| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 10580, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1890359168241966, | |
| "grad_norm": 0.37213513255119324, | |
| "learning_rate": 0.0003, | |
| "loss": 1.9201, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 0.5696528553962708, | |
| "learning_rate": 0.0003, | |
| "loss": 1.8659, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.44589683413505554, | |
| "learning_rate": 0.0003, | |
| "loss": 1.8382, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.4036942720413208, | |
| "learning_rate": 0.0003, | |
| "loss": 1.8431, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.45892009139060974, | |
| "learning_rate": 0.0003, | |
| "loss": 1.8369, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.575076923076923, | |
| "eval_loss": 1.6031557321548462, | |
| "eval_runtime": 5.71, | |
| "eval_samples_per_second": 87.566, | |
| "eval_steps_per_second": 11.033, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.4805261194705963, | |
| "learning_rate": 0.0003, | |
| "loss": 1.7006, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 0.4189806878566742, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6276, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.4777927100658417, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6294, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.4425320327281952, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6196, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 0.5979380011558533, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6451, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5745641025641026, | |
| "eval_loss": 1.635698676109314, | |
| "eval_runtime": 5.9951, | |
| "eval_samples_per_second": 83.401, | |
| "eval_steps_per_second": 10.509, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.5178828239440918, | |
| "learning_rate": 0.0003, | |
| "loss": 1.514, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 0.7963550090789795, | |
| "learning_rate": 0.0003, | |
| "loss": 1.3597, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 0.7210434675216675, | |
| "learning_rate": 0.0003, | |
| "loss": 1.3774, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 0.6484540700912476, | |
| "learning_rate": 0.0003, | |
| "loss": 1.374, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 0.6798349618911743, | |
| "learning_rate": 0.0003, | |
| "loss": 1.3703, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5716410256410256, | |
| "eval_loss": 1.7677161693572998, | |
| "eval_runtime": 5.9596, | |
| "eval_samples_per_second": 83.899, | |
| "eval_steps_per_second": 10.571, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0245746691871456, | |
| "grad_norm": 0.6527106761932373, | |
| "learning_rate": 0.0003, | |
| "loss": 1.3419, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.213610586011342, | |
| "grad_norm": 0.6613947749137878, | |
| "learning_rate": 0.0003, | |
| "loss": 1.0932, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.402646502835539, | |
| "grad_norm": 0.7362242341041565, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1188, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.5916824196597354, | |
| "grad_norm": 0.6629425287246704, | |
| "learning_rate": 0.0003, | |
| "loss": 1.142, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.780718336483932, | |
| "grad_norm": 0.789070725440979, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1661, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9697542533081287, | |
| "grad_norm": 0.7567113637924194, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1817, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5718461538461539, | |
| "eval_loss": 1.8587489128112793, | |
| "eval_runtime": 5.7528, | |
| "eval_samples_per_second": 86.914, | |
| "eval_steps_per_second": 10.951, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.158790170132325, | |
| "grad_norm": 0.7041072249412537, | |
| "learning_rate": 0.0003, | |
| "loss": 0.9335, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 0.7739288210868835, | |
| "learning_rate": 0.0003, | |
| "loss": 0.9257, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.536862003780718, | |
| "grad_norm": 0.9699936509132385, | |
| "learning_rate": 0.0003, | |
| "loss": 0.9234, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.725897920604915, | |
| "grad_norm": 0.8464515209197998, | |
| "learning_rate": 0.0003, | |
| "loss": 0.9547, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.914933837429111, | |
| "grad_norm": 0.7238239049911499, | |
| "learning_rate": 0.0003, | |
| "loss": 0.9674, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5712820512820512, | |
| "eval_loss": 1.931915283203125, | |
| "eval_runtime": 5.7956, | |
| "eval_samples_per_second": 86.272, | |
| "eval_steps_per_second": 10.87, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.103969754253308, | |
| "grad_norm": 0.8031799793243408, | |
| "learning_rate": 0.0003, | |
| "loss": 0.8353, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.293005671077505, | |
| "grad_norm": 0.862354576587677, | |
| "learning_rate": 0.0003, | |
| "loss": 0.731, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.482041587901701, | |
| "grad_norm": 0.9067243337631226, | |
| "learning_rate": 0.0003, | |
| "loss": 0.7515, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.671077504725898, | |
| "grad_norm": 0.9791676998138428, | |
| "learning_rate": 0.0003, | |
| "loss": 0.7769, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.8601134215500945, | |
| "grad_norm": 0.9806828498840332, | |
| "learning_rate": 0.0003, | |
| "loss": 0.7936, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5704102564102564, | |
| "eval_loss": 1.993375301361084, | |
| "eval_runtime": 5.8166, | |
| "eval_samples_per_second": 85.96, | |
| "eval_steps_per_second": 10.831, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.049149338374291, | |
| "grad_norm": 0.8496273159980774, | |
| "learning_rate": 0.0003, | |
| "loss": 0.749, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.238185255198488, | |
| "grad_norm": 0.8800780177116394, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5822, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.427221172022684, | |
| "grad_norm": 0.8136658668518066, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6114, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.616257088846881, | |
| "grad_norm": 1.0112674236297607, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6329, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.805293005671078, | |
| "grad_norm": 0.8546850681304932, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6499, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.994328922495274, | |
| "grad_norm": 0.947127640247345, | |
| "learning_rate": 0.0003, | |
| "loss": 0.67, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5683589743589743, | |
| "eval_loss": 2.046682357788086, | |
| "eval_runtime": 5.9678, | |
| "eval_samples_per_second": 83.784, | |
| "eval_steps_per_second": 10.557, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.183364839319471, | |
| "grad_norm": 1.0099776983261108, | |
| "learning_rate": 0.0003, | |
| "loss": 0.477, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.3724007561436675, | |
| "grad_norm": 0.8788864016532898, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4951, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.561436672967864, | |
| "grad_norm": 0.9243162274360657, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5141, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.750472589792061, | |
| "grad_norm": 1.0089187622070312, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5317, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.939508506616257, | |
| "grad_norm": 0.9674586057662964, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5604, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5692820512820512, | |
| "eval_loss": 2.121832847595215, | |
| "eval_runtime": 5.6926, | |
| "eval_samples_per_second": 87.833, | |
| "eval_steps_per_second": 11.067, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.128544423440454, | |
| "grad_norm": 0.8500985503196716, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4532, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.31758034026465, | |
| "grad_norm": 0.8404316902160645, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4173, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.506616257088847, | |
| "grad_norm": 0.9724924564361572, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4386, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.9974843859672546, | |
| "learning_rate": 0.0003, | |
| "loss": 0.453, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.88468809073724, | |
| "grad_norm": 0.9835983514785767, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4747, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5682051282051283, | |
| "eval_loss": 2.1342432498931885, | |
| "eval_runtime": 5.6644, | |
| "eval_samples_per_second": 88.271, | |
| "eval_steps_per_second": 11.122, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.073724007561436, | |
| "grad_norm": 0.7832907438278198, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4297, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.262759924385634, | |
| "grad_norm": 0.9495198130607605, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3607, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.45179584120983, | |
| "grad_norm": 0.8801999688148499, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3836, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.640831758034027, | |
| "grad_norm": 0.9312089681625366, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4013, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 9.829867674858223, | |
| "grad_norm": 1.0668907165527344, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4191, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5673846153846154, | |
| "eval_loss": 2.1679139137268066, | |
| "eval_runtime": 5.6062, | |
| "eval_samples_per_second": 89.186, | |
| "eval_steps_per_second": 11.237, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.01890359168242, | |
| "grad_norm": 0.8078038096427917, | |
| "learning_rate": 0.0003, | |
| "loss": 0.42, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.207939508506616, | |
| "grad_norm": 0.9549182653427124, | |
| "learning_rate": 0.0003, | |
| "loss": 0.329, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.396975425330814, | |
| "grad_norm": 1.0073105096817017, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3521, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.58601134215501, | |
| "grad_norm": 1.0606422424316406, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3671, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.775047258979207, | |
| "grad_norm": 0.8719451427459717, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3818, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 10.964083175803403, | |
| "grad_norm": 1.0333566665649414, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3971, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5657948717948718, | |
| "eval_loss": 2.2080512046813965, | |
| "eval_runtime": 5.6645, | |
| "eval_samples_per_second": 88.27, | |
| "eval_steps_per_second": 11.122, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.1531190926276, | |
| "grad_norm": 0.84638911485672, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3323, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.342155009451796, | |
| "grad_norm": 0.8573931455612183, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3304, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.531190926275993, | |
| "grad_norm": 1.0611106157302856, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3432, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.720226843100189, | |
| "grad_norm": 0.8618783950805664, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3601, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.909262759924385, | |
| "grad_norm": 1.0319652557373047, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3753, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5663589743589743, | |
| "eval_loss": 2.184002637863159, | |
| "eval_runtime": 5.7032, | |
| "eval_samples_per_second": 87.67, | |
| "eval_steps_per_second": 11.046, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.098298676748582, | |
| "grad_norm": 0.9860585331916809, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3454, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.287334593572778, | |
| "grad_norm": 0.8168879747390747, | |
| "learning_rate": 0.0003, | |
| "loss": 0.317, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.476370510396976, | |
| "grad_norm": 0.8994132876396179, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3286, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.665406427221171, | |
| "grad_norm": 0.9378811120986938, | |
| "learning_rate": 0.0003, | |
| "loss": 0.34, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 12.854442344045369, | |
| "grad_norm": 0.8742515444755554, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3571, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5633846153846154, | |
| "eval_loss": 2.232431411743164, | |
| "eval_runtime": 5.6334, | |
| "eval_samples_per_second": 88.756, | |
| "eval_steps_per_second": 11.183, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 0.8000278472900391, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3518, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.232514177693762, | |
| "grad_norm": 0.7633320689201355, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3046, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.421550094517958, | |
| "grad_norm": 0.9149838089942932, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3217, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.610586011342155, | |
| "grad_norm": 0.9108119010925293, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3315, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.799621928166351, | |
| "grad_norm": 0.8181828260421753, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3454, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 13.988657844990549, | |
| "grad_norm": 0.9064919352531433, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3526, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5631794871794872, | |
| "eval_loss": 2.2189676761627197, | |
| "eval_runtime": 5.8178, | |
| "eval_samples_per_second": 85.943, | |
| "eval_steps_per_second": 10.829, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.177693761814744, | |
| "grad_norm": 0.8295121788978577, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3011, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.366729678638942, | |
| "grad_norm": 0.78739994764328, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3091, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.555765595463138, | |
| "grad_norm": 0.9230947494506836, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3178, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 14.744801512287335, | |
| "grad_norm": 0.8687440752983093, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3298, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.93383742911153, | |
| "grad_norm": 0.7883646488189697, | |
| "learning_rate": 0.0003, | |
| "loss": 0.35, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5638974358974359, | |
| "eval_loss": 2.208575487136841, | |
| "eval_runtime": 5.8338, | |
| "eval_samples_per_second": 85.707, | |
| "eval_steps_per_second": 10.799, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.122873345935728, | |
| "grad_norm": 0.8580202460289001, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3089, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.311909262759924, | |
| "grad_norm": 1.0185339450836182, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2966, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.500945179584122, | |
| "grad_norm": 1.0962390899658203, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3135, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.689981096408317, | |
| "grad_norm": 0.9543150663375854, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3258, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 15.879017013232515, | |
| "grad_norm": 0.9073179364204407, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3323, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5653846153846154, | |
| "eval_loss": 2.2654531002044678, | |
| "eval_runtime": 5.7427, | |
| "eval_samples_per_second": 87.067, | |
| "eval_steps_per_second": 10.97, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.068052930056712, | |
| "grad_norm": 0.9761775135993958, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3298, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.257088846880908, | |
| "grad_norm": 0.8946503400802612, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2945, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.446124763705104, | |
| "grad_norm": 0.8116426467895508, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3051, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 16.6351606805293, | |
| "grad_norm": 0.8611814379692078, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3145, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.8241965973535, | |
| "grad_norm": 0.9202536940574646, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3281, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5667179487179487, | |
| "eval_loss": 2.244356870651245, | |
| "eval_runtime": 5.9955, | |
| "eval_samples_per_second": 83.396, | |
| "eval_steps_per_second": 10.508, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.013232514177695, | |
| "grad_norm": 0.7778891324996948, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3344, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.20226843100189, | |
| "grad_norm": 0.7352868914604187, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2843, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.7922298908233643, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2941, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.58034026465028, | |
| "grad_norm": 1.0191751718521118, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3164, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 17.76937618147448, | |
| "grad_norm": 1.0436335802078247, | |
| "learning_rate": 0.0003, | |
| "loss": 0.321, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.958412098298677, | |
| "grad_norm": 0.8792089223861694, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3328, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5626153846153846, | |
| "eval_loss": 2.2596943378448486, | |
| "eval_runtime": 6.0856, | |
| "eval_samples_per_second": 82.161, | |
| "eval_steps_per_second": 10.352, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.147448015122873, | |
| "grad_norm": 0.6146367788314819, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2967, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.33648393194707, | |
| "grad_norm": 0.8163793683052063, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2958, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.525519848771268, | |
| "grad_norm": 0.8955701589584351, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3065, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.714555765595463, | |
| "grad_norm": 0.7934162020683289, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3121, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 18.90359168241966, | |
| "grad_norm": 1.07945716381073, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3305, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5633333333333334, | |
| "eval_loss": 2.2682220935821533, | |
| "eval_runtime": 5.8794, | |
| "eval_samples_per_second": 85.042, | |
| "eval_steps_per_second": 10.715, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.092627599243855, | |
| "grad_norm": 0.694956362247467, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3069, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.281663516068054, | |
| "grad_norm": 0.8144003748893738, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2879, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.47069943289225, | |
| "grad_norm": 0.7943779826164246, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3002, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 19.659735349716446, | |
| "grad_norm": 0.7781754732131958, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3107, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.84877126654064, | |
| "grad_norm": 0.7696000933647156, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3228, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5640512820512821, | |
| "eval_loss": 2.253227472305298, | |
| "eval_runtime": 5.7574, | |
| "eval_samples_per_second": 86.845, | |
| "eval_steps_per_second": 10.943, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10580, | |
| "total_flos": 6.517631969856061e+17, | |
| "train_loss": 0.6476908660340625, | |
| "train_runtime": 24281.4349, | |
| "train_samples_per_second": 13.941, | |
| "train_steps_per_second": 0.436 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10580, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 6.517631969856061e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |