| { | |
| "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": 0.2485148161649704, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 0.2885477840900421, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8505, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.3204280436038971, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8005, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.3903968632221222, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7786, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.37639909982681274, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7331, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6365112107623319, | |
| "eval_loss": 1.427091121673584, | |
| "eval_runtime": 6.6532, | |
| "eval_samples_per_second": 75.152, | |
| "eval_steps_per_second": 9.469, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.5027719736099243, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4811, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 0.6268895268440247, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3494, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.5936668515205383, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3268, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.61969393491745, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3109, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 0.6521849036216736, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3037, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6845739910313902, | |
| "eval_loss": 1.0687369108200073, | |
| "eval_runtime": 6.6004, | |
| "eval_samples_per_second": 75.753, | |
| "eval_steps_per_second": 9.545, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.6594504117965698, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0942, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 0.7692387104034424, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8554, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 0.7720690965652466, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8748, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 0.8158366680145264, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8853, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 0.8173924684524536, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8818, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7216233183856502, | |
| "eval_loss": 0.8142201900482178, | |
| "eval_runtime": 7.0748, | |
| "eval_samples_per_second": 70.673, | |
| "eval_steps_per_second": 8.905, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0245746691871456, | |
| "grad_norm": 0.8073566555976868, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8192, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.213610586011342, | |
| "grad_norm": 0.7835751175880432, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5222, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.402646502835539, | |
| "grad_norm": 0.8399431109428406, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5699, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.5916824196597354, | |
| "grad_norm": 0.8880508542060852, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6016, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.780718336483932, | |
| "grad_norm": 1.0154162645339966, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6219, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9697542533081287, | |
| "grad_norm": 0.8903167843818665, | |
| "learning_rate": 0.0005, | |
| "loss": 0.6397, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7470134529147983, | |
| "eval_loss": 0.6636369824409485, | |
| "eval_runtime": 6.2123, | |
| "eval_samples_per_second": 80.485, | |
| "eval_steps_per_second": 10.141, | |
| "step": 2116 | |
| }, | |
| { | |
| "epoch": 4.158790170132325, | |
| "grad_norm": 0.7903006672859192, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4242, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 0.8569557666778564, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4102, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.536862003780718, | |
| "grad_norm": 0.8884899616241455, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4266, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.725897920604915, | |
| "grad_norm": 0.9051417708396912, | |
| "learning_rate": 0.0005, | |
| "loss": 0.456, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.914933837429111, | |
| "grad_norm": 0.8909860849380493, | |
| "learning_rate": 0.0005, | |
| "loss": 0.4735, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7667174887892376, | |
| "eval_loss": 0.5546546578407288, | |
| "eval_runtime": 6.1339, | |
| "eval_samples_per_second": 81.514, | |
| "eval_steps_per_second": 10.271, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 5.103969754253308, | |
| "grad_norm": 0.8669963479042053, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3835, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.293005671077505, | |
| "grad_norm": 0.7596257925033569, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3091, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.482041587901701, | |
| "grad_norm": 0.8810439705848694, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3356, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.671077504725898, | |
| "grad_norm": 0.866624653339386, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3585, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.8601134215500945, | |
| "grad_norm": 1.0449254512786865, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3798, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7763946188340807, | |
| "eval_loss": 0.5001726746559143, | |
| "eval_runtime": 7.0927, | |
| "eval_samples_per_second": 70.495, | |
| "eval_steps_per_second": 8.882, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 6.049149338374291, | |
| "grad_norm": 0.7229480147361755, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3581, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.238185255198488, | |
| "grad_norm": 0.8279250264167786, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2561, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.427221172022684, | |
| "grad_norm": 0.7875299453735352, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2796, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.616257088846881, | |
| "grad_norm": 0.9155901074409485, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3019, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.805293005671078, | |
| "grad_norm": 0.7736573815345764, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3212, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.994328922495274, | |
| "grad_norm": 1.005894422531128, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3409, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7801165919282511, | |
| "eval_loss": 0.48504143953323364, | |
| "eval_runtime": 6.6015, | |
| "eval_samples_per_second": 75.74, | |
| "eval_steps_per_second": 9.543, | |
| "step": 3703 | |
| }, | |
| { | |
| "epoch": 7.183364839319471, | |
| "grad_norm": 0.7929958701133728, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2328, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.3724007561436675, | |
| "grad_norm": 0.8098764419555664, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2474, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.561436672967864, | |
| "grad_norm": 0.7900761365890503, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2645, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.750472589792061, | |
| "grad_norm": 0.8573282957077026, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2817, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.939508506616257, | |
| "grad_norm": 0.9526512622833252, | |
| "learning_rate": 0.0005, | |
| "loss": 0.3054, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7835156950672646, | |
| "eval_loss": 0.46914222836494446, | |
| "eval_runtime": 6.6138, | |
| "eval_samples_per_second": 75.599, | |
| "eval_steps_per_second": 9.526, | |
| "step": 4232 | |
| }, | |
| { | |
| "epoch": 8.128544423440454, | |
| "grad_norm": 0.7902705669403076, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2491, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 8.31758034026465, | |
| "grad_norm": 0.7070533633232117, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2221, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.506616257088847, | |
| "grad_norm": 0.8409714102745056, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2471, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "grad_norm": 0.8545922636985779, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2634, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.88468809073724, | |
| "grad_norm": 0.9577164053916931, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2803, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.785865470852018, | |
| "eval_loss": 0.4636962115764618, | |
| "eval_runtime": 6.6143, | |
| "eval_samples_per_second": 75.594, | |
| "eval_steps_per_second": 9.525, | |
| "step": 4761 | |
| }, | |
| { | |
| "epoch": 9.073724007561436, | |
| "grad_norm": 0.7526606917381287, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2587, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.262759924385634, | |
| "grad_norm": 0.881596565246582, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2101, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 9.45179584120983, | |
| "grad_norm": 0.8427467942237854, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2218, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.640831758034027, | |
| "grad_norm": 0.8922020792961121, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2471, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 9.829867674858223, | |
| "grad_norm": 0.9664803147315979, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2637, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7877040358744395, | |
| "eval_loss": 0.45324718952178955, | |
| "eval_runtime": 6.6378, | |
| "eval_samples_per_second": 75.326, | |
| "eval_steps_per_second": 9.491, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 10.01890359168242, | |
| "grad_norm": 0.7240998148918152, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2689, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 10.207939508506616, | |
| "grad_norm": 0.8262352347373962, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2022, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.396975425330814, | |
| "grad_norm": 0.8235085010528564, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2113, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 10.58601134215501, | |
| "grad_norm": 0.9357341527938843, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2282, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.775047258979207, | |
| "grad_norm": 0.8699725866317749, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2499, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 10.964083175803403, | |
| "grad_norm": 0.9279125928878784, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2661, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7878834080717488, | |
| "eval_loss": 0.4667767286300659, | |
| "eval_runtime": 6.6569, | |
| "eval_samples_per_second": 75.11, | |
| "eval_steps_per_second": 9.464, | |
| "step": 5819 | |
| }, | |
| { | |
| "epoch": 11.1531190926276, | |
| "grad_norm": 0.7377325892448425, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2085, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 11.342155009451796, | |
| "grad_norm": 0.8089397549629211, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2027, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.531190926275993, | |
| "grad_norm": 0.8074979782104492, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2183, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 11.720226843100189, | |
| "grad_norm": 0.8173251152038574, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2373, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.909262759924385, | |
| "grad_norm": 0.8041278123855591, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2513, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7892825112107623, | |
| "eval_loss": 0.4647226631641388, | |
| "eval_runtime": 6.1139, | |
| "eval_samples_per_second": 81.78, | |
| "eval_steps_per_second": 10.304, | |
| "step": 6348 | |
| }, | |
| { | |
| "epoch": 12.098298676748582, | |
| "grad_norm": 0.7303513288497925, | |
| "learning_rate": 0.0005, | |
| "loss": 0.221, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.287334593572778, | |
| "grad_norm": 0.8052433133125305, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1969, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 12.476370510396976, | |
| "grad_norm": 0.9060468077659607, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2078, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.665406427221171, | |
| "grad_norm": 0.8090572357177734, | |
| "learning_rate": 0.0005, | |
| "loss": 0.228, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 12.854442344045369, | |
| "grad_norm": 0.8910106420516968, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2424, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7896681614349775, | |
| "eval_loss": 0.46145549416542053, | |
| "eval_runtime": 7.2577, | |
| "eval_samples_per_second": 68.892, | |
| "eval_steps_per_second": 8.68, | |
| "step": 6877 | |
| }, | |
| { | |
| "epoch": 13.043478260869565, | |
| "grad_norm": 0.7991313934326172, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2395, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 13.232514177693762, | |
| "grad_norm": 0.6423863768577576, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1903, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.421550094517958, | |
| "grad_norm": 0.8329731225967407, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2021, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 13.610586011342155, | |
| "grad_norm": 0.8105093240737915, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2198, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.799621928166351, | |
| "grad_norm": 0.8448224067687988, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2317, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 13.988657844990549, | |
| "grad_norm": 0.8879454731941223, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2499, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7894170403587444, | |
| "eval_loss": 0.4546208679676056, | |
| "eval_runtime": 6.6163, | |
| "eval_samples_per_second": 75.571, | |
| "eval_steps_per_second": 9.522, | |
| "step": 7406 | |
| }, | |
| { | |
| "epoch": 14.177693761814744, | |
| "grad_norm": 0.7737810611724854, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1916, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 14.366729678638942, | |
| "grad_norm": 0.8142408132553101, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1968, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.555765595463138, | |
| "grad_norm": 0.8974208831787109, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2089, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 14.744801512287335, | |
| "grad_norm": 0.8206737637519836, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2268, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.93383742911153, | |
| "grad_norm": 0.8444597721099854, | |
| "learning_rate": 0.0005, | |
| "loss": 0.235, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7896233183856503, | |
| "eval_loss": 0.4667895436286926, | |
| "eval_runtime": 6.086, | |
| "eval_samples_per_second": 82.156, | |
| "eval_steps_per_second": 10.352, | |
| "step": 7935 | |
| }, | |
| { | |
| "epoch": 15.122873345935728, | |
| "grad_norm": 0.8915912508964539, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2052, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.311909262759924, | |
| "grad_norm": 0.8511343598365784, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1887, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 15.500945179584122, | |
| "grad_norm": 0.8454872965812683, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1967, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.689981096408317, | |
| "grad_norm": 0.7670953869819641, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2177, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 15.879017013232515, | |
| "grad_norm": 0.8440077900886536, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2317, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7913183856502242, | |
| "eval_loss": 0.4509560465812683, | |
| "eval_runtime": 6.1751, | |
| "eval_samples_per_second": 80.97, | |
| "eval_steps_per_second": 10.202, | |
| "step": 8464 | |
| }, | |
| { | |
| "epoch": 16.068052930056712, | |
| "grad_norm": 0.7304275035858154, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2194, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 16.257088846880908, | |
| "grad_norm": 0.6791017651557922, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1821, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.446124763705104, | |
| "grad_norm": 0.7813127040863037, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1952, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 16.6351606805293, | |
| "grad_norm": 0.8328565955162048, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2107, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.8241965973535, | |
| "grad_norm": 0.7615934014320374, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2225, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7914618834080718, | |
| "eval_loss": 0.44965559244155884, | |
| "eval_runtime": 6.0983, | |
| "eval_samples_per_second": 81.99, | |
| "eval_steps_per_second": 10.331, | |
| "step": 8993 | |
| }, | |
| { | |
| "epoch": 17.013232514177695, | |
| "grad_norm": 0.6526479125022888, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2357, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 17.20226843100189, | |
| "grad_norm": 0.8103135228157043, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1796, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.7900732159614563, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1894, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.58034026465028, | |
| "grad_norm": 0.8373563885688782, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2048, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 17.76937618147448, | |
| "grad_norm": 0.8260546922683716, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2213, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.958412098298677, | |
| "grad_norm": 0.8767030835151672, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2358, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7915695067264574, | |
| "eval_loss": 0.4474881589412689, | |
| "eval_runtime": 6.1364, | |
| "eval_samples_per_second": 81.481, | |
| "eval_steps_per_second": 10.267, | |
| "step": 9522 | |
| }, | |
| { | |
| "epoch": 18.147448015122873, | |
| "grad_norm": 0.7694826722145081, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1931, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 18.33648393194707, | |
| "grad_norm": 0.7357311248779297, | |
| "learning_rate": 0.0005, | |
| "loss": 0.187, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 18.525519848771268, | |
| "grad_norm": 0.7614474892616272, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1971, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.714555765595463, | |
| "grad_norm": 0.8685888648033142, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2127, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 18.90359168241966, | |
| "grad_norm": 0.8261685371398926, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2253, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7918116591928251, | |
| "eval_loss": 0.45286619663238525, | |
| "eval_runtime": 7.1709, | |
| "eval_samples_per_second": 69.726, | |
| "eval_steps_per_second": 8.785, | |
| "step": 10051 | |
| }, | |
| { | |
| "epoch": 19.092627599243855, | |
| "grad_norm": 0.6555222868919373, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2055, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 19.281663516068054, | |
| "grad_norm": 0.6589702367782593, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1828, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.47069943289225, | |
| "grad_norm": 0.8058968782424927, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1932, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 19.659735349716446, | |
| "grad_norm": 0.7728559374809265, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2066, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.84877126654064, | |
| "grad_norm": 0.7878997325897217, | |
| "learning_rate": 0.0005, | |
| "loss": 0.2172, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7918744394618834, | |
| "eval_loss": 0.45650961995124817, | |
| "eval_runtime": 6.6035, | |
| "eval_samples_per_second": 75.718, | |
| "eval_steps_per_second": 9.54, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10580, | |
| "total_flos": 9.64242245391745e+17, | |
| "train_loss": 0.42894544871858475, | |
| "train_runtime": 22582.9978, | |
| "train_samples_per_second": 14.989, | |
| "train_steps_per_second": 0.468 | |
| } | |
| ], | |
| "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 | |
| } | |