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.667860340196956, | |
| "grad_norm": 0.2968049943447113, | |
| "learning_rate": 3.8400094998848616e-05, | |
| "loss": 0.0178, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.6768128916741274, | |
| "grad_norm": 0.26228010654449463, | |
| "learning_rate": 3.807867966425611e-05, | |
| "loss": 0.0182, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 2.6857654431512983, | |
| "grad_norm": 0.3043522238731384, | |
| "learning_rate": 3.775778580683481e-05, | |
| "loss": 0.0186, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.694717994628469, | |
| "grad_norm": 0.2954307794570923, | |
| "learning_rate": 3.743742746352156e-05, | |
| "loss": 0.0227, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 2.70367054610564, | |
| "grad_norm": 0.2497381865978241, | |
| "learning_rate": 3.711761864782817e-05, | |
| "loss": 0.0214, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.712623097582811, | |
| "grad_norm": 0.3127233684062958, | |
| "learning_rate": 3.679837334922825e-05, | |
| "loss": 0.019, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.7215756490599823, | |
| "grad_norm": 0.18976886570453644, | |
| "learning_rate": 3.647970553254538e-05, | |
| "loss": 0.0159, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.730528200537153, | |
| "grad_norm": 0.17379194498062134, | |
| "learning_rate": 3.61616291373422e-05, | |
| "loss": 0.0212, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.739480752014324, | |
| "grad_norm": 0.2456379234790802, | |
| "learning_rate": 3.584415807731065e-05, | |
| "loss": 0.0204, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.748433303491495, | |
| "grad_norm": 0.35977035760879517, | |
| "learning_rate": 3.552730623966337e-05, | |
| "loss": 0.0253, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.757385854968666, | |
| "grad_norm": 0.3271699547767639, | |
| "learning_rate": 3.521108748452617e-05, | |
| "loss": 0.0209, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.766338406445837, | |
| "grad_norm": 0.17938630282878876, | |
| "learning_rate": 3.489551564433186e-05, | |
| "loss": 0.0223, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.775290957923008, | |
| "grad_norm": 0.3276021480560303, | |
| "learning_rate": 3.4580604523215006e-05, | |
| "loss": 0.0217, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.784243509400179, | |
| "grad_norm": 0.25213152170181274, | |
| "learning_rate": 3.4266367896408216e-05, | |
| "loss": 0.0233, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 2.7931960608773503, | |
| "grad_norm": 0.19577741622924805, | |
| "learning_rate": 3.3952819509639534e-05, | |
| "loss": 0.0168, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.802148612354521, | |
| "grad_norm": 0.33538851141929626, | |
| "learning_rate": 3.3639973078531165e-05, | |
| "loss": 0.0215, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 2.811101163831692, | |
| "grad_norm": 0.3436635434627533, | |
| "learning_rate": 3.332784228799947e-05, | |
| "loss": 0.0208, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.820053715308863, | |
| "grad_norm": 0.25322389602661133, | |
| "learning_rate": 3.301644079165638e-05, | |
| "loss": 0.0191, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.829006266786034, | |
| "grad_norm": 0.3786560595035553, | |
| "learning_rate": 3.27057822112122e-05, | |
| "loss": 0.0188, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.837958818263205, | |
| "grad_norm": 0.20764805376529694, | |
| "learning_rate": 3.239588013587958e-05, | |
| "loss": 0.0218, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 2.846911369740376, | |
| "grad_norm": 0.17294557392597198, | |
| "learning_rate": 3.208674812177926e-05, | |
| "loss": 0.02, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.855863921217547, | |
| "grad_norm": 0.19973523914813995, | |
| "learning_rate": 3.177839969134698e-05, | |
| "loss": 0.0186, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.864816472694718, | |
| "grad_norm": 0.30084753036499023, | |
| "learning_rate": 3.1470848332742e-05, | |
| "loss": 0.0205, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.8737690241718887, | |
| "grad_norm": 0.371510773897171, | |
| "learning_rate": 3.116410749925708e-05, | |
| "loss": 0.0205, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.88272157564906, | |
| "grad_norm": 0.24790452420711517, | |
| "learning_rate": 3.085819060872995e-05, | |
| "loss": 0.017, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.891674127126231, | |
| "grad_norm": 0.18967312574386597, | |
| "learning_rate": 3.055311104295648e-05, | |
| "loss": 0.018, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 2.900626678603402, | |
| "grad_norm": 0.20764872431755066, | |
| "learning_rate": 3.024888214710517e-05, | |
| "loss": 0.0174, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.909579230080573, | |
| "grad_norm": 0.2953355610370636, | |
| "learning_rate": 2.994551722913349e-05, | |
| "loss": 0.0151, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.918531781557744, | |
| "grad_norm": 0.237528458237648, | |
| "learning_rate": 2.9643029559205727e-05, | |
| "loss": 0.023, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 2.927484333034915, | |
| "grad_norm": 0.225080668926239, | |
| "learning_rate": 2.934143236911248e-05, | |
| "loss": 0.0174, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 2.936436884512086, | |
| "grad_norm": 0.1836756318807602, | |
| "learning_rate": 2.90407388516919e-05, | |
| "loss": 0.0163, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 2.9453894359892567, | |
| "grad_norm": 0.1732194870710373, | |
| "learning_rate": 2.8740962160252495e-05, | |
| "loss": 0.0183, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 2.954341987466428, | |
| "grad_norm": 0.42215636372566223, | |
| "learning_rate": 2.844211540799797e-05, | |
| "loss": 0.0169, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.963294538943599, | |
| "grad_norm": 0.15775205194950104, | |
| "learning_rate": 2.8144211667453368e-05, | |
| "loss": 0.0162, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 2.97224709042077, | |
| "grad_norm": 0.2212500423192978, | |
| "learning_rate": 2.7847263969893344e-05, | |
| "loss": 0.0152, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 2.981199641897941, | |
| "grad_norm": 0.43953937292099, | |
| "learning_rate": 2.7551285304772206e-05, | |
| "loss": 0.0168, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 2.990152193375112, | |
| "grad_norm": 0.2736935317516327, | |
| "learning_rate": 2.7256288619155567e-05, | |
| "loss": 0.018, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 2.999104744852283, | |
| "grad_norm": 0.2877485156059265, | |
| "learning_rate": 2.6962286817154158e-05, | |
| "loss": 0.0203, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 3.008057296329454, | |
| "grad_norm": 0.21356728672981262, | |
| "learning_rate": 2.6669292759359166e-05, | |
| "loss": 0.0175, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 3.0170098478066247, | |
| "grad_norm": 0.29351386427879333, | |
| "learning_rate": 2.637731926227993e-05, | |
| "loss": 0.0255, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 3.025962399283796, | |
| "grad_norm": 0.30802270770072937, | |
| "learning_rate": 2.6086379097783033e-05, | |
| "loss": 0.0185, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 3.034914950760967, | |
| "grad_norm": 0.23744232952594757, | |
| "learning_rate": 2.579648499253377e-05, | |
| "loss": 0.0172, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 3.043867502238138, | |
| "grad_norm": 0.33891722559928894, | |
| "learning_rate": 2.5507649627439466e-05, | |
| "loss": 0.0201, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.0528200537153087, | |
| "grad_norm": 0.12712599337100983, | |
| "learning_rate": 2.5219885637094653e-05, | |
| "loss": 0.0181, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 3.06177260519248, | |
| "grad_norm": 0.25749579071998596, | |
| "learning_rate": 2.4933205609228533e-05, | |
| "loss": 0.0188, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 3.070725156669651, | |
| "grad_norm": 0.25427231192588806, | |
| "learning_rate": 2.464762208415419e-05, | |
| "loss": 0.0149, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 3.079677708146822, | |
| "grad_norm": 0.2582094669342041, | |
| "learning_rate": 2.4363147554220213e-05, | |
| "loss": 0.0184, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 3.0886302596239927, | |
| "grad_norm": 0.268612265586853, | |
| "learning_rate": 2.407979446326411e-05, | |
| "loss": 0.0147, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 3.0975828111011636, | |
| "grad_norm": 0.22941374778747559, | |
| "learning_rate": 2.379757520606799e-05, | |
| "loss": 0.0195, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 3.106535362578335, | |
| "grad_norm": 0.2100794017314911, | |
| "learning_rate": 2.3516502127816455e-05, | |
| "loss": 0.016, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 3.115487914055506, | |
| "grad_norm": 0.13210788369178772, | |
| "learning_rate": 2.323658752355647e-05, | |
| "loss": 0.017, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 3.1244404655326767, | |
| "grad_norm": 0.18918994069099426, | |
| "learning_rate": 2.2957843637659654e-05, | |
| "loss": 0.0165, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 3.1333930170098476, | |
| "grad_norm": 0.3535359501838684, | |
| "learning_rate": 2.2680282663286552e-05, | |
| "loss": 0.0209, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.142345568487019, | |
| "grad_norm": 0.2309960275888443, | |
| "learning_rate": 2.2403916741853364e-05, | |
| "loss": 0.0167, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 3.15129811996419, | |
| "grad_norm": 0.21955038607120514, | |
| "learning_rate": 2.2128757962500817e-05, | |
| "loss": 0.0164, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 3.1602506714413607, | |
| "grad_norm": 0.24366547167301178, | |
| "learning_rate": 2.1854818361565275e-05, | |
| "loss": 0.015, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 3.1692032229185316, | |
| "grad_norm": 0.25391441583633423, | |
| "learning_rate": 2.1582109922052364e-05, | |
| "loss": 0.0192, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 3.178155774395703, | |
| "grad_norm": 0.2967330515384674, | |
| "learning_rate": 2.1310644573112635e-05, | |
| "loss": 0.0168, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 3.187108325872874, | |
| "grad_norm": 0.2360481321811676, | |
| "learning_rate": 2.1040434189519924e-05, | |
| "loss": 0.0163, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 3.1960608773500447, | |
| "grad_norm": 0.2346932291984558, | |
| "learning_rate": 2.0771490591151733e-05, | |
| "loss": 0.0181, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 3.2050134288272156, | |
| "grad_norm": 0.24760273098945618, | |
| "learning_rate": 2.0503825542472317e-05, | |
| "loss": 0.0175, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 3.213965980304387, | |
| "grad_norm": 0.19240783154964447, | |
| "learning_rate": 2.023745075201805e-05, | |
| "loss": 0.0169, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 3.222918531781558, | |
| "grad_norm": 0.2748640477657318, | |
| "learning_rate": 1.9972377871885157e-05, | |
| "loss": 0.0179, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.2318710832587287, | |
| "grad_norm": 0.22245590388774872, | |
| "learning_rate": 1.970861849722017e-05, | |
| "loss": 0.0136, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 3.2408236347358996, | |
| "grad_norm": 0.24653933942317963, | |
| "learning_rate": 1.9446184165712587e-05, | |
| "loss": 0.0161, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 3.2497761862130705, | |
| "grad_norm": 0.21308885514736176, | |
| "learning_rate": 1.9185086357090214e-05, | |
| "loss": 0.0156, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 3.258728737690242, | |
| "grad_norm": 0.1842220425605774, | |
| "learning_rate": 1.8925336492617057e-05, | |
| "loss": 0.0198, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 3.2676812891674127, | |
| "grad_norm": 0.28430524468421936, | |
| "learning_rate": 1.8666945934593666e-05, | |
| "loss": 0.0174, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 3.2766338406445836, | |
| "grad_norm": 0.22021664679050446, | |
| "learning_rate": 1.8409925985860126e-05, | |
| "loss": 0.0143, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 3.2855863921217545, | |
| "grad_norm": 0.26781344413757324, | |
| "learning_rate": 1.8154287889301603e-05, | |
| "loss": 0.015, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 3.294538943598926, | |
| "grad_norm": 0.19385065138339996, | |
| "learning_rate": 1.7900042827356612e-05, | |
| "loss": 0.0169, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 3.3034914950760967, | |
| "grad_norm": 0.1983576864004135, | |
| "learning_rate": 1.76472019215278e-05, | |
| "loss": 0.0164, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 3.3124440465532676, | |
| "grad_norm": 0.2234923541545868, | |
| "learning_rate": 1.739577623189545e-05, | |
| "loss": 0.0158, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.3213965980304385, | |
| "grad_norm": 0.163887158036232, | |
| "learning_rate": 1.7145776756633768e-05, | |
| "loss": 0.0201, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 3.33034914950761, | |
| "grad_norm": 0.1897057443857193, | |
| "learning_rate": 1.6897214431529646e-05, | |
| "loss": 0.0175, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 3.3393017009847807, | |
| "grad_norm": 0.21445390582084656, | |
| "learning_rate": 1.6650100129504475e-05, | |
| "loss": 0.0138, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 3.3482542524619516, | |
| "grad_norm": 0.2707969546318054, | |
| "learning_rate": 1.6404444660138335e-05, | |
| "loss": 0.0162, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 3.3572068039391225, | |
| "grad_norm": 0.3275142312049866, | |
| "learning_rate": 1.616025876919725e-05, | |
| "loss": 0.0158, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 3.3661593554162934, | |
| "grad_norm": 0.23695893585681915, | |
| "learning_rate": 1.5917553138163172e-05, | |
| "loss": 0.0174, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 3.3751119068934647, | |
| "grad_norm": 0.17175161838531494, | |
| "learning_rate": 1.5676338383766632e-05, | |
| "loss": 0.0197, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 3.3840644583706356, | |
| "grad_norm": 0.17486472427845, | |
| "learning_rate": 1.5436625057522447e-05, | |
| "loss": 0.0145, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 3.3930170098478065, | |
| "grad_norm": 0.2262464016675949, | |
| "learning_rate": 1.519842364526804e-05, | |
| "loss": 0.0129, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 3.401969561324978, | |
| "grad_norm": 0.2933650016784668, | |
| "learning_rate": 1.4961744566704855e-05, | |
| "loss": 0.0163, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.4109221128021487, | |
| "grad_norm": 0.1765672117471695, | |
| "learning_rate": 1.4726598174942551e-05, | |
| "loss": 0.0132, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 3.4198746642793196, | |
| "grad_norm": 0.22736941277980804, | |
| "learning_rate": 1.4492994756046035e-05, | |
| "loss": 0.0171, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 3.4288272157564905, | |
| "grad_norm": 0.16248083114624023, | |
| "learning_rate": 1.4260944528585645e-05, | |
| "loss": 0.0167, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 3.4377797672336614, | |
| "grad_norm": 0.2661954164505005, | |
| "learning_rate": 1.4030457643190048e-05, | |
| "loss": 0.0141, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 3.4467323187108327, | |
| "grad_norm": 0.1929498016834259, | |
| "learning_rate": 1.3801544182102311e-05, | |
| "loss": 0.0154, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 3.4556848701880036, | |
| "grad_norm": 0.21110036969184875, | |
| "learning_rate": 1.3574214158738763e-05, | |
| "loss": 0.0168, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 3.4646374216651745, | |
| "grad_norm": 0.35853829979896545, | |
| "learning_rate": 1.3348477517251101e-05, | |
| "loss": 0.0158, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 3.4735899731423454, | |
| "grad_norm": 0.20026364922523499, | |
| "learning_rate": 1.312434413209131e-05, | |
| "loss": 0.0138, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 3.4825425246195167, | |
| "grad_norm": 0.2420613169670105, | |
| "learning_rate": 1.2901823807579727e-05, | |
| "loss": 0.0146, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 3.4914950760966876, | |
| "grad_norm": 0.2799060344696045, | |
| "learning_rate": 1.2680926277476245e-05, | |
| "loss": 0.0159, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.5004476275738585, | |
| "grad_norm": 0.2773265242576599, | |
| "learning_rate": 1.2461661204554397e-05, | |
| "loss": 0.0151, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 3.5094001790510294, | |
| "grad_norm": 0.18590183556079865, | |
| "learning_rate": 1.2244038180178835e-05, | |
| "loss": 0.0147, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 3.5183527305282007, | |
| "grad_norm": 0.19557642936706543, | |
| "learning_rate": 1.2028066723885612e-05, | |
| "loss": 0.0154, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 3.5273052820053716, | |
| "grad_norm": 0.22107233107089996, | |
| "learning_rate": 1.1813756282965888e-05, | |
| "loss": 0.0154, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 3.5362578334825425, | |
| "grad_norm": 0.18053828179836273, | |
| "learning_rate": 1.1601116232052638e-05, | |
| "loss": 0.0165, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 3.5452103849597134, | |
| "grad_norm": 0.21140411496162415, | |
| "learning_rate": 1.1390155872710517e-05, | |
| "loss": 0.0134, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.5541629364368843, | |
| "grad_norm": 0.16849522292613983, | |
| "learning_rate": 1.1180884433029087e-05, | |
| "loss": 0.0165, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 3.5631154879140556, | |
| "grad_norm": 0.14933831989765167, | |
| "learning_rate": 1.097331106721904e-05, | |
| "loss": 0.0124, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 3.5720680393912265, | |
| "grad_norm": 0.217430979013443, | |
| "learning_rate": 1.0767444855211862e-05, | |
| "loss": 0.0135, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 3.5810205908683974, | |
| "grad_norm": 0.21716763079166412, | |
| "learning_rate": 1.0563294802262558e-05, | |
| "loss": 0.0167, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.5899731423455687, | |
| "grad_norm": 0.20996397733688354, | |
| "learning_rate": 1.0360869838555809e-05, | |
| "loss": 0.0131, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 3.5989256938227396, | |
| "grad_norm": 0.22892490029335022, | |
| "learning_rate": 1.0160178818815313e-05, | |
| "loss": 0.0134, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 3.6078782452999105, | |
| "grad_norm": 0.22495168447494507, | |
| "learning_rate": 9.961230521916387e-06, | |
| "loss": 0.0146, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 3.6168307967770814, | |
| "grad_norm": 0.1184273287653923, | |
| "learning_rate": 9.764033650502074e-06, | |
| "loss": 0.0143, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 3.6257833482542523, | |
| "grad_norm": 0.18436811864376068, | |
| "learning_rate": 9.568596830602344e-06, | |
| "loss": 0.0114, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 3.6347358997314236, | |
| "grad_norm": 0.2092517912387848, | |
| "learning_rate": 9.37492861125681e-06, | |
| "loss": 0.0164, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 3.6436884512085945, | |
| "grad_norm": 0.1160958930850029, | |
| "learning_rate": 9.183037464140804e-06, | |
| "loss": 0.0162, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 3.6526410026857654, | |
| "grad_norm": 0.24437487125396729, | |
| "learning_rate": 8.992931783194735e-06, | |
| "loss": 0.0131, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 3.6615935541629363, | |
| "grad_norm": 0.185224711894989, | |
| "learning_rate": 8.80461988425696e-06, | |
| "loss": 0.0155, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 3.670546105640107, | |
| "grad_norm": 0.2819221019744873, | |
| "learning_rate": 8.618110004699974e-06, | |
| "loss": 0.0148, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 3.6794986571172785, | |
| "grad_norm": 0.24776029586791992, | |
| "learning_rate": 8.4334103030701e-06, | |
| "loss": 0.0162, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 3.6884512085944494, | |
| "grad_norm": 0.19672740995883942, | |
| "learning_rate": 8.25052885873066e-06, | |
| "loss": 0.0136, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 3.6974037600716203, | |
| "grad_norm": 0.26846009492874146, | |
| "learning_rate": 8.06947367150846e-06, | |
| "loss": 0.0139, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 3.7063563115487916, | |
| "grad_norm": 0.15493078529834747, | |
| "learning_rate": 7.890252661343938e-06, | |
| "loss": 0.0121, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 3.7153088630259625, | |
| "grad_norm": 0.19229039549827576, | |
| "learning_rate": 7.712873667944681e-06, | |
| "loss": 0.0137, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 3.7242614145031334, | |
| "grad_norm": 0.2311052680015564, | |
| "learning_rate": 7.537344450442469e-06, | |
| "loss": 0.0104, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 3.7332139659803043, | |
| "grad_norm": 0.19902154803276062, | |
| "learning_rate": 7.36367268705393e-06, | |
| "loss": 0.0173, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 3.742166517457475, | |
| "grad_norm": 0.2320551574230194, | |
| "learning_rate": 7.1918659747446e-06, | |
| "loss": 0.0144, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 3.7511190689346465, | |
| "grad_norm": 0.1548873782157898, | |
| "learning_rate": 7.021931828896666e-06, | |
| "loss": 0.0161, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 3.7600716204118174, | |
| "grad_norm": 0.30775901675224304, | |
| "learning_rate": 6.8538776829801584e-06, | |
| "loss": 0.0134, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.7690241718889883, | |
| "grad_norm": 0.36373430490493774, | |
| "learning_rate": 6.687710888227849e-06, | |
| "loss": 0.0115, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 3.777976723366159, | |
| "grad_norm": 0.20033209025859833, | |
| "learning_rate": 6.5234387133136565e-06, | |
| "loss": 0.0109, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 3.78692927484333, | |
| "grad_norm": 0.1912829428911209, | |
| "learning_rate": 6.361068344034665e-06, | |
| "loss": 0.0101, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 3.7958818263205014, | |
| "grad_norm": 0.2168465107679367, | |
| "learning_rate": 6.200606882996846e-06, | |
| "loss": 0.0117, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 3.8048343777976723, | |
| "grad_norm": 0.17642278969287872, | |
| "learning_rate": 6.042061349304312e-06, | |
| "loss": 0.0144, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 3.813786929274843, | |
| "grad_norm": 0.1212242841720581, | |
| "learning_rate": 5.885438678252342e-06, | |
| "loss": 0.0172, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 3.8227394807520145, | |
| "grad_norm": 0.1348852664232254, | |
| "learning_rate": 5.730745721023939e-06, | |
| "loss": 0.0129, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 3.8316920322291854, | |
| "grad_norm": 0.15375061333179474, | |
| "learning_rate": 5.577989244390192e-06, | |
| "loss": 0.0115, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 3.8406445837063563, | |
| "grad_norm": 0.16176094114780426, | |
| "learning_rate": 5.4271759304142635e-06, | |
| "loss": 0.0112, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 3.849597135183527, | |
| "grad_norm": 0.23048561811447144, | |
| "learning_rate": 5.278312376159051e-06, | |
| "loss": 0.0131, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 3.858549686660698, | |
| "grad_norm": 0.1969141662120819, | |
| "learning_rate": 5.1314050933986944e-06, | |
| "loss": 0.0109, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 3.8675022381378694, | |
| "grad_norm": 0.15142923593521118, | |
| "learning_rate": 4.986460508333634e-06, | |
| "loss": 0.0117, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 3.8764547896150403, | |
| "grad_norm": 0.13823889195919037, | |
| "learning_rate": 4.843484961309597e-06, | |
| "loss": 0.0119, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 3.885407341092211, | |
| "grad_norm": 0.2543256878852844, | |
| "learning_rate": 4.702484706540161e-06, | |
| "loss": 0.0126, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 3.8943598925693825, | |
| "grad_norm": 0.24175454676151276, | |
| "learning_rate": 4.563465911833259e-06, | |
| "loss": 0.0105, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 3.9033124440465534, | |
| "grad_norm": 0.22203406691551208, | |
| "learning_rate": 4.426434658321344e-06, | |
| "loss": 0.0112, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 3.9122649955237243, | |
| "grad_norm": 0.2877592146396637, | |
| "learning_rate": 4.2913969401953466e-06, | |
| "loss": 0.0124, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 3.921217547000895, | |
| "grad_norm": 0.15440590679645538, | |
| "learning_rate": 4.15835866444253e-06, | |
| "loss": 0.0116, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 3.930170098478066, | |
| "grad_norm": 0.2942441999912262, | |
| "learning_rate": 4.027325650588043e-06, | |
| "loss": 0.017, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 3.9391226499552374, | |
| "grad_norm": 0.3283434212207794, | |
| "learning_rate": 3.898303630440419e-06, | |
| "loss": 0.0166, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 3.9480752014324083, | |
| "grad_norm": 0.12953047454357147, | |
| "learning_rate": 3.7712982478407877e-06, | |
| "loss": 0.0122, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 3.957027752909579, | |
| "grad_norm": 0.2435809224843979, | |
| "learning_rate": 3.6463150584160053e-06, | |
| "loss": 0.0137, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 3.96598030438675, | |
| "grad_norm": 0.3056238293647766, | |
| "learning_rate": 3.5233595293356957e-06, | |
| "loss": 0.0131, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 3.974932855863921, | |
| "grad_norm": 0.2567388117313385, | |
| "learning_rate": 3.4024370390730033e-06, | |
| "loss": 0.0185, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 3.9838854073410923, | |
| "grad_norm": 0.22484621405601501, | |
| "learning_rate": 3.2835528771693992e-06, | |
| "loss": 0.0119, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 3.992837958818263, | |
| "grad_norm": 0.14950992166996002, | |
| "learning_rate": 3.1667122440032505e-06, | |
| "loss": 0.0118, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 4.001790510295434, | |
| "grad_norm": 0.18850022554397583, | |
| "learning_rate": 3.051920250562351e-06, | |
| "loss": 0.0125, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 4.010743061772605, | |
| "grad_norm": 0.21999414265155792, | |
| "learning_rate": 2.939181918220385e-06, | |
| "loss": 0.0126, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 4.019695613249776, | |
| "grad_norm": 0.2078053206205368, | |
| "learning_rate": 2.8285021785172226e-06, | |
| "loss": 0.0134, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 4.028648164726947, | |
| "grad_norm": 0.11813908815383911, | |
| "learning_rate": 2.7198858729432288e-06, | |
| "loss": 0.0135, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.0376007162041185, | |
| "grad_norm": 0.1995633989572525, | |
| "learning_rate": 2.6133377527274905e-06, | |
| "loss": 0.0155, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 4.046553267681289, | |
| "grad_norm": 0.22360952198505402, | |
| "learning_rate": 2.5088624786299366e-06, | |
| "loss": 0.0136, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 4.05550581915846, | |
| "grad_norm": 0.1910884976387024, | |
| "learning_rate": 2.406464620737531e-06, | |
| "loss": 0.0141, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 4.064458370635631, | |
| "grad_norm": 0.16512851417064667, | |
| "learning_rate": 2.3061486582642734e-06, | |
| "loss": 0.0113, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 4.073410922112802, | |
| "grad_norm": 0.16644251346588135, | |
| "learning_rate": 2.2079189793553667e-06, | |
| "loss": 0.0103, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 4.082363473589973, | |
| "grad_norm": 0.19846302270889282, | |
| "learning_rate": 2.111779880895165e-06, | |
| "loss": 0.0102, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 4.091316025067144, | |
| "grad_norm": 0.11550256609916687, | |
| "learning_rate": 2.01773556831929e-06, | |
| "loss": 0.0135, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 4.100268576544315, | |
| "grad_norm": 0.1841946244239807, | |
| "learning_rate": 1.9257901554306513e-06, | |
| "loss": 0.0104, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 4.1092211280214865, | |
| "grad_norm": 0.16650405526161194, | |
| "learning_rate": 1.835947664219445e-06, | |
| "loss": 0.0127, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 4.118173679498657, | |
| "grad_norm": 0.197437584400177, | |
| "learning_rate": 1.748212024687307e-06, | |
| "loss": 0.0122, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.127126230975828, | |
| "grad_norm": 0.08538883924484253, | |
| "learning_rate": 1.6625870746753147e-06, | |
| "loss": 0.0115, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 4.136078782452999, | |
| "grad_norm": 0.1866847425699234, | |
| "learning_rate": 1.5790765596961853e-06, | |
| "loss": 0.0125, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 4.14503133393017, | |
| "grad_norm": 0.19738441705703735, | |
| "learning_rate": 1.4976841327703717e-06, | |
| "loss": 0.0137, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 4.153983885407341, | |
| "grad_norm": 0.19295786321163177, | |
| "learning_rate": 1.4184133542663014e-06, | |
| "loss": 0.0127, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 4.162936436884512, | |
| "grad_norm": 0.25274568796157837, | |
| "learning_rate": 1.341267691744641e-06, | |
| "loss": 0.0115, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 4.171888988361683, | |
| "grad_norm": 0.1629364937543869, | |
| "learning_rate": 1.2662505198065666e-06, | |
| "loss": 0.0172, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 4.180841539838854, | |
| "grad_norm": 0.22567766904830933, | |
| "learning_rate": 1.193365119946216e-06, | |
| "loss": 0.0126, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 4.189794091316025, | |
| "grad_norm": 0.14044791460037231, | |
| "learning_rate": 1.1226146804070859e-06, | |
| "loss": 0.0084, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 4.198746642793196, | |
| "grad_norm": 0.10704066604375839, | |
| "learning_rate": 1.0540022960426111e-06, | |
| "loss": 0.015, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 4.207699194270367, | |
| "grad_norm": 0.17678844928741455, | |
| "learning_rate": 9.875309681807443e-07, | |
| "loss": 0.0165, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.216651745747538, | |
| "grad_norm": 0.2521614730358124, | |
| "learning_rate": 9.232036044927061e-07, | |
| "loss": 0.0116, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 4.225604297224709, | |
| "grad_norm": 0.1365845501422882, | |
| "learning_rate": 8.610230188657919e-07, | |
| "loss": 0.0136, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 4.23455684870188, | |
| "grad_norm": 0.17501398921012878, | |
| "learning_rate": 8.009919312802372e-07, | |
| "loss": 0.0127, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 4.243509400179051, | |
| "grad_norm": 0.1747271865606308, | |
| "learning_rate": 7.431129676902904e-07, | |
| "loss": 0.0109, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 4.252461951656222, | |
| "grad_norm": 0.31210875511169434, | |
| "learning_rate": 6.873886599093215e-07, | |
| "loss": 0.0143, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 4.261414503133393, | |
| "grad_norm": 0.1589571237564087, | |
| "learning_rate": 6.338214454990776e-07, | |
| "loss": 0.0115, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 4.270367054610564, | |
| "grad_norm": 0.1521824449300766, | |
| "learning_rate": 5.82413667663051e-07, | |
| "loss": 0.0122, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 4.279319606087735, | |
| "grad_norm": 0.20766985416412354, | |
| "learning_rate": 5.331675751439725e-07, | |
| "loss": 0.0145, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 4.288272157564906, | |
| "grad_norm": 0.2259196788072586, | |
| "learning_rate": 4.86085322125479e-07, | |
| "loss": 0.0112, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 4.2972247090420765, | |
| "grad_norm": 0.17303107678890228, | |
| "learning_rate": 4.411689681378284e-07, | |
| "loss": 0.0147, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.306177260519248, | |
| "grad_norm": 0.18465794622898102, | |
| "learning_rate": 3.9842047796786466e-07, | |
| "loss": 0.0114, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 4.315129811996419, | |
| "grad_norm": 0.1356513351202011, | |
| "learning_rate": 3.578417215730323e-07, | |
| "loss": 0.0089, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 4.32408236347359, | |
| "grad_norm": 0.2126903235912323, | |
| "learning_rate": 3.1943447399958027e-07, | |
| "loss": 0.0143, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 4.333034914950761, | |
| "grad_norm": 0.23088884353637695, | |
| "learning_rate": 2.8320041530495724e-07, | |
| "loss": 0.0198, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 4.341987466427932, | |
| "grad_norm": 0.151603102684021, | |
| "learning_rate": 2.491411304842539e-07, | |
| "loss": 0.0143, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 4.350940017905103, | |
| "grad_norm": 0.16578730940818787, | |
| "learning_rate": 2.1725810940094183e-07, | |
| "loss": 0.0141, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 4.359892569382274, | |
| "grad_norm": 0.11706045269966125, | |
| "learning_rate": 1.8755274672164202e-07, | |
| "loss": 0.0123, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 4.3688451208594445, | |
| "grad_norm": 0.21942616999149323, | |
| "learning_rate": 1.600263418551573e-07, | |
| "loss": 0.0163, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 4.377797672336616, | |
| "grad_norm": 0.163126140832901, | |
| "learning_rate": 1.346800988955954e-07, | |
| "loss": 0.0121, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 4.386750223813787, | |
| "grad_norm": 0.18442213535308838, | |
| "learning_rate": 1.1151512656975005e-07, | |
| "loss": 0.0119, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 4.395702775290958, | |
| "grad_norm": 0.1644607037305832, | |
| "learning_rate": 9.053243818853973e-08, | |
| "loss": 0.0125, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 4.404655326768129, | |
| "grad_norm": 0.1691417694091797, | |
| "learning_rate": 7.173295160273763e-08, | |
| "loss": 0.0089, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 4.4136078782453, | |
| "grad_norm": 0.19320355355739594, | |
| "learning_rate": 5.511748916279258e-08, | |
| "loss": 0.013, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 4.422560429722471, | |
| "grad_norm": 0.30131739377975464, | |
| "learning_rate": 4.068677768285234e-08, | |
| "loss": 0.012, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 4.431512981199642, | |
| "grad_norm": 0.22268839180469513, | |
| "learning_rate": 2.844144840898344e-08, | |
| "loss": 0.0148, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 4.4404655326768125, | |
| "grad_norm": 0.15642546117305756, | |
| "learning_rate": 1.8382036991559936e-08, | |
| "loss": 0.0089, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 4.449418084153984, | |
| "grad_norm": 0.14433293044567108, | |
| "learning_rate": 1.0508983461832156e-08, | |
| "loss": 0.0147, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 4.458370635631155, | |
| "grad_norm": 0.12535181641578674, | |
| "learning_rate": 4.822632212653222e-09, | |
| "loss": 0.0152, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 4.467323187108326, | |
| "grad_norm": 0.23488545417785645, | |
| "learning_rate": 1.3232319834632912e-09, | |
| "loss": 0.0127, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 4.476275738585497, | |
| "grad_norm": 0.22346431016921997, | |
| "learning_rate": 1.0935849353854721e-11, | |
| "loss": 0.0127, | |
| "step": 5000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |