{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.2738853503184713, "eval_steps": 10000000, "global_step": 1000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.012738853503184714, "grad_norm": 8.56943733236697, "learning_rate": 6.369426751592357e-09, "loss": 2.0819, "step": 10 }, { "epoch": 0.025477707006369428, "grad_norm": 8.50801711736602, "learning_rate": 1.2738853503184714e-08, "loss": 2.1089, "step": 20 }, { "epoch": 0.03821656050955414, "grad_norm": 8.306574604791226, "learning_rate": 1.910828025477707e-08, "loss": 2.1193, "step": 30 }, { "epoch": 0.050955414012738856, "grad_norm": 8.685229549619896, "learning_rate": 2.5477707006369427e-08, "loss": 2.096, "step": 40 }, { "epoch": 0.06369426751592357, "grad_norm": 8.546769161931346, "learning_rate": 3.184713375796178e-08, "loss": 2.0835, "step": 50 }, { "epoch": 0.07643312101910828, "grad_norm": 8.719632286997141, "learning_rate": 3.821656050955414e-08, "loss": 2.1084, "step": 60 }, { "epoch": 0.08917197452229299, "grad_norm": 8.465502680877355, "learning_rate": 4.458598726114649e-08, "loss": 2.106, "step": 70 }, { "epoch": 0.10191082802547771, "grad_norm": 8.160116558852692, "learning_rate": 5.0955414012738854e-08, "loss": 2.1084, "step": 80 }, { "epoch": 0.11464968152866242, "grad_norm": 7.879463909039044, "learning_rate": 5.732484076433121e-08, "loss": 2.0807, "step": 90 }, { "epoch": 0.12738853503184713, "grad_norm": 8.024989565954654, "learning_rate": 6.369426751592356e-08, "loss": 2.0799, "step": 100 }, { "epoch": 0.14012738853503184, "grad_norm": 8.024652217381297, "learning_rate": 7.006369426751591e-08, "loss": 2.0851, "step": 110 }, { "epoch": 0.15286624203821655, "grad_norm": 7.1712069050488125, "learning_rate": 7.643312101910828e-08, "loss": 2.0872, "step": 120 }, { "epoch": 0.16560509554140126, "grad_norm": 7.280737795114974, "learning_rate": 8.280254777070063e-08, "loss": 2.0832, "step": 130 }, { "epoch": 0.17834394904458598, "grad_norm": 7.0433682583690205, "learning_rate": 8.917197452229298e-08, "loss": 2.069, "step": 140 }, { "epoch": 0.1910828025477707, "grad_norm": 6.719930027351258, "learning_rate": 9.554140127388536e-08, "loss": 2.0392, "step": 150 }, { "epoch": 0.20382165605095542, "grad_norm": 5.705591268858002, "learning_rate": 1.0191082802547771e-07, "loss": 2.0493, "step": 160 }, { "epoch": 0.21656050955414013, "grad_norm": 4.575997886363505, "learning_rate": 1.0828025477707006e-07, "loss": 1.983, "step": 170 }, { "epoch": 0.22929936305732485, "grad_norm": 4.418462096774936, "learning_rate": 1.1464968152866242e-07, "loss": 2.0093, "step": 180 }, { "epoch": 0.24203821656050956, "grad_norm": 4.4097948630855255, "learning_rate": 1.2101910828025477e-07, "loss": 2.0138, "step": 190 }, { "epoch": 0.25477707006369427, "grad_norm": 3.5259545363334284, "learning_rate": 1.2738853503184713e-07, "loss": 1.972, "step": 200 }, { "epoch": 0.267515923566879, "grad_norm": 3.0391815131588698, "learning_rate": 1.3375796178343948e-07, "loss": 1.9749, "step": 210 }, { "epoch": 0.2802547770700637, "grad_norm": 2.6254404775057973, "learning_rate": 1.4012738853503183e-07, "loss": 1.9917, "step": 220 }, { "epoch": 0.2929936305732484, "grad_norm": 2.4024744566371505, "learning_rate": 1.464968152866242e-07, "loss": 1.9432, "step": 230 }, { "epoch": 0.3057324840764331, "grad_norm": 2.2547074325053256, "learning_rate": 1.5286624203821656e-07, "loss": 1.9573, "step": 240 }, { "epoch": 0.3184713375796178, "grad_norm": 2.177178569468059, "learning_rate": 1.592356687898089e-07, "loss": 1.9387, "step": 250 }, { "epoch": 0.33121019108280253, "grad_norm": 2.1464165722160495, "learning_rate": 1.6560509554140126e-07, "loss": 1.8581, "step": 260 }, { "epoch": 0.34394904458598724, "grad_norm": 2.087209962861719, "learning_rate": 1.719745222929936e-07, "loss": 1.8983, "step": 270 }, { "epoch": 0.35668789808917195, "grad_norm": 1.9318089948993342, "learning_rate": 1.7834394904458596e-07, "loss": 1.8978, "step": 280 }, { "epoch": 0.36942675159235666, "grad_norm": 1.9797292859696498, "learning_rate": 1.847133757961783e-07, "loss": 1.8881, "step": 290 }, { "epoch": 0.3821656050955414, "grad_norm": 1.8708460589289673, "learning_rate": 1.9108280254777072e-07, "loss": 1.8309, "step": 300 }, { "epoch": 0.39490445859872614, "grad_norm": 1.8593376000097654, "learning_rate": 1.9745222929936307e-07, "loss": 1.8694, "step": 310 }, { "epoch": 0.40764331210191085, "grad_norm": 1.8013443944148582, "learning_rate": 2.0382165605095542e-07, "loss": 1.8743, "step": 320 }, { "epoch": 0.42038216560509556, "grad_norm": 1.7803529845296482, "learning_rate": 2.1019108280254777e-07, "loss": 1.9007, "step": 330 }, { "epoch": 0.43312101910828027, "grad_norm": 1.7991102975353765, "learning_rate": 2.1656050955414012e-07, "loss": 1.8497, "step": 340 }, { "epoch": 0.445859872611465, "grad_norm": 1.8069009405537868, "learning_rate": 2.2292993630573247e-07, "loss": 1.8562, "step": 350 }, { "epoch": 0.4585987261146497, "grad_norm": 1.6273753515480374, "learning_rate": 2.2929936305732485e-07, "loss": 1.8603, "step": 360 }, { "epoch": 0.4713375796178344, "grad_norm": 1.779189816867112, "learning_rate": 2.356687898089172e-07, "loss": 1.8829, "step": 370 }, { "epoch": 0.4840764331210191, "grad_norm": 1.696286378605514, "learning_rate": 2.4203821656050955e-07, "loss": 1.8154, "step": 380 }, { "epoch": 0.4968152866242038, "grad_norm": 1.6332996140543181, "learning_rate": 2.484076433121019e-07, "loss": 1.839, "step": 390 }, { "epoch": 0.5095541401273885, "grad_norm": 1.6812932949046533, "learning_rate": 2.5477707006369425e-07, "loss": 1.8085, "step": 400 }, { "epoch": 0.5222929936305732, "grad_norm": 1.6802631887025108, "learning_rate": 2.611464968152866e-07, "loss": 1.8687, "step": 410 }, { "epoch": 0.535031847133758, "grad_norm": 2.096297259072495, "learning_rate": 2.6751592356687895e-07, "loss": 1.8425, "step": 420 }, { "epoch": 0.5477707006369427, "grad_norm": 1.6250115457169498, "learning_rate": 2.738853503184713e-07, "loss": 1.8526, "step": 430 }, { "epoch": 0.5605095541401274, "grad_norm": 1.9381854040760949, "learning_rate": 2.8025477707006366e-07, "loss": 1.831, "step": 440 }, { "epoch": 0.5732484076433121, "grad_norm": 1.6807327820364515, "learning_rate": 2.86624203821656e-07, "loss": 1.8398, "step": 450 }, { "epoch": 0.5859872611464968, "grad_norm": 1.6107440193241551, "learning_rate": 2.929936305732484e-07, "loss": 1.8267, "step": 460 }, { "epoch": 0.5987261146496815, "grad_norm": 1.5977967583411985, "learning_rate": 2.9936305732484076e-07, "loss": 1.8355, "step": 470 }, { "epoch": 0.6114649681528662, "grad_norm": 1.9497460290846542, "learning_rate": 3.057324840764331e-07, "loss": 1.8377, "step": 480 }, { "epoch": 0.6242038216560509, "grad_norm": 1.6895490491683507, "learning_rate": 3.1210191082802546e-07, "loss": 1.8417, "step": 490 }, { "epoch": 0.6369426751592356, "grad_norm": 1.6275787734591152, "learning_rate": 3.184713375796178e-07, "loss": 1.7875, "step": 500 }, { "epoch": 0.6496815286624203, "grad_norm": 1.6968078173121193, "learning_rate": 3.2484076433121017e-07, "loss": 1.8465, "step": 510 }, { "epoch": 0.6624203821656051, "grad_norm": 1.9081221408565316, "learning_rate": 3.312101910828025e-07, "loss": 1.8044, "step": 520 }, { "epoch": 0.6751592356687898, "grad_norm": 1.7268827395679611, "learning_rate": 3.3757961783439487e-07, "loss": 1.8424, "step": 530 }, { "epoch": 0.6878980891719745, "grad_norm": 1.588010653604496, "learning_rate": 3.439490445859872e-07, "loss": 1.7895, "step": 540 }, { "epoch": 0.7006369426751592, "grad_norm": 1.5048397892480878, "learning_rate": 3.5031847133757957e-07, "loss": 1.8253, "step": 550 }, { "epoch": 0.7133757961783439, "grad_norm": 1.707374118535617, "learning_rate": 3.566878980891719e-07, "loss": 1.8082, "step": 560 }, { "epoch": 0.7261146496815286, "grad_norm": 2.2197114565201517, "learning_rate": 3.6305732484076427e-07, "loss": 1.779, "step": 570 }, { "epoch": 0.7388535031847133, "grad_norm": 1.7191872664512349, "learning_rate": 3.694267515923566e-07, "loss": 1.8024, "step": 580 }, { "epoch": 0.7515923566878981, "grad_norm": 1.408384846451388, "learning_rate": 3.757961783439491e-07, "loss": 1.7709, "step": 590 }, { "epoch": 0.7643312101910829, "grad_norm": 1.4961385659351782, "learning_rate": 3.8216560509554143e-07, "loss": 1.8185, "step": 600 }, { "epoch": 0.7770700636942676, "grad_norm": 1.4966732383387238, "learning_rate": 3.885350318471338e-07, "loss": 1.8015, "step": 610 }, { "epoch": 0.7898089171974523, "grad_norm": 1.5806969964681516, "learning_rate": 3.9490445859872613e-07, "loss": 1.8667, "step": 620 }, { "epoch": 0.802547770700637, "grad_norm": 2.1393134484625773, "learning_rate": 4.012738853503185e-07, "loss": 1.7756, "step": 630 }, { "epoch": 0.8152866242038217, "grad_norm": 1.5449303617675734, "learning_rate": 4.0764331210191083e-07, "loss": 1.8137, "step": 640 }, { "epoch": 0.8280254777070064, "grad_norm": 1.6616798824953516, "learning_rate": 4.140127388535032e-07, "loss": 1.8229, "step": 650 }, { "epoch": 0.8407643312101911, "grad_norm": 1.4727929666770696, "learning_rate": 4.2038216560509554e-07, "loss": 1.7946, "step": 660 }, { "epoch": 0.8535031847133758, "grad_norm": 1.5063558977122367, "learning_rate": 4.267515923566879e-07, "loss": 1.8227, "step": 670 }, { "epoch": 0.8662420382165605, "grad_norm": 1.9073170511583624, "learning_rate": 4.3312101910828024e-07, "loss": 1.8304, "step": 680 }, { "epoch": 0.8789808917197452, "grad_norm": 1.600454635611546, "learning_rate": 4.394904458598726e-07, "loss": 1.8316, "step": 690 }, { "epoch": 0.89171974522293, "grad_norm": 1.612214525129711, "learning_rate": 4.4585987261146494e-07, "loss": 1.7905, "step": 700 }, { "epoch": 0.9044585987261147, "grad_norm": 1.5805218517687452, "learning_rate": 4.522292993630573e-07, "loss": 1.8001, "step": 710 }, { "epoch": 0.9171974522292994, "grad_norm": 1.5226575805751894, "learning_rate": 4.585987261146497e-07, "loss": 1.802, "step": 720 }, { "epoch": 0.9299363057324841, "grad_norm": 2.1066153446994282, "learning_rate": 4.6496815286624205e-07, "loss": 1.824, "step": 730 }, { "epoch": 0.9426751592356688, "grad_norm": 1.5098647042463382, "learning_rate": 4.713375796178344e-07, "loss": 1.7902, "step": 740 }, { "epoch": 0.9554140127388535, "grad_norm": 1.7579270222970877, "learning_rate": 4.777070063694267e-07, "loss": 1.7669, "step": 750 }, { "epoch": 0.9681528662420382, "grad_norm": 1.510473250719683, "learning_rate": 4.840764331210191e-07, "loss": 1.8189, "step": 760 }, { "epoch": 0.9808917197452229, "grad_norm": 2.0426167655298406, "learning_rate": 4.904458598726115e-07, "loss": 1.7999, "step": 770 }, { "epoch": 0.9936305732484076, "grad_norm": 2.1864882903503235, "learning_rate": 4.968152866242038e-07, "loss": 1.8172, "step": 780 }, { "epoch": 1.0063694267515924, "grad_norm": 1.7106458473804984, "learning_rate": 5.031847133757962e-07, "loss": 1.8049, "step": 790 }, { "epoch": 1.019108280254777, "grad_norm": 2.4536538299974238, "learning_rate": 5.095541401273885e-07, "loss": 1.7972, "step": 800 }, { "epoch": 1.0318471337579618, "grad_norm": 2.2703003766318073, "learning_rate": 5.159235668789809e-07, "loss": 1.8039, "step": 810 }, { "epoch": 1.0445859872611465, "grad_norm": 2.04435931962724, "learning_rate": 5.222929936305732e-07, "loss": 1.783, "step": 820 }, { "epoch": 1.0573248407643312, "grad_norm": 1.5182977898972327, "learning_rate": 5.286624203821656e-07, "loss": 1.7601, "step": 830 }, { "epoch": 1.070063694267516, "grad_norm": 2.3955602915546574, "learning_rate": 5.350318471337579e-07, "loss": 1.778, "step": 840 }, { "epoch": 1.0828025477707006, "grad_norm": 2.1173201633271943, "learning_rate": 5.414012738853503e-07, "loss": 1.8397, "step": 850 }, { "epoch": 1.0955414012738853, "grad_norm": 2.438031001903818, "learning_rate": 5.477707006369426e-07, "loss": 1.7873, "step": 860 }, { "epoch": 1.10828025477707, "grad_norm": 2.761676188309256, "learning_rate": 5.54140127388535e-07, "loss": 1.8517, "step": 870 }, { "epoch": 1.1210191082802548, "grad_norm": 1.8173434166079478, "learning_rate": 5.605095541401273e-07, "loss": 1.8146, "step": 880 }, { "epoch": 1.1337579617834395, "grad_norm": 2.592129456495398, "learning_rate": 5.668789808917197e-07, "loss": 1.79, "step": 890 }, { "epoch": 1.1464968152866242, "grad_norm": 2.196256448135963, "learning_rate": 5.73248407643312e-07, "loss": 1.8153, "step": 900 }, { "epoch": 1.1592356687898089, "grad_norm": 2.25064776189445, "learning_rate": 5.796178343949044e-07, "loss": 1.8099, "step": 910 }, { "epoch": 1.1719745222929936, "grad_norm": 1.4088922808400457, "learning_rate": 5.859872611464968e-07, "loss": 1.8257, "step": 920 }, { "epoch": 1.1847133757961783, "grad_norm": 2.559249339951868, "learning_rate": 5.923566878980892e-07, "loss": 1.7803, "step": 930 }, { "epoch": 1.197452229299363, "grad_norm": 1.6812495216892827, "learning_rate": 5.987261146496815e-07, "loss": 1.8071, "step": 940 }, { "epoch": 1.2101910828025477, "grad_norm": 2.9865102020317735, "learning_rate": 6.050955414012739e-07, "loss": 1.8165, "step": 950 }, { "epoch": 1.2229299363057324, "grad_norm": 7.283684341077877, "learning_rate": 6.114649681528662e-07, "loss": 1.7945, "step": 960 }, { "epoch": 1.2356687898089171, "grad_norm": 1.5609113071701854, "learning_rate": 6.178343949044586e-07, "loss": 1.7963, "step": 970 }, { "epoch": 1.2484076433121019, "grad_norm": 2.931531535579773, "learning_rate": 6.242038216560509e-07, "loss": 1.7795, "step": 980 }, { "epoch": 1.2611464968152866, "grad_norm": 3.432859884879658, "learning_rate": 6.305732484076433e-07, "loss": 1.8424, "step": 990 }, { "epoch": 1.2738853503184713, "grad_norm": 3.4669063068366928, "learning_rate": 6.369426751592356e-07, "loss": 1.7614, "step": 1000 } ], "logging_steps": 10, "max_steps": 15700, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 43703909744640.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }