| { | |
| "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": 0.45108333230018616, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9159, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 0.6339304447174072, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8756, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.5402090549468994, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8511, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.43492260575294495, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8581, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.5317010879516602, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8583, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5726153846153846, | |
| "eval_loss": 1.6375669240951538, | |
| "eval_runtime": 5.9206, | |
| "eval_samples_per_second": 84.451, | |
| "eval_steps_per_second": 10.641, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.5848847031593323, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6655, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 0.642599880695343, | |
| "learning_rate": 0.0005, | |
| "loss": 1.581, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.5563580393791199, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5945, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.576246976852417, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5962, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 0.6526830196380615, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6329, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5712820512820512, | |
| "eval_loss": 1.6881264448165894, | |
| "eval_runtime": 6.002, | |
| "eval_samples_per_second": 83.305, | |
| "eval_steps_per_second": 10.496, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.614975094795227, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4729, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 0.7056101560592651, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2845, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 0.8371734619140625, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3214, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 0.8234328627586365, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3326, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 0.7623058557510376, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3464, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5663076923076923, | |
| "eval_loss": 1.8256189823150635, | |
| "eval_runtime": 5.9355, | |
| "eval_samples_per_second": 84.239, | |
| "eval_steps_per_second": 10.614, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0245746691871456, | |
| "grad_norm": 0.8697965145111084, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3199, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.213610586011342, | |
| "grad_norm": 0.7469469308853149, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0191, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.402646502835539, | |
| "grad_norm": 0.8800496459007263, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0606, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.5916824196597354, | |
| "grad_norm": 0.7127139568328857, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0988, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.780718336483932, | |
| "grad_norm": 0.9053574204444885, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1355, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9697542533081287, | |
| "grad_norm": 0.9217483997344971, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1624, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5651794871794872, | |
| "eval_loss": 1.9222664833068848, | |
| "eval_runtime": 5.7781, | |
| "eval_samples_per_second": 86.533, | |
| "eval_steps_per_second": 10.903, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.158790170132325, | |
| "grad_norm": 0.8502592444419861, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8795, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 0.8051011562347412, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8713, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.536862003780718, | |
| "grad_norm": 1.0111184120178223, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8974, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.725897920604915, | |
| "grad_norm": 0.9804710745811462, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9396, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.914933837429111, | |
| "grad_norm": 1.1629979610443115, | |
| "learning_rate": 0.0005, | |
| "loss": 0.964, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5642564102564103, | |
| "eval_loss": 1.9719713926315308, | |
| "eval_runtime": 6.1201, | |
| "eval_samples_per_second": 81.698, | |
| "eval_steps_per_second": 10.294, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.103969754253308, | |
| "grad_norm": 0.8145145177841187, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8208, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.293005671077505, | |
| "grad_norm": 0.7975667715072632, | |
| "learning_rate": 0.0005, | |
| "loss": 0.7022, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.482041587901701, | |
| "grad_norm": 0.8971696496009827, | |
| "learning_rate": 0.0005, | |
| "loss": 0.7435, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.671077504725898, | |
| "grad_norm": 1.0234569311141968, | |
| "learning_rate": 0.0005, | |
| "loss": 0.7798, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.8601134215500945, | |
| "grad_norm": 1.0058211088180542, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8117, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5647179487179487, | |
| "eval_loss": 2.001645565032959, | |
| "eval_runtime": 6.1282, | |
| "eval_samples_per_second": 81.59, | |
| "eval_steps_per_second": 10.28, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.049149338374291, | |
| "grad_norm": 0.9231036901473999, | |
| "learning_rate": 0.0005, | |
| "loss": 0.7673, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.238185255198488, | |
| "grad_norm": 0.9058907628059387, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5828, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.427221172022684, | |
| "grad_norm": 0.8709693551063538, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6272, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.616257088846881, | |
| "grad_norm": 0.9873590469360352, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6599, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.805293005671078, | |
| "grad_norm": 0.9257177114486694, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6918, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.994328922495274, | |
| "grad_norm": 1.0284956693649292, | |
| "learning_rate": 0.0005, | |
| "loss": 0.7242, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5638974358974359, | |
| "eval_loss": 2.0784664154052734, | |
| "eval_runtime": 6.1613, | |
| "eval_samples_per_second": 81.152, | |
| "eval_steps_per_second": 10.225, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.183364839319471, | |
| "grad_norm": 1.0036342144012451, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5099, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.3724007561436675, | |
| "grad_norm": 0.9070968627929688, | |
| "learning_rate": 0.0005, | |
| "loss": 0.541, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.561436672967864, | |
| "grad_norm": 1.035317063331604, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5681, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.750472589792061, | |
| "grad_norm": 1.0337880849838257, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5991, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.939508506616257, | |
| "grad_norm": 0.9707123637199402, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6381, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5644615384615385, | |
| "eval_loss": 2.0954012870788574, | |
| "eval_runtime": 5.9732, | |
| "eval_samples_per_second": 83.708, | |
| "eval_steps_per_second": 10.547, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.128544423440454, | |
| "grad_norm": 0.9024428129196167, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5113, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.31758034026465, | |
| "grad_norm": 0.8392314314842224, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4752, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.506616257088847, | |
| "grad_norm": 0.9310020208358765, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5088, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 1.115989327430725, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5358, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.88468809073724, | |
| "grad_norm": 1.1282142400741577, | |
| "learning_rate": 0.0005, | |
| "loss": 0.573, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5623076923076923, | |
| "eval_loss": 2.106734275817871, | |
| "eval_runtime": 5.9846, | |
| "eval_samples_per_second": 83.548, | |
| "eval_steps_per_second": 10.527, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.073724007561436, | |
| "grad_norm": 0.7718724012374878, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5165, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.262759924385634, | |
| "grad_norm": 1.0482347011566162, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4291, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.45179584120983, | |
| "grad_norm": 1.1258679628372192, | |
| "learning_rate": 0.0005, | |
| "loss": 0.467, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.640831758034027, | |
| "grad_norm": 1.0340867042541504, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4992, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 9.829867674858223, | |
| "grad_norm": 0.9999015927314758, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5269, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5646153846153846, | |
| "eval_loss": 2.1355865001678467, | |
| "eval_runtime": 5.6591, | |
| "eval_samples_per_second": 88.353, | |
| "eval_steps_per_second": 11.132, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.01890359168242, | |
| "grad_norm": 0.8213431239128113, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5325, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.207939508506616, | |
| "grad_norm": 1.023992896080017, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3966, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.396975425330814, | |
| "grad_norm": 0.8211169242858887, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4307, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.58601134215501, | |
| "grad_norm": 0.9888399839401245, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4604, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.775047258979207, | |
| "grad_norm": 0.9565384387969971, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4888, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 10.964083175803403, | |
| "grad_norm": 1.0463844537734985, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5144, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5616410256410257, | |
| "eval_loss": 2.195107936859131, | |
| "eval_runtime": 5.722, | |
| "eval_samples_per_second": 87.382, | |
| "eval_steps_per_second": 11.01, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.1531190926276, | |
| "grad_norm": 0.9159551858901978, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4094, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.342155009451796, | |
| "grad_norm": 0.9027566909790039, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4083, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.531190926275993, | |
| "grad_norm": 0.9325262904167175, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4339, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.720226843100189, | |
| "grad_norm": 1.0173691511154175, | |
| "learning_rate": 0.0005, | |
| "loss": 0.464, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.909262759924385, | |
| "grad_norm": 0.9742277264595032, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4887, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5631282051282052, | |
| "eval_loss": 2.1778857707977295, | |
| "eval_runtime": 6.0538, | |
| "eval_samples_per_second": 82.593, | |
| "eval_steps_per_second": 10.407, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.098298676748582, | |
| "grad_norm": 1.0347404479980469, | |
| "learning_rate": 0.0005, | |
| "loss": 0.438, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.287334593572778, | |
| "grad_norm": 0.8733549118041992, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3893, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.476370510396976, | |
| "grad_norm": 0.8693263530731201, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4142, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.665406427221171, | |
| "grad_norm": 1.1597702503204346, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4379, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 12.854442344045369, | |
| "grad_norm": 1.018873929977417, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4636, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.561076923076923, | |
| "eval_loss": 2.175729274749756, | |
| "eval_runtime": 5.683, | |
| "eval_samples_per_second": 87.981, | |
| "eval_steps_per_second": 11.086, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 0.8772552013397217, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4552, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.232514177693762, | |
| "grad_norm": 0.8089916110038757, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3698, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.421550094517958, | |
| "grad_norm": 1.0659642219543457, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3971, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.610586011342155, | |
| "grad_norm": 1.2195113897323608, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4213, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.799621928166351, | |
| "grad_norm": 1.1026618480682373, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4487, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 13.988657844990549, | |
| "grad_norm": 1.120952844619751, | |
| "learning_rate": 0.0005, | |
| "loss": 0.467, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5624102564102564, | |
| "eval_loss": 2.1781177520751953, | |
| "eval_runtime": 5.7664, | |
| "eval_samples_per_second": 86.709, | |
| "eval_steps_per_second": 10.925, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.177693761814744, | |
| "grad_norm": 0.9397806525230408, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3685, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.366729678638942, | |
| "grad_norm": 0.8353122472763062, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3828, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.555765595463138, | |
| "grad_norm": 1.0425219535827637, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4026, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 14.744801512287335, | |
| "grad_norm": 1.0660319328308105, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4287, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.93383742911153, | |
| "grad_norm": 1.131705403327942, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4613, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5611794871794872, | |
| "eval_loss": 2.231194019317627, | |
| "eval_runtime": 5.8033, | |
| "eval_samples_per_second": 86.158, | |
| "eval_steps_per_second": 10.856, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.122873345935728, | |
| "grad_norm": 0.8532615303993225, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3889, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.311909262759924, | |
| "grad_norm": 0.9682427048683167, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3675, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.500945179584122, | |
| "grad_norm": 0.9799471497535706, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3968, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.689981096408317, | |
| "grad_norm": 0.9278863072395325, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4198, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 15.879017013232515, | |
| "grad_norm": 1.0345618724822998, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4405, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.562923076923077, | |
| "eval_loss": 2.1799755096435547, | |
| "eval_runtime": 5.8871, | |
| "eval_samples_per_second": 84.931, | |
| "eval_steps_per_second": 10.701, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.068052930056712, | |
| "grad_norm": 0.9081653356552124, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4241, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.257088846880908, | |
| "grad_norm": 0.8579433560371399, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3634, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.446124763705104, | |
| "grad_norm": 1.042361855506897, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3837, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 16.6351606805293, | |
| "grad_norm": 1.032463550567627, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4071, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.8241965973535, | |
| "grad_norm": 1.0003693103790283, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4308, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5627692307692308, | |
| "eval_loss": 2.1960062980651855, | |
| "eval_runtime": 6.0711, | |
| "eval_samples_per_second": 82.358, | |
| "eval_steps_per_second": 10.377, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.013232514177695, | |
| "grad_norm": 0.793249249458313, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4471, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.20226843100189, | |
| "grad_norm": 0.9027940630912781, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3499, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.9174290299415588, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3698, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.58034026465028, | |
| "grad_norm": 1.135392665863037, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4027, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 17.76937618147448, | |
| "grad_norm": 0.9852614998817444, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4179, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.958412098298677, | |
| "grad_norm": 0.9487043619155884, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4401, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.560974358974359, | |
| "eval_loss": 2.2354702949523926, | |
| "eval_runtime": 5.7571, | |
| "eval_samples_per_second": 86.85, | |
| "eval_steps_per_second": 10.943, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.147448015122873, | |
| "grad_norm": 0.7868739366531372, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3637, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.33648393194707, | |
| "grad_norm": 0.962329626083374, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3642, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.525519848771268, | |
| "grad_norm": 0.9594590663909912, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3865, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.714555765595463, | |
| "grad_norm": 1.0574795007705688, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4081, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 18.90359168241966, | |
| "grad_norm": 1.0566685199737549, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4334, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5608205128205128, | |
| "eval_loss": 2.2380332946777344, | |
| "eval_runtime": 5.7808, | |
| "eval_samples_per_second": 86.493, | |
| "eval_steps_per_second": 10.898, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.092627599243855, | |
| "grad_norm": 0.6632242798805237, | |
| "learning_rate": 0.0005, | |
| "loss": 0.393, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.281663516068054, | |
| "grad_norm": 0.8786609768867493, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3528, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.47069943289225, | |
| "grad_norm": 1.0305085182189941, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3752, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 19.659735349716446, | |
| "grad_norm": 0.9528347849845886, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4023, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.84877126654064, | |
| "grad_norm": 0.9051763415336609, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4218, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.560974358974359, | |
| "eval_loss": 2.2417104244232178, | |
| "eval_runtime": 5.7479, | |
| "eval_samples_per_second": 86.989, | |
| "eval_steps_per_second": 10.961, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10580, | |
| "total_flos": 6.517631969856061e+17, | |
| "train_loss": 0.6977483630405708, | |
| "train_runtime": 24342.3236, | |
| "train_samples_per_second": 13.906, | |
| "train_steps_per_second": 0.435 | |
| } | |
| ], | |
| "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 | |
| } | |