| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 62, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03225806451612903, | |
| "grad_norm": 31.199046559439136, | |
| "learning_rate": 9.993582535855265e-06, | |
| "loss": 1.6497, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 17.46234723437062, | |
| "learning_rate": 9.974346616959476e-06, | |
| "loss": 1.1688, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0967741935483871, | |
| "grad_norm": 17.2132760171181, | |
| "learning_rate": 9.942341621640558e-06, | |
| "loss": 1.2213, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 8.814975880215384, | |
| "learning_rate": 9.897649706262474e-06, | |
| "loss": 0.852, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 9.032603485694555, | |
| "learning_rate": 9.840385594331022e-06, | |
| "loss": 0.8313, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 5.078284135758567, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6852, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.22580645161290322, | |
| "grad_norm": 3.427300813881487, | |
| "learning_rate": 9.688760660735403e-06, | |
| "loss": 0.5022, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 4.2152309516284845, | |
| "learning_rate": 9.594789058101154e-06, | |
| "loss": 0.5572, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2903225806451613, | |
| "grad_norm": 3.085786333052321, | |
| "learning_rate": 9.48902269785371e-06, | |
| "loss": 0.4309, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 3.125642693243869, | |
| "learning_rate": 9.371733080722911e-06, | |
| "loss": 0.5173, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3548387096774194, | |
| "grad_norm": 3.277088815175521, | |
| "learning_rate": 9.243221287473755e-06, | |
| "loss": 0.4889, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 2.357479595943573, | |
| "learning_rate": 9.103817206036383e-06, | |
| "loss": 0.3478, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.41935483870967744, | |
| "grad_norm": 2.0539004083223156, | |
| "learning_rate": 8.953878684688492e-06, | |
| "loss": 0.3496, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 2.52448299705888, | |
| "learning_rate": 8.793790613463956e-06, | |
| "loss": 0.398, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 2.355140249988772, | |
| "learning_rate": 8.6239639361456e-06, | |
| "loss": 0.3568, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 2.4469993078971326, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.3884, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5483870967741935, | |
| "grad_norm": 2.52296370250433, | |
| "learning_rate": 8.256862413611113e-06, | |
| "loss": 0.3657, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 3.3355903081537135, | |
| "learning_rate": 8.060529912738316e-06, | |
| "loss": 0.5187, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6129032258064516, | |
| "grad_norm": 2.484683322570902, | |
| "learning_rate": 7.856341075473963e-06, | |
| "loss": 0.4143, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 1.7487341996348138, | |
| "learning_rate": 7.644820051634813e-06, | |
| "loss": 0.2972, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6774193548387096, | |
| "grad_norm": 2.0583109478127204, | |
| "learning_rate": 7.4265098126554065e-06, | |
| "loss": 0.3197, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 2.0344563885911575, | |
| "learning_rate": 7.201970757788172e-06, | |
| "loss": 0.3395, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.7419354838709677, | |
| "grad_norm": 2.058990191861192, | |
| "learning_rate": 6.971779275566593e-06, | |
| "loss": 0.3394, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 1.9074433438418341, | |
| "learning_rate": 6.736526264224101e-06, | |
| "loss": 0.3442, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 2.0162577599888323, | |
| "learning_rate": 6.496815614866792e-06, | |
| "loss": 0.2795, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 2.4742505970882163, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.3968, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8709677419354839, | |
| "grad_norm": 2.207808258328297, | |
| "learning_rate": 6.006492600443301e-06, | |
| "loss": 0.3594, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 1.7696328864440694, | |
| "learning_rate": 5.757138887522884e-06, | |
| "loss": 0.3114, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.9354838709677419, | |
| "grad_norm": 2.0275212616065645, | |
| "learning_rate": 5.505841609937162e-06, | |
| "loss": 0.3478, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 1.6298425948210777, | |
| "learning_rate": 5.253245844193564e-06, | |
| "loss": 0.2626, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.5467460791719727, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2845, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.032258064516129, | |
| "grad_norm": 1.5586457967216936, | |
| "learning_rate": 4.746754155806437e-06, | |
| "loss": 0.2419, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.064516129032258, | |
| "grad_norm": 1.888001852915147, | |
| "learning_rate": 4.49415839006284e-06, | |
| "loss": 0.2208, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.096774193548387, | |
| "grad_norm": 1.4011698191222255, | |
| "learning_rate": 4.2428611124771184e-06, | |
| "loss": 0.2144, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.129032258064516, | |
| "grad_norm": 1.7487392529140704, | |
| "learning_rate": 3.993507399556699e-06, | |
| "loss": 0.2356, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1612903225806452, | |
| "grad_norm": 2.111614480418175, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.2874, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1935483870967742, | |
| "grad_norm": 1.317740002801753, | |
| "learning_rate": 3.5031843851332105e-06, | |
| "loss": 0.1728, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.2258064516129032, | |
| "grad_norm": 1.7399051773302943, | |
| "learning_rate": 3.2634737357758994e-06, | |
| "loss": 0.2604, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.2580645161290323, | |
| "grad_norm": 1.8124616181419573, | |
| "learning_rate": 3.0282207244334084e-06, | |
| "loss": 0.2461, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 2.2245741657359934, | |
| "learning_rate": 2.7980292422118282e-06, | |
| "loss": 0.2703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.3225806451612903, | |
| "grad_norm": 1.6789960223565612, | |
| "learning_rate": 2.573490187344596e-06, | |
| "loss": 0.2038, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.3548387096774195, | |
| "grad_norm": 1.79239256078375, | |
| "learning_rate": 2.3551799483651894e-06, | |
| "loss": 0.2239, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3870967741935485, | |
| "grad_norm": 3.076290808429015, | |
| "learning_rate": 2.1436589245260375e-06, | |
| "loss": 0.2957, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.4193548387096775, | |
| "grad_norm": 1.5602198674820165, | |
| "learning_rate": 1.9394700872616856e-06, | |
| "loss": 0.2107, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.4516129032258065, | |
| "grad_norm": 1.577639906284918, | |
| "learning_rate": 1.74313758638889e-06, | |
| "loss": 0.2238, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4838709677419355, | |
| "grad_norm": 1.5054440694141868, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.2035, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.5161290322580645, | |
| "grad_norm": 1.92724369253792, | |
| "learning_rate": 1.3760360638544012e-06, | |
| "loss": 0.256, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.5483870967741935, | |
| "grad_norm": 1.52533329195425, | |
| "learning_rate": 1.2062093865360458e-06, | |
| "loss": 0.2078, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.5806451612903225, | |
| "grad_norm": 2.3222181174482044, | |
| "learning_rate": 1.046121315311508e-06, | |
| "loss": 0.2611, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.6129032258064515, | |
| "grad_norm": 1.4315191687571915, | |
| "learning_rate": 8.961827939636198e-07, | |
| "loss": 0.1965, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.6451612903225805, | |
| "grad_norm": 1.5414636545885392, | |
| "learning_rate": 7.567787125262449e-07, | |
| "loss": 0.2, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.6774193548387095, | |
| "grad_norm": 1.4835478438790177, | |
| "learning_rate": 6.282669192770896e-07, | |
| "loss": 0.2509, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.7096774193548387, | |
| "grad_norm": 1.4946263792553225, | |
| "learning_rate": 5.109773021462921e-07, | |
| "loss": 0.2085, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.7419354838709677, | |
| "grad_norm": 1.7678787704558188, | |
| "learning_rate": 4.05210941898847e-07, | |
| "loss": 0.2467, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.7741935483870968, | |
| "grad_norm": 1.4867440873899518, | |
| "learning_rate": 3.112393392645985e-07, | |
| "loss": 0.2363, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.8064516129032258, | |
| "grad_norm": 1.4803775636293333, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.1855, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.838709677419355, | |
| "grad_norm": 1.4490266945389694, | |
| "learning_rate": 1.5961440566897913e-07, | |
| "loss": 0.1821, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.870967741935484, | |
| "grad_norm": 1.7145630167827153, | |
| "learning_rate": 1.0235029373752758e-07, | |
| "loss": 0.2157, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.903225806451613, | |
| "grad_norm": 1.5925676389651224, | |
| "learning_rate": 5.7658378359443104e-08, | |
| "loss": 0.2158, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "grad_norm": 1.4959223286590795, | |
| "learning_rate": 2.5653383040524228e-08, | |
| "loss": 0.2202, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.967741935483871, | |
| "grad_norm": 2.013681318495715, | |
| "learning_rate": 6.417464144736208e-09, | |
| "loss": 0.2533, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.4127405994063447, | |
| "learning_rate": 0.0, | |
| "loss": 0.2159, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 62, | |
| "total_flos": 2138633404416.0, | |
| "train_loss": 0.370798627936071, | |
| "train_runtime": 218.5623, | |
| "train_samples_per_second": 2.251, | |
| "train_steps_per_second": 0.284 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 62, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 70000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2138633404416.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |