| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 39, | |
| "global_step": 384, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0026041666666666665, | |
| "grad_norm": 0.3952034496250024, | |
| "learning_rate": 5e-06, | |
| "loss": 2.4903, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026041666666666665, | |
| "eval_loss": 2.554558753967285, | |
| "eval_runtime": 65.1161, | |
| "eval_samples_per_second": 1.229, | |
| "eval_steps_per_second": 0.154, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005208333333333333, | |
| "grad_norm": 0.3885918424668888, | |
| "learning_rate": 1e-05, | |
| "loss": 2.4825, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0078125, | |
| "grad_norm": 0.3847099050493279, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.5035, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010416666666666666, | |
| "grad_norm": 0.4235545463515579, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5577, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013020833333333334, | |
| "grad_norm": 0.3801745772810421, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.5297, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015625, | |
| "grad_norm": 0.3512277500716236, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4777, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.018229166666666668, | |
| "grad_norm": 0.28888015786994375, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.4556, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 0.25350975750236143, | |
| "learning_rate": 4e-05, | |
| "loss": 2.5022, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0234375, | |
| "grad_norm": 0.22691977505165736, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.4143, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.026041666666666668, | |
| "grad_norm": 0.20864896327745933, | |
| "learning_rate": 5e-05, | |
| "loss": 2.4545, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.028645833333333332, | |
| "grad_norm": 0.20008113121425344, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.4939, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 0.24785418789288546, | |
| "learning_rate": 6e-05, | |
| "loss": 2.3184, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.033854166666666664, | |
| "grad_norm": 0.2153715911656377, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.4738, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.036458333333333336, | |
| "grad_norm": 0.22249287774026938, | |
| "learning_rate": 7e-05, | |
| "loss": 2.4814, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0390625, | |
| "grad_norm": 0.2076426443768558, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.3529, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.041666666666666664, | |
| "grad_norm": 0.2246776060857838, | |
| "learning_rate": 8e-05, | |
| "loss": 2.2371, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.044270833333333336, | |
| "grad_norm": 0.18383862266336323, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.4787, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "grad_norm": 0.20631688287583885, | |
| "learning_rate": 9e-05, | |
| "loss": 2.6178, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.049479166666666664, | |
| "grad_norm": 0.18016643348558348, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.4891, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.052083333333333336, | |
| "grad_norm": 0.14534618682097772, | |
| "learning_rate": 0.0001, | |
| "loss": 2.4522, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0546875, | |
| "grad_norm": 0.15428537805246717, | |
| "learning_rate": 9.999832398924833e-05, | |
| "loss": 2.4111, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.057291666666666664, | |
| "grad_norm": 0.1631103718894274, | |
| "learning_rate": 9.999329608183822e-05, | |
| "loss": 2.387, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.059895833333333336, | |
| "grad_norm": 0.1552602669275131, | |
| "learning_rate": 9.998491665229539e-05, | |
| "loss": 2.3648, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.1704183704935776, | |
| "learning_rate": 9.997318632479817e-05, | |
| "loss": 2.4997, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06510416666666667, | |
| "grad_norm": 0.1585395745370806, | |
| "learning_rate": 9.995810597313128e-05, | |
| "loss": 2.5326, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06770833333333333, | |
| "grad_norm": 0.1344528253347993, | |
| "learning_rate": 9.993967672062052e-05, | |
| "loss": 2.5174, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0703125, | |
| "grad_norm": 0.14705487245235002, | |
| "learning_rate": 9.991789994004929e-05, | |
| "loss": 2.3446, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07291666666666667, | |
| "grad_norm": 0.1461054029585455, | |
| "learning_rate": 9.989277725355615e-05, | |
| "loss": 2.4945, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07552083333333333, | |
| "grad_norm": 0.14508526627249885, | |
| "learning_rate": 9.986431053251411e-05, | |
| "loss": 2.3801, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.078125, | |
| "grad_norm": 0.14473824067661442, | |
| "learning_rate": 9.983250189739119e-05, | |
| "loss": 2.4877, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08072916666666667, | |
| "grad_norm": 0.18945595196350265, | |
| "learning_rate": 9.97973537175925e-05, | |
| "loss": 2.5833, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 0.1541697712272958, | |
| "learning_rate": 9.975886861128368e-05, | |
| "loss": 2.4619, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0859375, | |
| "grad_norm": 0.16052736172586615, | |
| "learning_rate": 9.971704944519594e-05, | |
| "loss": 2.5037, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08854166666666667, | |
| "grad_norm": 0.15543243937875972, | |
| "learning_rate": 9.967189933441243e-05, | |
| "loss": 2.4374, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09114583333333333, | |
| "grad_norm": 0.16705335802827087, | |
| "learning_rate": 9.962342164213639e-05, | |
| "loss": 2.5955, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 0.14789348505066385, | |
| "learning_rate": 9.957161997944034e-05, | |
| "loss": 2.4322, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09635416666666667, | |
| "grad_norm": 0.15495422007897772, | |
| "learning_rate": 9.95164982049974e-05, | |
| "loss": 2.4307, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09895833333333333, | |
| "grad_norm": 0.15049524925558916, | |
| "learning_rate": 9.94580604247936e-05, | |
| "loss": 2.3423, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1015625, | |
| "grad_norm": 0.1444251980172912, | |
| "learning_rate": 9.939631099182219e-05, | |
| "loss": 2.2882, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1015625, | |
| "eval_loss": 2.4301750659942627, | |
| "eval_runtime": 65.3323, | |
| "eval_samples_per_second": 1.225, | |
| "eval_steps_per_second": 0.153, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 0.1475367612200888, | |
| "learning_rate": 9.933125450575932e-05, | |
| "loss": 2.4509, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10677083333333333, | |
| "grad_norm": 0.16356380426266176, | |
| "learning_rate": 9.926289581262147e-05, | |
| "loss": 2.5307, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.109375, | |
| "grad_norm": 0.166302003601292, | |
| "learning_rate": 9.919124000440438e-05, | |
| "loss": 2.526, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11197916666666667, | |
| "grad_norm": 0.14645402434215465, | |
| "learning_rate": 9.91162924187038e-05, | |
| "loss": 2.2958, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11458333333333333, | |
| "grad_norm": 0.2119539720051403, | |
| "learning_rate": 9.903805863831799e-05, | |
| "loss": 2.3299, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1171875, | |
| "grad_norm": 0.151653634781893, | |
| "learning_rate": 9.895654449083166e-05, | |
| "loss": 2.4133, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11979166666666667, | |
| "grad_norm": 0.1547044553169797, | |
| "learning_rate": 9.887175604818206e-05, | |
| "loss": 2.5952, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.12239583333333333, | |
| "grad_norm": 0.16664262013704687, | |
| "learning_rate": 9.87836996262066e-05, | |
| "loss": 2.4714, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.15551192268579045, | |
| "learning_rate": 9.869238178417235e-05, | |
| "loss": 2.3432, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12760416666666666, | |
| "grad_norm": 0.14968525834314664, | |
| "learning_rate": 9.859780932428756e-05, | |
| "loss": 2.4037, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13020833333333334, | |
| "grad_norm": 0.1685655914159011, | |
| "learning_rate": 9.849998929119488e-05, | |
| "loss": 2.4032, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1328125, | |
| "grad_norm": 0.13644233785328275, | |
| "learning_rate": 9.839892897144663e-05, | |
| "loss": 2.4095, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13541666666666666, | |
| "grad_norm": 0.14364815720124613, | |
| "learning_rate": 9.829463589296203e-05, | |
| "loss": 2.4398, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13802083333333334, | |
| "grad_norm": 0.14915929369602085, | |
| "learning_rate": 9.818711782446645e-05, | |
| "loss": 2.4816, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.140625, | |
| "grad_norm": 0.14935461072359194, | |
| "learning_rate": 9.80763827749127e-05, | |
| "loss": 2.3426, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14322916666666666, | |
| "grad_norm": 0.15190176296605723, | |
| "learning_rate": 9.796243899288456e-05, | |
| "loss": 2.4705, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14583333333333334, | |
| "grad_norm": 0.14303795896023264, | |
| "learning_rate": 9.784529496598214e-05, | |
| "loss": 2.5108, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1484375, | |
| "grad_norm": 0.1556316812368993, | |
| "learning_rate": 9.772495942018985e-05, | |
| "loss": 2.3112, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.15104166666666666, | |
| "grad_norm": 0.1483632196038671, | |
| "learning_rate": 9.760144131922628e-05, | |
| "loss": 2.2325, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15364583333333334, | |
| "grad_norm": 0.12410345584803052, | |
| "learning_rate": 9.747474986387654e-05, | |
| "loss": 2.3749, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 0.16335694957610425, | |
| "learning_rate": 9.734489449130695e-05, | |
| "loss": 2.4645, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15885416666666666, | |
| "grad_norm": 0.1479375196060614, | |
| "learning_rate": 9.721188487436195e-05, | |
| "loss": 2.3834, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16145833333333334, | |
| "grad_norm": 0.13750323689269844, | |
| "learning_rate": 9.707573092084368e-05, | |
| "loss": 2.3592, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1640625, | |
| "grad_norm": 0.16027916915470333, | |
| "learning_rate": 9.693644277277391e-05, | |
| "loss": 2.4954, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.14229365410212697, | |
| "learning_rate": 9.679403080563861e-05, | |
| "loss": 2.4404, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.16927083333333334, | |
| "grad_norm": 0.153686499622388, | |
| "learning_rate": 9.6648505627615e-05, | |
| "loss": 2.2282, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.171875, | |
| "grad_norm": 0.16590644992700104, | |
| "learning_rate": 9.649987807878148e-05, | |
| "loss": 2.5126, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17447916666666666, | |
| "grad_norm": 0.15502972088699477, | |
| "learning_rate": 9.634815923030997e-05, | |
| "loss": 2.3749, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.17708333333333334, | |
| "grad_norm": 0.14756465898580834, | |
| "learning_rate": 9.61933603836415e-05, | |
| "loss": 2.3476, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1796875, | |
| "grad_norm": 0.14919525754908758, | |
| "learning_rate": 9.603549306964407e-05, | |
| "loss": 2.3509, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.18229166666666666, | |
| "grad_norm": 0.13861494211266717, | |
| "learning_rate": 9.587456904775393e-05, | |
| "loss": 2.2449, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18489583333333334, | |
| "grad_norm": 0.14821518394829722, | |
| "learning_rate": 9.57106003050996e-05, | |
| "loss": 2.3813, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.16324527196573113, | |
| "learning_rate": 9.554359905560886e-05, | |
| "loss": 2.477, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.19010416666666666, | |
| "grad_norm": 0.14892748528676938, | |
| "learning_rate": 9.537357773909906e-05, | |
| "loss": 2.264, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.19270833333333334, | |
| "grad_norm": 0.1284546434475265, | |
| "learning_rate": 9.520054902035035e-05, | |
| "loss": 2.4267, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1953125, | |
| "grad_norm": 0.14293551215488642, | |
| "learning_rate": 9.502452578816244e-05, | |
| "loss": 2.5131, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19791666666666666, | |
| "grad_norm": 0.14775108859068506, | |
| "learning_rate": 9.484552115439445e-05, | |
| "loss": 2.4076, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.20052083333333334, | |
| "grad_norm": 0.15478800535673032, | |
| "learning_rate": 9.466354845298817e-05, | |
| "loss": 2.5762, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.203125, | |
| "grad_norm": 0.16835921991226063, | |
| "learning_rate": 9.44786212389749e-05, | |
| "loss": 2.3752, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.203125, | |
| "eval_loss": 2.4171054363250732, | |
| "eval_runtime": 65.4109, | |
| "eval_samples_per_second": 1.223, | |
| "eval_steps_per_second": 0.153, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.20572916666666666, | |
| "grad_norm": 0.14467889640535875, | |
| "learning_rate": 9.42907532874657e-05, | |
| "loss": 2.4769, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 0.13503613492125413, | |
| "learning_rate": 9.40999585926253e-05, | |
| "loss": 2.4346, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2109375, | |
| "grad_norm": 0.1517728973900508, | |
| "learning_rate": 9.390625136662972e-05, | |
| "loss": 2.261, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.21354166666666666, | |
| "grad_norm": 0.1466563383653839, | |
| "learning_rate": 9.370964603860753e-05, | |
| "loss": 2.5261, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.21614583333333334, | |
| "grad_norm": 0.14618631155344838, | |
| "learning_rate": 9.351015725356514e-05, | |
| "loss": 2.4269, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.21875, | |
| "grad_norm": 0.15601390292939937, | |
| "learning_rate": 9.33077998712958e-05, | |
| "loss": 2.3773, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.22135416666666666, | |
| "grad_norm": 0.14415725413526403, | |
| "learning_rate": 9.310258896527278e-05, | |
| "loss": 2.3473, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.22395833333333334, | |
| "grad_norm": 0.13470412653239092, | |
| "learning_rate": 9.289453982152653e-05, | |
| "loss": 2.3472, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2265625, | |
| "grad_norm": 0.15570776289510233, | |
| "learning_rate": 9.2683667937506e-05, | |
| "loss": 2.4414, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.22916666666666666, | |
| "grad_norm": 0.14612626399487072, | |
| "learning_rate": 9.246998902092428e-05, | |
| "loss": 2.4655, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.23177083333333334, | |
| "grad_norm": 0.18267625807633459, | |
| "learning_rate": 9.22535189885886e-05, | |
| "loss": 2.2771, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.234375, | |
| "grad_norm": 0.16065010725957032, | |
| "learning_rate": 9.203427396521454e-05, | |
| "loss": 2.4194, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23697916666666666, | |
| "grad_norm": 0.18321820032984756, | |
| "learning_rate": 9.181227028222508e-05, | |
| "loss": 2.4456, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.23958333333333334, | |
| "grad_norm": 0.13929464732035163, | |
| "learning_rate": 9.158752447653397e-05, | |
| "loss": 2.3989, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2421875, | |
| "grad_norm": 0.15585898969003706, | |
| "learning_rate": 9.136005328931395e-05, | |
| "loss": 2.4777, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.24479166666666666, | |
| "grad_norm": 0.14793073723102618, | |
| "learning_rate": 9.112987366474972e-05, | |
| "loss": 2.379, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.24739583333333334, | |
| "grad_norm": 0.14075868568691305, | |
| "learning_rate": 9.089700274877574e-05, | |
| "loss": 2.3281, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.15327339815586916, | |
| "learning_rate": 9.066145788779908e-05, | |
| "loss": 2.4517, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2526041666666667, | |
| "grad_norm": 0.15505433166839452, | |
| "learning_rate": 9.042325662740726e-05, | |
| "loss": 2.2919, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.2552083333333333, | |
| "grad_norm": 0.1468856057426848, | |
| "learning_rate": 9.018241671106134e-05, | |
| "loss": 2.3889, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2578125, | |
| "grad_norm": 0.1478488710337695, | |
| "learning_rate": 8.993895607877418e-05, | |
| "loss": 2.3595, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.2604166666666667, | |
| "grad_norm": 0.175359520279898, | |
| "learning_rate": 8.969289286577408e-05, | |
| "loss": 2.553, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2630208333333333, | |
| "grad_norm": 0.14503182442700818, | |
| "learning_rate": 8.9444245401154e-05, | |
| "loss": 2.3573, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.265625, | |
| "grad_norm": 0.14177870868430487, | |
| "learning_rate": 8.919303220650606e-05, | |
| "loss": 2.3274, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2682291666666667, | |
| "grad_norm": 0.1779007031093982, | |
| "learning_rate": 8.893927199454207e-05, | |
| "loss": 2.2008, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2708333333333333, | |
| "grad_norm": 0.14578869126284813, | |
| "learning_rate": 8.868298366769954e-05, | |
| "loss": 2.444, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2734375, | |
| "grad_norm": 0.14046914929392473, | |
| "learning_rate": 8.842418631673365e-05, | |
| "loss": 2.2844, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2760416666666667, | |
| "grad_norm": 0.13720708755401478, | |
| "learning_rate": 8.816289921929516e-05, | |
| "loss": 2.5047, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2786458333333333, | |
| "grad_norm": 0.13901967393067377, | |
| "learning_rate": 8.789914183849449e-05, | |
| "loss": 2.4602, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 0.14165507438668587, | |
| "learning_rate": 8.763293382145195e-05, | |
| "loss": 2.3135, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2838541666666667, | |
| "grad_norm": 0.1588574639328753, | |
| "learning_rate": 8.73642949978341e-05, | |
| "loss": 2.3731, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2864583333333333, | |
| "grad_norm": 0.1905241886636867, | |
| "learning_rate": 8.709324537837684e-05, | |
| "loss": 2.4152, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2890625, | |
| "grad_norm": 0.16919809232424363, | |
| "learning_rate": 8.681980515339464e-05, | |
| "loss": 2.2948, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2916666666666667, | |
| "grad_norm": 0.15114015395920236, | |
| "learning_rate": 8.654399469127673e-05, | |
| "loss": 2.4309, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2942708333333333, | |
| "grad_norm": 0.13890547261939354, | |
| "learning_rate": 8.626583453696976e-05, | |
| "loss": 2.4866, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.296875, | |
| "grad_norm": 0.1490334985258922, | |
| "learning_rate": 8.598534541044747e-05, | |
| "loss": 2.4748, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2994791666666667, | |
| "grad_norm": 0.15113023404371403, | |
| "learning_rate": 8.570254820516728e-05, | |
| "loss": 2.3583, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3020833333333333, | |
| "grad_norm": 0.14358757342098682, | |
| "learning_rate": 8.541746398651395e-05, | |
| "loss": 2.4005, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3046875, | |
| "grad_norm": 0.1494340804949735, | |
| "learning_rate": 8.513011399023036e-05, | |
| "loss": 2.3249, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3046875, | |
| "eval_loss": 2.4119415283203125, | |
| "eval_runtime": 65.4329, | |
| "eval_samples_per_second": 1.223, | |
| "eval_steps_per_second": 0.153, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3072916666666667, | |
| "grad_norm": 0.14857253164474957, | |
| "learning_rate": 8.484051962083579e-05, | |
| "loss": 2.3431, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3098958333333333, | |
| "grad_norm": 0.14191988741229736, | |
| "learning_rate": 8.454870245003141e-05, | |
| "loss": 2.3861, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.1516024307266096, | |
| "learning_rate": 8.425468421509349e-05, | |
| "loss": 2.411, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3151041666666667, | |
| "grad_norm": 0.15002982823803926, | |
| "learning_rate": 8.395848681725416e-05, | |
| "loss": 2.49, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3177083333333333, | |
| "grad_norm": 0.14904411125984457, | |
| "learning_rate": 8.366013232007002e-05, | |
| "loss": 2.2998, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3203125, | |
| "grad_norm": 0.15513959182528284, | |
| "learning_rate": 8.335964294777862e-05, | |
| "loss": 2.4748, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3229166666666667, | |
| "grad_norm": 0.16028597088352084, | |
| "learning_rate": 8.305704108364301e-05, | |
| "loss": 2.5415, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3255208333333333, | |
| "grad_norm": 0.1382845005964039, | |
| "learning_rate": 8.275234926828446e-05, | |
| "loss": 2.3863, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.328125, | |
| "grad_norm": 0.13146327268638525, | |
| "learning_rate": 8.244559019800328e-05, | |
| "loss": 2.2718, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3307291666666667, | |
| "grad_norm": 0.14918540441469405, | |
| "learning_rate": 8.213678672308841e-05, | |
| "loss": 2.29, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.14572964949235084, | |
| "learning_rate": 8.182596184611514e-05, | |
| "loss": 2.3865, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3359375, | |
| "grad_norm": 0.1467709610422986, | |
| "learning_rate": 8.151313872023172e-05, | |
| "loss": 2.3566, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3385416666666667, | |
| "grad_norm": 0.1696428030171741, | |
| "learning_rate": 8.119834064743469e-05, | |
| "loss": 2.4145, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3411458333333333, | |
| "grad_norm": 0.1567212274267596, | |
| "learning_rate": 8.088159107683314e-05, | |
| "loss": 2.3996, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.34375, | |
| "grad_norm": 0.16059171688564705, | |
| "learning_rate": 8.056291360290201e-05, | |
| "loss": 2.4796, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3463541666666667, | |
| "grad_norm": 0.17528470844982455, | |
| "learning_rate": 8.024233196372453e-05, | |
| "loss": 2.3711, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3489583333333333, | |
| "grad_norm": 0.15733889602776618, | |
| "learning_rate": 7.9919870039224e-05, | |
| "loss": 2.2677, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3515625, | |
| "grad_norm": 0.14873870088112393, | |
| "learning_rate": 7.959555184938495e-05, | |
| "loss": 2.4515, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3541666666666667, | |
| "grad_norm": 0.15417965735327782, | |
| "learning_rate": 7.926940155246397e-05, | |
| "loss": 2.4285, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3567708333333333, | |
| "grad_norm": 0.1431617281765595, | |
| "learning_rate": 7.894144344319014e-05, | |
| "loss": 2.2096, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.359375, | |
| "grad_norm": 0.1582812354896834, | |
| "learning_rate": 7.861170195095537e-05, | |
| "loss": 2.5397, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3619791666666667, | |
| "grad_norm": 0.17372786439002758, | |
| "learning_rate": 7.828020163799455e-05, | |
| "loss": 2.4293, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3645833333333333, | |
| "grad_norm": 0.12985922414383083, | |
| "learning_rate": 7.794696719755612e-05, | |
| "loss": 2.3064, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3671875, | |
| "grad_norm": 0.14444453543114258, | |
| "learning_rate": 7.761202345206249e-05, | |
| "loss": 2.4924, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3697916666666667, | |
| "grad_norm": 0.14361378320872797, | |
| "learning_rate": 7.727539535126118e-05, | |
| "loss": 2.485, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3723958333333333, | |
| "grad_norm": 0.14506765397834698, | |
| "learning_rate": 7.69371079703662e-05, | |
| "loss": 2.4081, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.15082242594669004, | |
| "learning_rate": 7.65971865081904e-05, | |
| "loss": 2.4042, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3776041666666667, | |
| "grad_norm": 0.14811116555183387, | |
| "learning_rate": 7.625565628526818e-05, | |
| "loss": 2.5335, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3802083333333333, | |
| "grad_norm": 0.14726265807121236, | |
| "learning_rate": 7.591254274196959e-05, | |
| "loss": 2.2424, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3828125, | |
| "grad_norm": 0.1582839530592079, | |
| "learning_rate": 7.556787143660521e-05, | |
| "loss": 2.3499, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3854166666666667, | |
| "grad_norm": 0.15670436913091046, | |
| "learning_rate": 7.522166804352226e-05, | |
| "loss": 2.4295, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3880208333333333, | |
| "grad_norm": 0.14730612216140554, | |
| "learning_rate": 7.487395835119231e-05, | |
| "loss": 2.3101, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.390625, | |
| "grad_norm": 0.15120580103738476, | |
| "learning_rate": 7.452476826029011e-05, | |
| "loss": 2.4888, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3932291666666667, | |
| "grad_norm": 0.15081791485599214, | |
| "learning_rate": 7.417412378176446e-05, | |
| "loss": 2.3946, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3958333333333333, | |
| "grad_norm": 0.13244742107918075, | |
| "learning_rate": 7.382205103490043e-05, | |
| "loss": 2.2704, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3984375, | |
| "grad_norm": 0.14571632653572605, | |
| "learning_rate": 7.346857624537407e-05, | |
| "loss": 2.4644, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4010416666666667, | |
| "grad_norm": 0.15252862923031207, | |
| "learning_rate": 7.311372574329854e-05, | |
| "loss": 2.554, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4036458333333333, | |
| "grad_norm": 0.14927980854491554, | |
| "learning_rate": 7.275752596126308e-05, | |
| "loss": 2.3804, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "grad_norm": 0.1406590744757482, | |
| "learning_rate": 7.240000343236385e-05, | |
| "loss": 2.2504, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "eval_loss": 2.408146619796753, | |
| "eval_runtime": 65.4856, | |
| "eval_samples_per_second": 1.222, | |
| "eval_steps_per_second": 0.153, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4088541666666667, | |
| "grad_norm": 0.13753178031418029, | |
| "learning_rate": 7.204118478822766e-05, | |
| "loss": 2.3063, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.4114583333333333, | |
| "grad_norm": 0.13719355865812055, | |
| "learning_rate": 7.168109675702806e-05, | |
| "loss": 2.1826, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4140625, | |
| "grad_norm": 0.14916916897375052, | |
| "learning_rate": 7.131976616149445e-05, | |
| "loss": 2.3635, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 0.1544520288957668, | |
| "learning_rate": 7.095721991691411e-05, | |
| "loss": 2.6067, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4192708333333333, | |
| "grad_norm": 0.1466811081471814, | |
| "learning_rate": 7.05934850291272e-05, | |
| "loss": 2.2636, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.421875, | |
| "grad_norm": 0.14245354657490636, | |
| "learning_rate": 7.022858859251517e-05, | |
| "loss": 2.3278, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4244791666666667, | |
| "grad_norm": 0.14310350882179557, | |
| "learning_rate": 6.986255778798253e-05, | |
| "loss": 2.2951, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4270833333333333, | |
| "grad_norm": 0.15028525220062852, | |
| "learning_rate": 6.949541988093208e-05, | |
| "loss": 2.3647, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4296875, | |
| "grad_norm": 0.13536590220383313, | |
| "learning_rate": 6.912720221923405e-05, | |
| "loss": 2.3635, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4322916666666667, | |
| "grad_norm": 0.14965390120450028, | |
| "learning_rate": 6.875793223118888e-05, | |
| "loss": 2.3191, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4348958333333333, | |
| "grad_norm": 0.15696846625410496, | |
| "learning_rate": 6.838763742348415e-05, | |
| "loss": 2.4342, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.16266249005182767, | |
| "learning_rate": 6.801634537914555e-05, | |
| "loss": 2.4487, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4401041666666667, | |
| "grad_norm": 0.15099304315918932, | |
| "learning_rate": 6.764408375548237e-05, | |
| "loss": 2.4774, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4427083333333333, | |
| "grad_norm": 0.1437502300792439, | |
| "learning_rate": 6.727088028202723e-05, | |
| "loss": 2.4369, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4453125, | |
| "grad_norm": 0.13339677555582233, | |
| "learning_rate": 6.68967627584705e-05, | |
| "loss": 2.3039, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4479166666666667, | |
| "grad_norm": 0.1416720700803862, | |
| "learning_rate": 6.652175905258963e-05, | |
| "loss": 2.2844, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4505208333333333, | |
| "grad_norm": 0.14502909820943563, | |
| "learning_rate": 6.614589709817317e-05, | |
| "loss": 2.4192, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.453125, | |
| "grad_norm": 0.14499193575249233, | |
| "learning_rate": 6.576920489294011e-05, | |
| "loss": 2.3057, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.4557291666666667, | |
| "grad_norm": 0.15038851044223103, | |
| "learning_rate": 6.539171049645426e-05, | |
| "loss": 2.4237, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4583333333333333, | |
| "grad_norm": 0.14553942382157312, | |
| "learning_rate": 6.501344202803414e-05, | |
| "loss": 2.4642, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4609375, | |
| "grad_norm": 0.13896484063095102, | |
| "learning_rate": 6.463442766465847e-05, | |
| "loss": 2.4132, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4635416666666667, | |
| "grad_norm": 0.14978646915320232, | |
| "learning_rate": 6.425469563886715e-05, | |
| "loss": 2.3692, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4661458333333333, | |
| "grad_norm": 0.13794418734075914, | |
| "learning_rate": 6.387427423665829e-05, | |
| "loss": 2.3524, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 0.1467143240226456, | |
| "learning_rate": 6.349319179538126e-05, | |
| "loss": 2.359, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4713541666666667, | |
| "grad_norm": 0.14326937356779176, | |
| "learning_rate": 6.311147670162576e-05, | |
| "loss": 2.5547, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4739583333333333, | |
| "grad_norm": 0.141220074236358, | |
| "learning_rate": 6.272915738910743e-05, | |
| "loss": 2.4299, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4765625, | |
| "grad_norm": 0.148901889828269, | |
| "learning_rate": 6.234626233654973e-05, | |
| "loss": 2.5296, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4791666666666667, | |
| "grad_norm": 0.15238094995154358, | |
| "learning_rate": 6.196282006556266e-05, | |
| "loss": 2.2642, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4817708333333333, | |
| "grad_norm": 0.14562487124583376, | |
| "learning_rate": 6.157885913851818e-05, | |
| "loss": 2.2485, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.484375, | |
| "grad_norm": 0.1557482340822006, | |
| "learning_rate": 6.119440815642258e-05, | |
| "loss": 2.5018, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4869791666666667, | |
| "grad_norm": 0.15557904288598579, | |
| "learning_rate": 6.080949575678606e-05, | |
| "loss": 2.3121, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4895833333333333, | |
| "grad_norm": 0.132624820989215, | |
| "learning_rate": 6.042415061148954e-05, | |
| "loss": 2.4456, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4921875, | |
| "grad_norm": 0.14540093060758663, | |
| "learning_rate": 6.003840142464886e-05, | |
| "loss": 2.4166, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4947916666666667, | |
| "grad_norm": 0.13698817848218522, | |
| "learning_rate": 5.9652276930476656e-05, | |
| "loss": 2.4569, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4973958333333333, | |
| "grad_norm": 0.13530133541335626, | |
| "learning_rate": 5.926580589114201e-05, | |
| "loss": 2.521, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.13985911131684764, | |
| "learning_rate": 5.8879017094627874e-05, | |
| "loss": 2.4112, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5026041666666666, | |
| "grad_norm": 0.13203633038116566, | |
| "learning_rate": 5.849193935258679e-05, | |
| "loss": 2.1839, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5052083333333334, | |
| "grad_norm": 0.16221179568567085, | |
| "learning_rate": 5.810460149819462e-05, | |
| "loss": 2.2793, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5078125, | |
| "grad_norm": 0.15402021583934355, | |
| "learning_rate": 5.771703238400288e-05, | |
| "loss": 2.3905, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5078125, | |
| "eval_loss": 2.4029593467712402, | |
| "eval_runtime": 65.2935, | |
| "eval_samples_per_second": 1.225, | |
| "eval_steps_per_second": 0.153, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5104166666666666, | |
| "grad_norm": 0.15606553454493674, | |
| "learning_rate": 5.7329260879789437e-05, | |
| "loss": 2.3709, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5130208333333334, | |
| "grad_norm": 0.1448236972696976, | |
| "learning_rate": 5.6941315870408066e-05, | |
| "loss": 2.4256, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.515625, | |
| "grad_norm": 0.15924118657730604, | |
| "learning_rate": 5.655322625363687e-05, | |
| "loss": 2.5228, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5182291666666666, | |
| "grad_norm": 0.17454487852950232, | |
| "learning_rate": 5.616502093802565e-05, | |
| "loss": 2.2945, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 0.14248471717781994, | |
| "learning_rate": 5.577672884074249e-05, | |
| "loss": 2.3514, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5234375, | |
| "grad_norm": 0.14873123082787126, | |
| "learning_rate": 5.538837888541986e-05, | |
| "loss": 2.4918, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5260416666666666, | |
| "grad_norm": 0.1600074911775423, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.3231, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5286458333333334, | |
| "grad_norm": 0.15123679319262007, | |
| "learning_rate": 5.461162111458016e-05, | |
| "loss": 2.3613, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.53125, | |
| "grad_norm": 0.13067031369476279, | |
| "learning_rate": 5.422327115925753e-05, | |
| "loss": 2.3407, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5338541666666666, | |
| "grad_norm": 0.1419685152456832, | |
| "learning_rate": 5.3834979061974376e-05, | |
| "loss": 2.4375, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5364583333333334, | |
| "grad_norm": 0.13456464948427768, | |
| "learning_rate": 5.3446773746363153e-05, | |
| "loss": 2.3542, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5390625, | |
| "grad_norm": 0.1394757337185396, | |
| "learning_rate": 5.305868412959195e-05, | |
| "loss": 2.6189, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5416666666666666, | |
| "grad_norm": 0.14661264347416875, | |
| "learning_rate": 5.2670739120210574e-05, | |
| "loss": 2.4666, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5442708333333334, | |
| "grad_norm": 0.15772002651226175, | |
| "learning_rate": 5.2282967615997125e-05, | |
| "loss": 2.2665, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.546875, | |
| "grad_norm": 0.1436207432226202, | |
| "learning_rate": 5.1895398501805383e-05, | |
| "loss": 2.4412, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5494791666666666, | |
| "grad_norm": 0.1627045485383769, | |
| "learning_rate": 5.150806064741323e-05, | |
| "loss": 2.4241, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5520833333333334, | |
| "grad_norm": 0.1608053633868765, | |
| "learning_rate": 5.112098290537213e-05, | |
| "loss": 2.2647, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5546875, | |
| "grad_norm": 0.15608204954248603, | |
| "learning_rate": 5.0734194108858e-05, | |
| "loss": 2.4768, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5572916666666666, | |
| "grad_norm": 0.1576523104437354, | |
| "learning_rate": 5.0347723069523355e-05, | |
| "loss": 2.2488, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5598958333333334, | |
| "grad_norm": 0.16267311756196431, | |
| "learning_rate": 4.9961598575351155e-05, | |
| "loss": 2.2745, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.14597656332784, | |
| "learning_rate": 4.9575849388510473e-05, | |
| "loss": 2.2226, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5651041666666666, | |
| "grad_norm": 0.12952717920836662, | |
| "learning_rate": 4.919050424321395e-05, | |
| "loss": 2.2863, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5677083333333334, | |
| "grad_norm": 0.14126095486957518, | |
| "learning_rate": 4.880559184357745e-05, | |
| "loss": 2.2773, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5703125, | |
| "grad_norm": 0.15894975244045567, | |
| "learning_rate": 4.842114086148185e-05, | |
| "loss": 2.4144, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5729166666666666, | |
| "grad_norm": 0.15187930552412898, | |
| "learning_rate": 4.803717993443734e-05, | |
| "loss": 2.2729, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5755208333333334, | |
| "grad_norm": 0.15567668429445186, | |
| "learning_rate": 4.765373766345028e-05, | |
| "loss": 2.4327, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.578125, | |
| "grad_norm": 0.1577838508745425, | |
| "learning_rate": 4.727084261089257e-05, | |
| "loss": 2.5534, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5807291666666666, | |
| "grad_norm": 0.1447851641278385, | |
| "learning_rate": 4.688852329837424e-05, | |
| "loss": 2.2928, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 0.13963925711509403, | |
| "learning_rate": 4.6506808204618754e-05, | |
| "loss": 2.3612, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5859375, | |
| "grad_norm": 0.17944687510007068, | |
| "learning_rate": 4.612572576334171e-05, | |
| "loss": 2.4195, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5885416666666666, | |
| "grad_norm": 0.15388075071309162, | |
| "learning_rate": 4.574530436113286e-05, | |
| "loss": 2.5277, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5911458333333334, | |
| "grad_norm": 0.12718481374460847, | |
| "learning_rate": 4.536557233534153e-05, | |
| "loss": 2.2991, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.59375, | |
| "grad_norm": 0.14948263293785125, | |
| "learning_rate": 4.498655797196586e-05, | |
| "loss": 2.4404, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5963541666666666, | |
| "grad_norm": 0.13608773847019048, | |
| "learning_rate": 4.460828950354577e-05, | |
| "loss": 2.2945, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5989583333333334, | |
| "grad_norm": 0.14126010771163286, | |
| "learning_rate": 4.423079510705992e-05, | |
| "loss": 2.2835, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6015625, | |
| "grad_norm": 0.15458611014177978, | |
| "learning_rate": 4.3854102901826834e-05, | |
| "loss": 2.4491, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6041666666666666, | |
| "grad_norm": 0.17078246755698387, | |
| "learning_rate": 4.3478240947410386e-05, | |
| "loss": 2.5331, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6067708333333334, | |
| "grad_norm": 0.14807435693469648, | |
| "learning_rate": 4.3103237241529506e-05, | |
| "loss": 2.4154, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.609375, | |
| "grad_norm": 0.15605308853260608, | |
| "learning_rate": 4.272911971797279e-05, | |
| "loss": 2.3354, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.609375, | |
| "eval_loss": 2.401761531829834, | |
| "eval_runtime": 65.3141, | |
| "eval_samples_per_second": 1.225, | |
| "eval_steps_per_second": 0.153, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6119791666666666, | |
| "grad_norm": 0.15343655698839206, | |
| "learning_rate": 4.235591624451763e-05, | |
| "loss": 2.5605, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6145833333333334, | |
| "grad_norm": 0.14304569592672528, | |
| "learning_rate": 4.198365462085446e-05, | |
| "loss": 2.5812, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6171875, | |
| "grad_norm": 0.15320835548074707, | |
| "learning_rate": 4.161236257651587e-05, | |
| "loss": 2.5275, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6197916666666666, | |
| "grad_norm": 0.15452569712518102, | |
| "learning_rate": 4.1242067768811134e-05, | |
| "loss": 2.4707, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6223958333333334, | |
| "grad_norm": 0.16320750887974791, | |
| "learning_rate": 4.0872797780765946e-05, | |
| "loss": 2.3996, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.14194397858802227, | |
| "learning_rate": 4.0504580119067933e-05, | |
| "loss": 2.2431, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6276041666666666, | |
| "grad_norm": 0.1437781011206105, | |
| "learning_rate": 4.01374422120175e-05, | |
| "loss": 2.3929, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6302083333333334, | |
| "grad_norm": 0.14383943780586664, | |
| "learning_rate": 3.977141140748484e-05, | |
| "loss": 2.3989, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6328125, | |
| "grad_norm": 0.16948819133813695, | |
| "learning_rate": 3.94065149708728e-05, | |
| "loss": 2.4256, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6354166666666666, | |
| "grad_norm": 0.14755837584042042, | |
| "learning_rate": 3.904278008308589e-05, | |
| "loss": 2.2711, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6380208333333334, | |
| "grad_norm": 0.15519908616035058, | |
| "learning_rate": 3.868023383850556e-05, | |
| "loss": 2.4623, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.640625, | |
| "grad_norm": 0.16021538951276384, | |
| "learning_rate": 3.831890324297197e-05, | |
| "loss": 2.3857, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6432291666666666, | |
| "grad_norm": 0.15886505001684953, | |
| "learning_rate": 3.795881521177236e-05, | |
| "loss": 2.5196, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6458333333333334, | |
| "grad_norm": 0.16005244689800305, | |
| "learning_rate": 3.7599996567636156e-05, | |
| "loss": 2.406, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6484375, | |
| "grad_norm": 0.1470078515999776, | |
| "learning_rate": 3.724247403873694e-05, | |
| "loss": 2.4975, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6510416666666666, | |
| "grad_norm": 0.1560311928142992, | |
| "learning_rate": 3.688627425670147e-05, | |
| "loss": 2.374, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6536458333333334, | |
| "grad_norm": 0.15349615074542047, | |
| "learning_rate": 3.653142375462596e-05, | |
| "loss": 2.4155, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.65625, | |
| "grad_norm": 0.20641136824203335, | |
| "learning_rate": 3.6177948965099585e-05, | |
| "loss": 2.4358, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6588541666666666, | |
| "grad_norm": 0.14326347914064022, | |
| "learning_rate": 3.582587621823558e-05, | |
| "loss": 2.4528, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6614583333333334, | |
| "grad_norm": 0.13097442902507145, | |
| "learning_rate": 3.547523173970989e-05, | |
| "loss": 2.3682, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6640625, | |
| "grad_norm": 0.14938841182330287, | |
| "learning_rate": 3.51260416488077e-05, | |
| "loss": 2.5273, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.1532472206020292, | |
| "learning_rate": 3.477833195647773e-05, | |
| "loss": 2.5301, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6692708333333334, | |
| "grad_norm": 0.12783586158700921, | |
| "learning_rate": 3.443212856339481e-05, | |
| "loss": 2.3279, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.671875, | |
| "grad_norm": 0.15547617846106007, | |
| "learning_rate": 3.408745725803042e-05, | |
| "loss": 2.4209, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6744791666666666, | |
| "grad_norm": 0.13531260542176757, | |
| "learning_rate": 3.3744343714731835e-05, | |
| "loss": 2.3595, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6770833333333334, | |
| "grad_norm": 0.1488109854224464, | |
| "learning_rate": 3.3402813491809623e-05, | |
| "loss": 2.2631, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6796875, | |
| "grad_norm": 0.1753652780376821, | |
| "learning_rate": 3.3062892029633817e-05, | |
| "loss": 2.4748, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6822916666666666, | |
| "grad_norm": 0.1593205802273226, | |
| "learning_rate": 3.272460464873884e-05, | |
| "loss": 2.4484, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6848958333333334, | |
| "grad_norm": 0.1401885264986934, | |
| "learning_rate": 3.238797654793752e-05, | |
| "loss": 2.4234, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 0.15910838717602993, | |
| "learning_rate": 3.205303280244389e-05, | |
| "loss": 2.4679, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6901041666666666, | |
| "grad_norm": 0.16179488061734165, | |
| "learning_rate": 3.1719798362005444e-05, | |
| "loss": 2.4883, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6927083333333334, | |
| "grad_norm": 0.15848452909780508, | |
| "learning_rate": 3.138829804904464e-05, | |
| "loss": 2.4583, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6953125, | |
| "grad_norm": 0.16307212504652477, | |
| "learning_rate": 3.105855655680986e-05, | |
| "loss": 2.3327, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6979166666666666, | |
| "grad_norm": 0.1467535420565889, | |
| "learning_rate": 3.073059844753604e-05, | |
| "loss": 2.4382, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.7005208333333334, | |
| "grad_norm": 0.14201414432531673, | |
| "learning_rate": 3.0404448150615063e-05, | |
| "loss": 2.3501, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.703125, | |
| "grad_norm": 0.1549923554458448, | |
| "learning_rate": 3.0080129960776017e-05, | |
| "loss": 2.396, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7057291666666666, | |
| "grad_norm": 0.1524034184779795, | |
| "learning_rate": 2.9757668036275477e-05, | |
| "loss": 2.2784, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7083333333333334, | |
| "grad_norm": 0.16256677012412982, | |
| "learning_rate": 2.9437086397097995e-05, | |
| "loss": 2.3027, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7109375, | |
| "grad_norm": 0.15633557288864075, | |
| "learning_rate": 2.9118408923166875e-05, | |
| "loss": 2.5473, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7109375, | |
| "eval_loss": 2.3996334075927734, | |
| "eval_runtime": 65.6177, | |
| "eval_samples_per_second": 1.219, | |
| "eval_steps_per_second": 0.152, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7135416666666666, | |
| "grad_norm": 0.14577207953482688, | |
| "learning_rate": 2.8801659352565335e-05, | |
| "loss": 2.3106, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7161458333333334, | |
| "grad_norm": 0.12720217228966463, | |
| "learning_rate": 2.848686127976829e-05, | |
| "loss": 2.4048, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.71875, | |
| "grad_norm": 0.1727762156703378, | |
| "learning_rate": 2.8174038153884862e-05, | |
| "loss": 2.3996, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7213541666666666, | |
| "grad_norm": 0.20872850658454511, | |
| "learning_rate": 2.786321327691158e-05, | |
| "loss": 2.5836, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7239583333333334, | |
| "grad_norm": 0.1591486153830447, | |
| "learning_rate": 2.7554409801996723e-05, | |
| "loss": 2.4478, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7265625, | |
| "grad_norm": 0.14032674083135735, | |
| "learning_rate": 2.7247650731715564e-05, | |
| "loss": 2.2852, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "grad_norm": 0.15626590715798697, | |
| "learning_rate": 2.6942958916356998e-05, | |
| "loss": 2.525, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7317708333333334, | |
| "grad_norm": 0.16623699130944486, | |
| "learning_rate": 2.66403570522214e-05, | |
| "loss": 2.3565, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.734375, | |
| "grad_norm": 0.1449533445649136, | |
| "learning_rate": 2.6339867679929997e-05, | |
| "loss": 2.2674, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7369791666666666, | |
| "grad_norm": 0.14290681116465945, | |
| "learning_rate": 2.6041513182745837e-05, | |
| "loss": 2.1482, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7395833333333334, | |
| "grad_norm": 0.13123439654719476, | |
| "learning_rate": 2.574531578490651e-05, | |
| "loss": 2.2889, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7421875, | |
| "grad_norm": 0.1510144014374668, | |
| "learning_rate": 2.54512975499686e-05, | |
| "loss": 2.3577, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7447916666666666, | |
| "grad_norm": 0.17738379736772392, | |
| "learning_rate": 2.515948037916423e-05, | |
| "loss": 2.3591, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7473958333333334, | |
| "grad_norm": 0.14040918825453863, | |
| "learning_rate": 2.4869886009769657e-05, | |
| "loss": 2.3019, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.15309282350105588, | |
| "learning_rate": 2.4582536013486054e-05, | |
| "loss": 2.3621, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.7526041666666666, | |
| "grad_norm": 0.15908796843842757, | |
| "learning_rate": 2.429745179483272e-05, | |
| "loss": 2.451, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7552083333333334, | |
| "grad_norm": 0.15934279432082493, | |
| "learning_rate": 2.4014654589552526e-05, | |
| "loss": 2.4775, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7578125, | |
| "grad_norm": 0.15682626436856428, | |
| "learning_rate": 2.3734165463030244e-05, | |
| "loss": 2.4, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7604166666666666, | |
| "grad_norm": 0.1688230501684105, | |
| "learning_rate": 2.345600530872328e-05, | |
| "loss": 2.4442, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7630208333333334, | |
| "grad_norm": 0.16115250940484133, | |
| "learning_rate": 2.3180194846605367e-05, | |
| "loss": 2.4088, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.765625, | |
| "grad_norm": 0.14208493985226064, | |
| "learning_rate": 2.290675462162318e-05, | |
| "loss": 2.2623, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7682291666666666, | |
| "grad_norm": 0.13703687813086513, | |
| "learning_rate": 2.263570500216591e-05, | |
| "loss": 2.486, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7708333333333334, | |
| "grad_norm": 0.16480865487948104, | |
| "learning_rate": 2.2367066178548072e-05, | |
| "loss": 2.321, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7734375, | |
| "grad_norm": 0.15426178206689278, | |
| "learning_rate": 2.2100858161505506e-05, | |
| "loss": 2.2805, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7760416666666666, | |
| "grad_norm": 0.14570862119709826, | |
| "learning_rate": 2.183710078070485e-05, | |
| "loss": 2.551, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7786458333333334, | |
| "grad_norm": 0.14473284935283884, | |
| "learning_rate": 2.157581368326635e-05, | |
| "loss": 2.1132, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 0.15820239745667702, | |
| "learning_rate": 2.1317016332300447e-05, | |
| "loss": 2.4685, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7838541666666666, | |
| "grad_norm": 0.14299466522368254, | |
| "learning_rate": 2.106072800545793e-05, | |
| "loss": 2.4417, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.7864583333333334, | |
| "grad_norm": 0.13794584283858527, | |
| "learning_rate": 2.080696779349396e-05, | |
| "loss": 2.3209, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7890625, | |
| "grad_norm": 0.14676514709546462, | |
| "learning_rate": 2.0555754598846027e-05, | |
| "loss": 2.3155, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7916666666666666, | |
| "grad_norm": 0.1551110645608698, | |
| "learning_rate": 2.030710713422592e-05, | |
| "loss": 2.5221, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.7942708333333334, | |
| "grad_norm": 0.15076270062351085, | |
| "learning_rate": 2.0061043921225828e-05, | |
| "loss": 2.3056, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.796875, | |
| "grad_norm": 0.14884431911568607, | |
| "learning_rate": 1.981758328893866e-05, | |
| "loss": 2.4404, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7994791666666666, | |
| "grad_norm": 0.14993854232241055, | |
| "learning_rate": 1.9576743372592747e-05, | |
| "loss": 2.2778, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8020833333333334, | |
| "grad_norm": 0.13916032144027837, | |
| "learning_rate": 1.933854211220094e-05, | |
| "loss": 2.3168, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8046875, | |
| "grad_norm": 0.1567961212337068, | |
| "learning_rate": 1.9102997251224282e-05, | |
| "loss": 2.3217, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8072916666666666, | |
| "grad_norm": 0.14185229006162023, | |
| "learning_rate": 1.8870126335250293e-05, | |
| "loss": 2.3056, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8098958333333334, | |
| "grad_norm": 0.14083323135716713, | |
| "learning_rate": 1.8639946710686064e-05, | |
| "loss": 2.5101, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 0.1428611888153054, | |
| "learning_rate": 1.841247552346603e-05, | |
| "loss": 2.4123, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "eval_loss": 2.398231029510498, | |
| "eval_runtime": 65.4627, | |
| "eval_samples_per_second": 1.222, | |
| "eval_steps_per_second": 0.153, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8151041666666666, | |
| "grad_norm": 0.13210051743324183, | |
| "learning_rate": 1.8187729717774925e-05, | |
| "loss": 2.2753, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8177083333333334, | |
| "grad_norm": 0.1474646492548143, | |
| "learning_rate": 1.7965726034785466e-05, | |
| "loss": 2.3933, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8203125, | |
| "grad_norm": 0.16871013478557104, | |
| "learning_rate": 1.7746481011411416e-05, | |
| "loss": 2.3614, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8229166666666666, | |
| "grad_norm": 0.15193041380102754, | |
| "learning_rate": 1.753001097907572e-05, | |
| "loss": 2.3525, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8255208333333334, | |
| "grad_norm": 0.14955167945389547, | |
| "learning_rate": 1.7316332062494016e-05, | |
| "loss": 2.4257, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.828125, | |
| "grad_norm": 0.1572944020219996, | |
| "learning_rate": 1.710546017847347e-05, | |
| "loss": 2.5182, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8307291666666666, | |
| "grad_norm": 0.1493341360803509, | |
| "learning_rate": 1.6897411034727218e-05, | |
| "loss": 2.3794, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.13561727835199758, | |
| "learning_rate": 1.66922001287042e-05, | |
| "loss": 2.2277, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8359375, | |
| "grad_norm": 0.13052163343368234, | |
| "learning_rate": 1.648984274643487e-05, | |
| "loss": 2.309, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8385416666666666, | |
| "grad_norm": 0.1578566178286769, | |
| "learning_rate": 1.629035396139247e-05, | |
| "loss": 2.4832, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8411458333333334, | |
| "grad_norm": 0.16163117665249763, | |
| "learning_rate": 1.6093748633370295e-05, | |
| "loss": 2.3515, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.84375, | |
| "grad_norm": 0.16589478114748782, | |
| "learning_rate": 1.5900041407374708e-05, | |
| "loss": 2.5438, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8463541666666666, | |
| "grad_norm": 0.13933316472296842, | |
| "learning_rate": 1.5709246712534315e-05, | |
| "loss": 2.4306, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8489583333333334, | |
| "grad_norm": 0.1606181863885322, | |
| "learning_rate": 1.5521378761025113e-05, | |
| "loss": 2.5248, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.8515625, | |
| "grad_norm": 0.15583835916491617, | |
| "learning_rate": 1.5336451547011838e-05, | |
| "loss": 2.57, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8541666666666666, | |
| "grad_norm": 0.13760770777048018, | |
| "learning_rate": 1.515447884560556e-05, | |
| "loss": 2.4217, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8567708333333334, | |
| "grad_norm": 0.14254277680719865, | |
| "learning_rate": 1.4975474211837561e-05, | |
| "loss": 2.5538, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.859375, | |
| "grad_norm": 0.15022329502854323, | |
| "learning_rate": 1.479945097964967e-05, | |
| "loss": 2.4875, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8619791666666666, | |
| "grad_norm": 0.14442101795365278, | |
| "learning_rate": 1.4626422260900962e-05, | |
| "loss": 2.4053, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8645833333333334, | |
| "grad_norm": 0.1339112894355035, | |
| "learning_rate": 1.4456400944391146e-05, | |
| "loss": 2.3938, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8671875, | |
| "grad_norm": 0.15065063782655086, | |
| "learning_rate": 1.4289399694900398e-05, | |
| "loss": 2.4288, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8697916666666666, | |
| "grad_norm": 0.15054714911892625, | |
| "learning_rate": 1.4125430952246071e-05, | |
| "loss": 2.4214, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8723958333333334, | |
| "grad_norm": 0.1560638225083581, | |
| "learning_rate": 1.3964506930355947e-05, | |
| "loss": 2.3693, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.17170476718899333, | |
| "learning_rate": 1.380663961635852e-05, | |
| "loss": 2.3902, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8776041666666666, | |
| "grad_norm": 0.14853468036224993, | |
| "learning_rate": 1.3651840769690028e-05, | |
| "loss": 2.5229, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8802083333333334, | |
| "grad_norm": 0.13862935132651075, | |
| "learning_rate": 1.350012192121854e-05, | |
| "loss": 2.272, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8828125, | |
| "grad_norm": 0.16359619179880716, | |
| "learning_rate": 1.3351494372384995e-05, | |
| "loss": 2.5031, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8854166666666666, | |
| "grad_norm": 0.15198032362712752, | |
| "learning_rate": 1.3205969194361395e-05, | |
| "loss": 2.355, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8880208333333334, | |
| "grad_norm": 0.12892013869101834, | |
| "learning_rate": 1.3063557227226094e-05, | |
| "loss": 2.3602, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.890625, | |
| "grad_norm": 0.1498170376596413, | |
| "learning_rate": 1.292426907915634e-05, | |
| "loss": 2.3912, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.8932291666666666, | |
| "grad_norm": 0.15638364831022664, | |
| "learning_rate": 1.2788115125638068e-05, | |
| "loss": 2.4359, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8958333333333334, | |
| "grad_norm": 0.151291537071441, | |
| "learning_rate": 1.2655105508693065e-05, | |
| "loss": 2.3082, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8984375, | |
| "grad_norm": 0.14865959228520678, | |
| "learning_rate": 1.252525013612346e-05, | |
| "loss": 2.3877, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9010416666666666, | |
| "grad_norm": 0.1459248998632338, | |
| "learning_rate": 1.2398558680773736e-05, | |
| "loss": 2.3293, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9036458333333334, | |
| "grad_norm": 0.14220792400443616, | |
| "learning_rate": 1.227504057981016e-05, | |
| "loss": 2.4427, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.90625, | |
| "grad_norm": 0.15415554255450412, | |
| "learning_rate": 1.2154705034017866e-05, | |
| "loss": 2.383, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9088541666666666, | |
| "grad_norm": 0.1457198339110242, | |
| "learning_rate": 1.203756100711545e-05, | |
| "loss": 2.3751, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9114583333333334, | |
| "grad_norm": 0.15986391066488098, | |
| "learning_rate": 1.1923617225087293e-05, | |
| "loss": 2.3891, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9140625, | |
| "grad_norm": 0.13681607679501942, | |
| "learning_rate": 1.1812882175533564e-05, | |
| "loss": 2.2878, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9140625, | |
| "eval_loss": 2.397103786468506, | |
| "eval_runtime": 65.1202, | |
| "eval_samples_per_second": 1.228, | |
| "eval_steps_per_second": 0.154, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.17312639183632889, | |
| "learning_rate": 1.1705364107037981e-05, | |
| "loss": 2.4481, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9192708333333334, | |
| "grad_norm": 0.17233940633090827, | |
| "learning_rate": 1.1601071028553371e-05, | |
| "loss": 2.4747, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.921875, | |
| "grad_norm": 0.13057391845213812, | |
| "learning_rate": 1.1500010708805123e-05, | |
| "loss": 2.3276, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9244791666666666, | |
| "grad_norm": 0.15218005054913405, | |
| "learning_rate": 1.1402190675712448e-05, | |
| "loss": 2.4584, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9270833333333334, | |
| "grad_norm": 0.1405908251620449, | |
| "learning_rate": 1.130761821582766e-05, | |
| "loss": 2.3576, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9296875, | |
| "grad_norm": 0.15487043477366738, | |
| "learning_rate": 1.1216300373793417e-05, | |
| "loss": 2.3773, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9322916666666666, | |
| "grad_norm": 0.15092606814376955, | |
| "learning_rate": 1.1128243951817937e-05, | |
| "loss": 2.2986, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9348958333333334, | |
| "grad_norm": 0.1484173556255145, | |
| "learning_rate": 1.1043455509168339e-05, | |
| "loss": 2.2237, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.14235753907847776, | |
| "learning_rate": 1.0961941361682013e-05, | |
| "loss": 2.4375, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9401041666666666, | |
| "grad_norm": 0.14142558045399545, | |
| "learning_rate": 1.0883707581296196e-05, | |
| "loss": 2.5165, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9427083333333334, | |
| "grad_norm": 0.13890140092039985, | |
| "learning_rate": 1.080875999559564e-05, | |
| "loss": 2.477, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9453125, | |
| "grad_norm": 0.149033997441562, | |
| "learning_rate": 1.0737104187378542e-05, | |
| "loss": 2.386, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9479166666666666, | |
| "grad_norm": 0.15829792856388608, | |
| "learning_rate": 1.066874549424068e-05, | |
| "loss": 2.2997, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9505208333333334, | |
| "grad_norm": 0.14793151985278888, | |
| "learning_rate": 1.0603689008177822e-05, | |
| "loss": 2.4599, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.953125, | |
| "grad_norm": 0.1556260763161562, | |
| "learning_rate": 1.0541939575206412e-05, | |
| "loss": 2.2611, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9557291666666666, | |
| "grad_norm": 0.16684067808748887, | |
| "learning_rate": 1.0483501795002612e-05, | |
| "loss": 2.4216, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.9583333333333334, | |
| "grad_norm": 0.14213262998853762, | |
| "learning_rate": 1.0428380020559658e-05, | |
| "loss": 2.4624, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9609375, | |
| "grad_norm": 0.16542015576320396, | |
| "learning_rate": 1.0376578357863627e-05, | |
| "loss": 2.087, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9635416666666666, | |
| "grad_norm": 0.16312011726686781, | |
| "learning_rate": 1.0328100665587574e-05, | |
| "loss": 2.3865, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9661458333333334, | |
| "grad_norm": 0.14758591702174165, | |
| "learning_rate": 1.0282950554804085e-05, | |
| "loss": 2.3726, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.96875, | |
| "grad_norm": 0.1316294327416778, | |
| "learning_rate": 1.0241131388716332e-05, | |
| "loss": 2.4155, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9713541666666666, | |
| "grad_norm": 0.14988844894688882, | |
| "learning_rate": 1.0202646282407505e-05, | |
| "loss": 2.3134, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9739583333333334, | |
| "grad_norm": 0.16468705638317635, | |
| "learning_rate": 1.016749810260881e-05, | |
| "loss": 2.3345, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9765625, | |
| "grad_norm": 0.16639080943133064, | |
| "learning_rate": 1.01356894674859e-05, | |
| "loss": 2.1808, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9791666666666666, | |
| "grad_norm": 0.1537739755991523, | |
| "learning_rate": 1.0107222746443862e-05, | |
| "loss": 2.3806, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.9817708333333334, | |
| "grad_norm": 0.1582641580303379, | |
| "learning_rate": 1.0082100059950713e-05, | |
| "loss": 2.4064, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.984375, | |
| "grad_norm": 0.16490051463401306, | |
| "learning_rate": 1.0060323279379476e-05, | |
| "loss": 2.3932, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9869791666666666, | |
| "grad_norm": 0.14677990127698765, | |
| "learning_rate": 1.0041894026868732e-05, | |
| "loss": 2.3932, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9895833333333334, | |
| "grad_norm": 0.17686734833911588, | |
| "learning_rate": 1.0026813675201832e-05, | |
| "loss": 2.2527, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9921875, | |
| "grad_norm": 0.15542797613175577, | |
| "learning_rate": 1.0015083347704623e-05, | |
| "loss": 2.4111, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.9947916666666666, | |
| "grad_norm": 0.1559166935312072, | |
| "learning_rate": 1.0006703918161775e-05, | |
| "loss": 2.4522, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9973958333333334, | |
| "grad_norm": 0.14284208961058514, | |
| "learning_rate": 1.000167601075169e-05, | |
| "loss": 2.3575, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.14031951339304571, | |
| "learning_rate": 1e-05, | |
| "loss": 2.3152, | |
| "step": 384 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 384, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 39, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 126942053400576.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |