| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.995418771290967, | |
| "eval_steps": 500, | |
| "global_step": 25500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.915426781519186e-06, | |
| "loss": 0.2741, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.038767650723457336, | |
| "eval_runtime": 118.2386, | |
| "eval_samples_per_second": 55.329, | |
| "eval_steps_per_second": 6.918, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 7.830853563038372e-06, | |
| "loss": 0.0414, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_loss": 0.02589261531829834, | |
| "eval_runtime": 118.8851, | |
| "eval_samples_per_second": 55.028, | |
| "eval_steps_per_second": 6.881, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.908086719973622e-06, | |
| "loss": 0.0321, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_loss": 0.023959027603268623, | |
| "eval_runtime": 119.5631, | |
| "eval_samples_per_second": 54.716, | |
| "eval_steps_per_second": 6.842, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.702003132470531e-06, | |
| "loss": 0.0291, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_loss": 0.022091126069426537, | |
| "eval_runtime": 118.9906, | |
| "eval_samples_per_second": 54.979, | |
| "eval_steps_per_second": 6.874, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.49591954496744e-06, | |
| "loss": 0.0289, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_loss": 0.021428581327199936, | |
| "eval_runtime": 119.0226, | |
| "eval_samples_per_second": 54.964, | |
| "eval_steps_per_second": 6.873, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.289835957464349e-06, | |
| "loss": 0.0262, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_loss": 0.021141424775123596, | |
| "eval_runtime": 118.9536, | |
| "eval_samples_per_second": 54.996, | |
| "eval_steps_per_second": 6.877, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.083752369961257e-06, | |
| "loss": 0.0247, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_loss": 0.01918744668364525, | |
| "eval_runtime": 119.1755, | |
| "eval_samples_per_second": 54.894, | |
| "eval_steps_per_second": 6.864, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 8.877668782458165e-06, | |
| "loss": 0.0242, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_loss": 0.01971018686890602, | |
| "eval_runtime": 118.8936, | |
| "eval_samples_per_second": 55.024, | |
| "eval_steps_per_second": 6.88, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 8.671585194955074e-06, | |
| "loss": 0.0226, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_loss": 0.019149309024214745, | |
| "eval_runtime": 118.5478, | |
| "eval_samples_per_second": 55.184, | |
| "eval_steps_per_second": 6.9, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.465501607451984e-06, | |
| "loss": 0.0241, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_loss": 0.019563956186175346, | |
| "eval_runtime": 117.7652, | |
| "eval_samples_per_second": 55.551, | |
| "eval_steps_per_second": 6.946, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.259418019948892e-06, | |
| "loss": 0.0233, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_loss": 0.020009223371744156, | |
| "eval_runtime": 118.0111, | |
| "eval_samples_per_second": 55.435, | |
| "eval_steps_per_second": 6.932, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.053334432445801e-06, | |
| "loss": 0.0224, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_loss": 0.019076339900493622, | |
| "eval_runtime": 118.6811, | |
| "eval_samples_per_second": 55.123, | |
| "eval_steps_per_second": 6.892, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.84725084494271e-06, | |
| "loss": 0.0225, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.018951497972011566, | |
| "eval_runtime": 118.9565, | |
| "eval_samples_per_second": 54.995, | |
| "eval_steps_per_second": 6.876, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 7.641167257439618e-06, | |
| "loss": 0.022, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 0.018673894926905632, | |
| "eval_runtime": 118.6206, | |
| "eval_samples_per_second": 55.151, | |
| "eval_steps_per_second": 6.896, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 7.435083669936527e-06, | |
| "loss": 0.0208, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 0.018179820850491524, | |
| "eval_runtime": 118.8159, | |
| "eval_samples_per_second": 55.06, | |
| "eval_steps_per_second": 6.885, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 7.229000082433436e-06, | |
| "loss": 0.022, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 0.01761673204600811, | |
| "eval_runtime": 119.0018, | |
| "eval_samples_per_second": 54.974, | |
| "eval_steps_per_second": 6.874, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.022916494930344e-06, | |
| "loss": 0.0225, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.01772836409509182, | |
| "eval_runtime": 119.0025, | |
| "eval_samples_per_second": 54.974, | |
| "eval_steps_per_second": 6.874, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 6.816832907427253e-06, | |
| "loss": 0.0174, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_loss": 0.01802617870271206, | |
| "eval_runtime": 118.9086, | |
| "eval_samples_per_second": 55.017, | |
| "eval_steps_per_second": 6.879, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 6.610749319924161e-06, | |
| "loss": 0.0172, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_loss": 0.017970656976103783, | |
| "eval_runtime": 118.9682, | |
| "eval_samples_per_second": 54.99, | |
| "eval_steps_per_second": 6.876, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 6.404665732421071e-06, | |
| "loss": 0.0167, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_loss": 0.018082452937960625, | |
| "eval_runtime": 118.2637, | |
| "eval_samples_per_second": 55.317, | |
| "eval_steps_per_second": 6.917, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 6.1985821449179794e-06, | |
| "loss": 0.0169, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_loss": 0.0184369795024395, | |
| "eval_runtime": 118.1503, | |
| "eval_samples_per_second": 55.37, | |
| "eval_steps_per_second": 6.923, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 5.992498557414887e-06, | |
| "loss": 0.0172, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_loss": 0.017873156815767288, | |
| "eval_runtime": 118.4849, | |
| "eval_samples_per_second": 55.214, | |
| "eval_steps_per_second": 6.904, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 5.786414969911797e-06, | |
| "loss": 0.0165, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "eval_loss": 0.018475396558642387, | |
| "eval_runtime": 118.2528, | |
| "eval_samples_per_second": 55.322, | |
| "eval_steps_per_second": 6.917, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 5.5803313824087056e-06, | |
| "loss": 0.0182, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_loss": 0.01798292063176632, | |
| "eval_runtime": 118.8845, | |
| "eval_samples_per_second": 55.028, | |
| "eval_steps_per_second": 6.881, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 5.374247794905613e-06, | |
| "loss": 0.0168, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_loss": 0.01794307678937912, | |
| "eval_runtime": 118.5742, | |
| "eval_samples_per_second": 55.172, | |
| "eval_steps_per_second": 6.899, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 5.168164207402523e-06, | |
| "loss": 0.0174, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_loss": 0.017786763608455658, | |
| "eval_runtime": 118.992, | |
| "eval_samples_per_second": 54.978, | |
| "eval_steps_per_second": 6.874, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 4.962080619899432e-06, | |
| "loss": 0.0161, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_loss": 0.017610570415854454, | |
| "eval_runtime": 118.7482, | |
| "eval_samples_per_second": 55.091, | |
| "eval_steps_per_second": 6.889, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 4.75599703239634e-06, | |
| "loss": 0.0171, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 0.018315177410840988, | |
| "eval_runtime": 118.5603, | |
| "eval_samples_per_second": 55.179, | |
| "eval_steps_per_second": 6.899, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 4.549913444893249e-06, | |
| "loss": 0.0159, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_loss": 0.01768197864294052, | |
| "eval_runtime": 118.811, | |
| "eval_samples_per_second": 55.062, | |
| "eval_steps_per_second": 6.885, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 4.343829857390158e-06, | |
| "loss": 0.0157, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_loss": 0.01793872006237507, | |
| "eval_runtime": 117.8803, | |
| "eval_samples_per_second": 55.497, | |
| "eval_steps_per_second": 6.939, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 4.1377462698870665e-06, | |
| "loss": 0.0163, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "eval_loss": 0.0176596250385046, | |
| "eval_runtime": 117.7426, | |
| "eval_samples_per_second": 55.562, | |
| "eval_steps_per_second": 6.947, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 3.931662682383975e-06, | |
| "loss": 0.017, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "eval_loss": 0.017588863149285316, | |
| "eval_runtime": 118.319, | |
| "eval_samples_per_second": 55.291, | |
| "eval_steps_per_second": 6.914, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 3.725579094880884e-06, | |
| "loss": 0.0163, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "eval_loss": 0.0175350159406662, | |
| "eval_runtime": 118.676, | |
| "eval_samples_per_second": 55.125, | |
| "eval_steps_per_second": 6.893, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.5194955073777926e-06, | |
| "loss": 0.0162, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.017314020544290543, | |
| "eval_runtime": 118.3195, | |
| "eval_samples_per_second": 55.291, | |
| "eval_steps_per_second": 6.913, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 3.3134119198747017e-06, | |
| "loss": 0.0144, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "eval_loss": 0.017457639798521996, | |
| "eval_runtime": 118.5977, | |
| "eval_samples_per_second": 55.161, | |
| "eval_steps_per_second": 6.897, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 3.1073283323716105e-06, | |
| "loss": 0.0139, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "eval_loss": 0.01773645542562008, | |
| "eval_runtime": 118.6104, | |
| "eval_samples_per_second": 55.155, | |
| "eval_steps_per_second": 6.897, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2.9012447448685187e-06, | |
| "loss": 0.0145, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_loss": 0.0180921982973814, | |
| "eval_runtime": 118.3455, | |
| "eval_samples_per_second": 55.279, | |
| "eval_steps_per_second": 6.912, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 2.6951611573654274e-06, | |
| "loss": 0.0138, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "eval_loss": 0.01829521358013153, | |
| "eval_runtime": 118.9023, | |
| "eval_samples_per_second": 55.02, | |
| "eval_steps_per_second": 6.88, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 2.489077569862336e-06, | |
| "loss": 0.0131, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "eval_loss": 0.01797698438167572, | |
| "eval_runtime": 118.4871, | |
| "eval_samples_per_second": 55.213, | |
| "eval_steps_per_second": 6.904, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 2.2829939823592453e-06, | |
| "loss": 0.0135, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_loss": 0.018181076273322105, | |
| "eval_runtime": 117.5911, | |
| "eval_samples_per_second": 55.633, | |
| "eval_steps_per_second": 6.956, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 2.076910394856154e-06, | |
| "loss": 0.0134, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "eval_loss": 0.017410971224308014, | |
| "eval_runtime": 117.7062, | |
| "eval_samples_per_second": 55.579, | |
| "eval_steps_per_second": 6.95, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 1.8708268073530625e-06, | |
| "loss": 0.0139, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "eval_loss": 0.01747230626642704, | |
| "eval_runtime": 117.9312, | |
| "eval_samples_per_second": 55.473, | |
| "eval_steps_per_second": 6.936, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 1.6647432198499714e-06, | |
| "loss": 0.013, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_loss": 0.017725400626659393, | |
| "eval_runtime": 118.3008, | |
| "eval_samples_per_second": 55.3, | |
| "eval_steps_per_second": 6.915, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 1.45865963234688e-06, | |
| "loss": 0.0138, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "eval_loss": 0.017776617780327797, | |
| "eval_runtime": 118.343, | |
| "eval_samples_per_second": 55.28, | |
| "eval_steps_per_second": 6.912, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 1.2525760448437888e-06, | |
| "loss": 0.014, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "eval_loss": 0.017718419432640076, | |
| "eval_runtime": 118.1253, | |
| "eval_samples_per_second": 55.382, | |
| "eval_steps_per_second": 6.925, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 1.0464924573406975e-06, | |
| "loss": 0.0134, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "eval_loss": 0.01749224029481411, | |
| "eval_runtime": 118.0122, | |
| "eval_samples_per_second": 55.435, | |
| "eval_steps_per_second": 6.931, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 8.404088698376061e-07, | |
| "loss": 0.0137, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "eval_loss": 0.01760455034673214, | |
| "eval_runtime": 118.2868, | |
| "eval_samples_per_second": 55.306, | |
| "eval_steps_per_second": 6.915, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 6.343252823345148e-07, | |
| "loss": 0.0131, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "eval_loss": 0.0176746416836977, | |
| "eval_runtime": 118.1089, | |
| "eval_samples_per_second": 55.39, | |
| "eval_steps_per_second": 6.926, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 4.2824169483142364e-07, | |
| "loss": 0.0137, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_loss": 0.01762452907860279, | |
| "eval_runtime": 118.4177, | |
| "eval_samples_per_second": 55.245, | |
| "eval_steps_per_second": 6.908, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 2.221581073283324e-07, | |
| "loss": 0.0135, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "eval_loss": 0.01761581189930439, | |
| "eval_runtime": 117.528, | |
| "eval_samples_per_second": 55.663, | |
| "eval_steps_per_second": 6.96, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.6074519825241118e-08, | |
| "loss": 0.014, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.017628999426960945, | |
| "eval_runtime": 118.0406, | |
| "eval_samples_per_second": 55.422, | |
| "eval_steps_per_second": 6.93, | |
| "step": 25500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 25539, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 2.428545951977472e+16, | |
| "train_batch_size": 18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |