| { | |
| "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": 0.24623066186904907, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0175, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 0.3154822587966919, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9025, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.2886754274368286, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8754, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.33380141854286194, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8795, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.3052491247653961, | |
| "learning_rate": 3e-05, | |
| "loss": 1.86, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.605677130044843, | |
| "eval_loss": 1.6920065879821777, | |
| "eval_runtime": 6.0828, | |
| "eval_samples_per_second": 82.199, | |
| "eval_steps_per_second": 10.357, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.38703393936157227, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8651, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 0.5257138609886169, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8238, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.5151342749595642, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8252, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.4788012206554413, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8178, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 0.506370484828949, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8271, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6118744394618834, | |
| "eval_loss": 1.6425946950912476, | |
| "eval_runtime": 6.042, | |
| "eval_samples_per_second": 82.754, | |
| "eval_steps_per_second": 10.427, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.537693440914154, | |
| "learning_rate": 3e-05, | |
| "loss": 1.795, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 0.6806929111480713, | |
| "learning_rate": 3e-05, | |
| "loss": 1.773, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 0.7064961194992065, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7609, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 0.7571425437927246, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7536, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 0.8069006204605103, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7324, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6163408071748879, | |
| "eval_loss": 1.59978187084198, | |
| "eval_runtime": 6.4372, | |
| "eval_samples_per_second": 77.673, | |
| "eval_steps_per_second": 9.787, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0245746691871456, | |
| "grad_norm": 0.7909080982208252, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7206, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.213610586011342, | |
| "grad_norm": 0.8461490869522095, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6696, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.402646502835539, | |
| "grad_norm": 0.9591927528381348, | |
| "learning_rate": 3e-05, | |
| "loss": 1.672, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.5916824196597354, | |
| "grad_norm": 1.0703200101852417, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6712, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.780718336483932, | |
| "grad_norm": 0.9409723281860352, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6877, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9697542533081287, | |
| "grad_norm": 0.9087811708450317, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6818, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.621982062780269, | |
| "eval_loss": 1.5579897165298462, | |
| "eval_runtime": 6.0535, | |
| "eval_samples_per_second": 82.597, | |
| "eval_steps_per_second": 10.407, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.158790170132325, | |
| "grad_norm": 1.0507543087005615, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6195, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 1.1318222284317017, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6039, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.536862003780718, | |
| "grad_norm": 1.1028773784637451, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5924, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.725897920604915, | |
| "grad_norm": 1.1365679502487183, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5903, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.914933837429111, | |
| "grad_norm": 1.0776621103286743, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5864, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6278206278026905, | |
| "eval_loss": 1.5211377143859863, | |
| "eval_runtime": 6.0103, | |
| "eval_samples_per_second": 83.19, | |
| "eval_steps_per_second": 10.482, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.103969754253308, | |
| "grad_norm": 1.1829136610031128, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5478, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.293005671077505, | |
| "grad_norm": 1.1871124505996704, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5151, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.482041587901701, | |
| "grad_norm": 1.2089450359344482, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5159, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.671077504725898, | |
| "grad_norm": 1.3170595169067383, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5241, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.8601134215500945, | |
| "grad_norm": 1.439405083656311, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5204, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6326726457399103, | |
| "eval_loss": 1.4863040447235107, | |
| "eval_runtime": 6.0604, | |
| "eval_samples_per_second": 82.502, | |
| "eval_steps_per_second": 10.395, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.049149338374291, | |
| "grad_norm": 1.3573708534240723, | |
| "learning_rate": 3e-05, | |
| "loss": 1.506, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.238185255198488, | |
| "grad_norm": 1.3852615356445312, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4401, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.427221172022684, | |
| "grad_norm": 1.391204595565796, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4469, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.616257088846881, | |
| "grad_norm": 1.5318756103515625, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4608, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.805293005671078, | |
| "grad_norm": 1.4233124256134033, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4419, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.994328922495274, | |
| "grad_norm": 1.4444003105163574, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4481, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6371928251121076, | |
| "eval_loss": 1.4516844749450684, | |
| "eval_runtime": 6.0218, | |
| "eval_samples_per_second": 83.032, | |
| "eval_steps_per_second": 10.462, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.183364839319471, | |
| "grad_norm": 1.607857346534729, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3742, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.3724007561436675, | |
| "grad_norm": 1.5741080045700073, | |
| "learning_rate": 3e-05, | |
| "loss": 1.383, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.561436672967864, | |
| "grad_norm": 1.6515594720840454, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3734, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.750472589792061, | |
| "grad_norm": 1.704078197479248, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3527, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.939508506616257, | |
| "grad_norm": 1.7636491060256958, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3768, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6429058295964125, | |
| "eval_loss": 1.4120872020721436, | |
| "eval_runtime": 6.7429, | |
| "eval_samples_per_second": 74.152, | |
| "eval_steps_per_second": 9.343, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.128544423440454, | |
| "grad_norm": 1.691927433013916, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3254, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.31758034026465, | |
| "grad_norm": 1.695786476135254, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2819, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.506616257088847, | |
| "grad_norm": 1.8343569040298462, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3225, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 1.678654670715332, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2961, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.88468809073724, | |
| "grad_norm": 2.0212690830230713, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2946, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6481973094170403, | |
| "eval_loss": 1.3739463090896606, | |
| "eval_runtime": 6.0696, | |
| "eval_samples_per_second": 82.377, | |
| "eval_steps_per_second": 10.38, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.073724007561436, | |
| "grad_norm": 1.8944847583770752, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2831, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.262759924385634, | |
| "grad_norm": 1.9275157451629639, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2022, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.45179584120983, | |
| "grad_norm": 1.927822470664978, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2259, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.640831758034027, | |
| "grad_norm": 1.8706145286560059, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2231, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 9.829867674858223, | |
| "grad_norm": 1.858546257019043, | |
| "learning_rate": 3e-05, | |
| "loss": 1.243, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.653237668161435, | |
| "eval_loss": 1.3363806009292603, | |
| "eval_runtime": 6.6274, | |
| "eval_samples_per_second": 75.444, | |
| "eval_steps_per_second": 9.506, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.01890359168242, | |
| "grad_norm": 2.2799806594848633, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2223, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.207939508506616, | |
| "grad_norm": 2.0888822078704834, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1567, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.396975425330814, | |
| "grad_norm": 2.1389222145080566, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1421, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.58601134215501, | |
| "grad_norm": 2.253117322921753, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1525, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.775047258979207, | |
| "grad_norm": 2.3171472549438477, | |
| "learning_rate": 3e-05, | |
| "loss": 1.168, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 10.964083175803403, | |
| "grad_norm": 2.1451621055603027, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1425, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.6593991031390134, | |
| "eval_loss": 1.2968207597732544, | |
| "eval_runtime": 5.9969, | |
| "eval_samples_per_second": 83.377, | |
| "eval_steps_per_second": 10.506, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.1531190926276, | |
| "grad_norm": 2.0139310359954834, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0814, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.342155009451796, | |
| "grad_norm": 2.2693676948547363, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0731, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.531190926275993, | |
| "grad_norm": 2.4319376945495605, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0842, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.720226843100189, | |
| "grad_norm": 2.103640079498291, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0839, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.909262759924385, | |
| "grad_norm": 2.309508800506592, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0847, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.6652017937219731, | |
| "eval_loss": 1.2538820505142212, | |
| "eval_runtime": 6.0994, | |
| "eval_samples_per_second": 81.975, | |
| "eval_steps_per_second": 10.329, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.098298676748582, | |
| "grad_norm": 2.184951066970825, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0154, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.287334593572778, | |
| "grad_norm": 2.6142184734344482, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0129, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.476370510396976, | |
| "grad_norm": 2.377812147140503, | |
| "learning_rate": 3e-05, | |
| "loss": 1.01, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.665406427221171, | |
| "grad_norm": 2.252695083618164, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0132, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 12.854442344045369, | |
| "grad_norm": 2.4628734588623047, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0152, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.670609865470852, | |
| "eval_loss": 1.216369390487671, | |
| "eval_runtime": 6.115, | |
| "eval_samples_per_second": 81.766, | |
| "eval_steps_per_second": 10.302, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 2.2027320861816406, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0047, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.232514177693762, | |
| "grad_norm": 2.1473910808563232, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9308, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.421550094517958, | |
| "grad_norm": 2.633467435836792, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9265, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.610586011342155, | |
| "grad_norm": 2.5691654682159424, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9485, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.799621928166351, | |
| "grad_norm": 2.6028811931610107, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9502, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 13.988657844990549, | |
| "grad_norm": 2.6901166439056396, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9498, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.6757937219730942, | |
| "eval_loss": 1.176974892616272, | |
| "eval_runtime": 6.0622, | |
| "eval_samples_per_second": 82.479, | |
| "eval_steps_per_second": 10.392, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.177693761814744, | |
| "grad_norm": 2.624717950820923, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8708, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.366729678638942, | |
| "grad_norm": 2.5228078365325928, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8685, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.555765595463138, | |
| "grad_norm": 2.8440725803375244, | |
| "learning_rate": 3e-05, | |
| "loss": 0.88, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 14.744801512287335, | |
| "grad_norm": 2.5001230239868164, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8968, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.93383742911153, | |
| "grad_norm": 2.868446111679077, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8652, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.6820717488789237, | |
| "eval_loss": 1.1323232650756836, | |
| "eval_runtime": 6.0411, | |
| "eval_samples_per_second": 82.767, | |
| "eval_steps_per_second": 10.429, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.122873345935728, | |
| "grad_norm": 3.236562728881836, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8354, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.311909262759924, | |
| "grad_norm": 3.0811691284179688, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8042, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.500945179584122, | |
| "grad_norm": 2.6125237941741943, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8097, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.689981096408317, | |
| "grad_norm": 2.8925766944885254, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8133, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 15.879017013232515, | |
| "grad_norm": 2.927067518234253, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8265, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.690134529147982, | |
| "eval_loss": 1.0825910568237305, | |
| "eval_runtime": 6.0821, | |
| "eval_samples_per_second": 82.209, | |
| "eval_steps_per_second": 10.358, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.068052930056712, | |
| "grad_norm": 2.8401355743408203, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7933, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.257088846880908, | |
| "grad_norm": 3.2201480865478516, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7449, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.446124763705104, | |
| "grad_norm": 3.2903711795806885, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7565, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 16.6351606805293, | |
| "grad_norm": 3.153130292892456, | |
| "learning_rate": 3e-05, | |
| "loss": 0.773, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.8241965973535, | |
| "grad_norm": 2.8367865085601807, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7432, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.6959910313901345, | |
| "eval_loss": 1.04633367061615, | |
| "eval_runtime": 6.1125, | |
| "eval_samples_per_second": 81.8, | |
| "eval_steps_per_second": 10.307, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.013232514177695, | |
| "grad_norm": 3.268256664276123, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7576, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.20226843100189, | |
| "grad_norm": 3.0888004302978516, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6879, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 2.908324718475342, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6869, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.58034026465028, | |
| "grad_norm": 4.358062267303467, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7027, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 17.76937618147448, | |
| "grad_norm": 3.1521716117858887, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7188, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.958412098298677, | |
| "grad_norm": 3.1939432621002197, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7106, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7021973094170404, | |
| "eval_loss": 1.0097612142562866, | |
| "eval_runtime": 5.6404, | |
| "eval_samples_per_second": 88.646, | |
| "eval_steps_per_second": 11.169, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.147448015122873, | |
| "grad_norm": 3.3058857917785645, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6526, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.33648393194707, | |
| "grad_norm": 2.845630645751953, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6375, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.525519848771268, | |
| "grad_norm": 3.524501323699951, | |
| "learning_rate": 3e-05, | |
| "loss": 0.65, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.714555765595463, | |
| "grad_norm": 3.5869174003601074, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6529, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 18.90359168241966, | |
| "grad_norm": 2.8486292362213135, | |
| "learning_rate": 3e-05, | |
| "loss": 0.669, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7077937219730942, | |
| "eval_loss": 0.9695693254470825, | |
| "eval_runtime": 6.1993, | |
| "eval_samples_per_second": 80.654, | |
| "eval_steps_per_second": 10.162, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.092627599243855, | |
| "grad_norm": 2.8552756309509277, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6183, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.281663516068054, | |
| "grad_norm": 2.949084758758545, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5824, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.47069943289225, | |
| "grad_norm": 3.137747049331665, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6093, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 19.659735349716446, | |
| "grad_norm": 2.7827863693237305, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6155, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.84877126654064, | |
| "grad_norm": 3.2223191261291504, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6043, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7134618834080717, | |
| "eval_loss": 0.9358564019203186, | |
| "eval_runtime": 6.1149, | |
| "eval_samples_per_second": 81.767, | |
| "eval_steps_per_second": 10.303, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10580, | |
| "total_flos": 9.64242245391745e+17, | |
| "train_loss": 1.2100651977192927, | |
| "train_runtime": 22841.5406, | |
| "train_samples_per_second": 14.819, | |
| "train_steps_per_second": 0.463 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10580, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 9.64242245391745e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |