| { | |
| "best_metric": 0.856124222278595, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 1.2738853503184713, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006369426751592357, | |
| "grad_norm": 36.13005447387695, | |
| "learning_rate": 5e-06, | |
| "loss": 9.517, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006369426751592357, | |
| "eval_loss": 4.299320697784424, | |
| "eval_runtime": 22.4063, | |
| "eval_samples_per_second": 11.827, | |
| "eval_steps_per_second": 5.936, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012738853503184714, | |
| "grad_norm": 44.257057189941406, | |
| "learning_rate": 1e-05, | |
| "loss": 8.7487, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01910828025477707, | |
| "grad_norm": 42.081783294677734, | |
| "learning_rate": 1.5e-05, | |
| "loss": 9.2334, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 41.70063018798828, | |
| "learning_rate": 2e-05, | |
| "loss": 9.2985, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03184713375796178, | |
| "grad_norm": 24.960290908813477, | |
| "learning_rate": 2.5e-05, | |
| "loss": 8.3713, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 21.20467758178711, | |
| "learning_rate": 3e-05, | |
| "loss": 7.8663, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.044585987261146494, | |
| "grad_norm": 20.491474151611328, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.8621, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 19.02033042907715, | |
| "learning_rate": 4e-05, | |
| "loss": 7.1343, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05732484076433121, | |
| "grad_norm": 18.769779205322266, | |
| "learning_rate": 4.5e-05, | |
| "loss": 7.121, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06369426751592357, | |
| "grad_norm": 16.574338912963867, | |
| "learning_rate": 5e-05, | |
| "loss": 6.1092, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07006369426751592, | |
| "grad_norm": 16.516620635986328, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.1318, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 17.191225051879883, | |
| "learning_rate": 6e-05, | |
| "loss": 5.93, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08280254777070063, | |
| "grad_norm": 17.699655532836914, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 5.2682, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08917197452229299, | |
| "grad_norm": 18.904544830322266, | |
| "learning_rate": 7e-05, | |
| "loss": 6.294, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09554140127388536, | |
| "grad_norm": 16.874366760253906, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.1507, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 18.243234634399414, | |
| "learning_rate": 8e-05, | |
| "loss": 5.9768, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.10828025477707007, | |
| "grad_norm": 22.026344299316406, | |
| "learning_rate": 8.5e-05, | |
| "loss": 7.0462, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 18.958724975585938, | |
| "learning_rate": 9e-05, | |
| "loss": 5.301, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12101910828025478, | |
| "grad_norm": 18.320955276489258, | |
| "learning_rate": 9.5e-05, | |
| "loss": 5.9237, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 24.6405029296875, | |
| "learning_rate": 0.0001, | |
| "loss": 5.3109, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1337579617834395, | |
| "grad_norm": 39.88923645019531, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 5.7486, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14012738853503184, | |
| "grad_norm": 20.46506690979004, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 6.0248, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1464968152866242, | |
| "grad_norm": 20.671714782714844, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 6.1074, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 24.3786563873291, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 6.1998, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 24.025131225585938, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 5.2428, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16560509554140126, | |
| "grad_norm": 27.44234848022461, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 5.3396, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17197452229299362, | |
| "grad_norm": 27.521507263183594, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 5.9622, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 26.471494674682617, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 5.5839, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18471337579617833, | |
| "grad_norm": 26.029315948486328, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 5.4105, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 28.722759246826172, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 5.0631, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19745222929936307, | |
| "grad_norm": 30.768484115600586, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 5.6846, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 33.13035583496094, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 5.9204, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21019108280254778, | |
| "grad_norm": 28.801864624023438, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 4.7152, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.21656050955414013, | |
| "grad_norm": 29.887720108032227, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 4.2513, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2229299363057325, | |
| "grad_norm": 29.54290771484375, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 5.0721, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 37.83737564086914, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 4.5899, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2356687898089172, | |
| "grad_norm": 43.482608795166016, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 5.0616, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24203821656050956, | |
| "grad_norm": 50.641151428222656, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 6.4259, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2484076433121019, | |
| "grad_norm": 177.5625, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 13.0021, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 136.0759735107422, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 10.0178, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2611464968152866, | |
| "grad_norm": 55.65427017211914, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 5.7104, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 19.339778900146484, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 4.7599, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.27388535031847133, | |
| "grad_norm": 12.372485160827637, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 4.6529, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 9.565322875976562, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 4.4598, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.28662420382165604, | |
| "grad_norm": 8.054713249206543, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 4.507, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2929936305732484, | |
| "grad_norm": 8.118359565734863, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.9682, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.29936305732484075, | |
| "grad_norm": 7.510158061981201, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.9359, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 9.019046783447266, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 4.3822, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.31210191082802546, | |
| "grad_norm": 8.206843376159668, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 3.7949, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 8.42149829864502, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 4.563, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "eval_loss": 1.222476840019226, | |
| "eval_runtime": 22.9758, | |
| "eval_samples_per_second": 11.534, | |
| "eval_steps_per_second": 5.789, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3248407643312102, | |
| "grad_norm": 9.273707389831543, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 4.2956, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33121019108280253, | |
| "grad_norm": 8.437464714050293, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 4.1599, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3375796178343949, | |
| "grad_norm": 8.612960815429688, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 4.1081, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 11.117528915405273, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 4.3433, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3503184713375796, | |
| "grad_norm": 10.288155555725098, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 4.5473, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.35668789808917195, | |
| "grad_norm": 10.283195495605469, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.3766, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3630573248407643, | |
| "grad_norm": 11.619425773620605, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 4.5981, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.36942675159235666, | |
| "grad_norm": 12.74874210357666, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 4.98, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.37579617834394907, | |
| "grad_norm": 11.369183540344238, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 3.341, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 13.796540260314941, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 4.6845, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3885350318471338, | |
| "grad_norm": 13.996506690979004, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 4.5114, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.39490445859872614, | |
| "grad_norm": 16.951208114624023, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 5.0782, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4012738853503185, | |
| "grad_norm": 14.65240478515625, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 4.6793, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.40764331210191085, | |
| "grad_norm": 18.088653564453125, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 4.7098, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4140127388535032, | |
| "grad_norm": 17.526235580444336, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 4.4782, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 18.589906692504883, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 4.4513, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4267515923566879, | |
| "grad_norm": 22.394956588745117, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 4.823, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.43312101910828027, | |
| "grad_norm": 19.558523178100586, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 4.3261, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4394904458598726, | |
| "grad_norm": 19.380760192871094, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 4.6877, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.445859872611465, | |
| "grad_norm": 23.139272689819336, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 4.5216, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45222929936305734, | |
| "grad_norm": 20.861736297607422, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 4.1113, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 21.064624786376953, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 3.894, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.46496815286624205, | |
| "grad_norm": 23.0689697265625, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 4.061, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4713375796178344, | |
| "grad_norm": 25.696874618530273, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 4.2541, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 23.2997989654541, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 4.3783, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4840764331210191, | |
| "grad_norm": 33.244049072265625, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 3.9646, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.49044585987261147, | |
| "grad_norm": 34.73582077026367, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 4.3941, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 68.82785034179688, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 3.9351, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5031847133757962, | |
| "grad_norm": 48.5151252746582, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 6.4497, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 20.338134765625, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.1436, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5159235668789809, | |
| "grad_norm": 8.380877494812012, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 4.0942, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5222929936305732, | |
| "grad_norm": 6.018649101257324, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 3.5251, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5286624203821656, | |
| "grad_norm": 7.0183796882629395, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 3.6932, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 6.874508380889893, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.8955, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5414012738853503, | |
| "grad_norm": 7.094663143157959, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 4.4996, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5477707006369427, | |
| "grad_norm": 6.800449848175049, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 4.3487, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.554140127388535, | |
| "grad_norm": 6.512099742889404, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 3.7667, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5605095541401274, | |
| "grad_norm": 7.194104194641113, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 3.907, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5668789808917197, | |
| "grad_norm": 6.929054260253906, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 3.7032, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 6.628689289093018, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 3.3508, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5796178343949044, | |
| "grad_norm": 7.876663684844971, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 4.032, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5859872611464968, | |
| "grad_norm": 7.960112571716309, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.8419, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5923566878980892, | |
| "grad_norm": 8.09794807434082, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 3.9603, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5987261146496815, | |
| "grad_norm": 9.845130920410156, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 4.0276, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6050955414012739, | |
| "grad_norm": 8.297134399414062, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 3.4893, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 8.784958839416504, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 3.8555, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6178343949044586, | |
| "grad_norm": 10.812994003295898, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 3.7086, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6242038216560509, | |
| "grad_norm": 9.371301651000977, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 3.957, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6305732484076433, | |
| "grad_norm": 11.402350425720215, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 3.8245, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 10.444782257080078, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 3.4607, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "eval_loss": 0.9518508911132812, | |
| "eval_runtime": 22.9599, | |
| "eval_samples_per_second": 11.542, | |
| "eval_steps_per_second": 5.793, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.643312101910828, | |
| "grad_norm": 11.772903442382812, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 3.5184, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 15.014814376831055, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 4.02, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6560509554140127, | |
| "grad_norm": 13.781637191772461, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 3.6895, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6624203821656051, | |
| "grad_norm": 14.441632270812988, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 3.3828, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6687898089171974, | |
| "grad_norm": 16.388935089111328, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 3.8216, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6751592356687898, | |
| "grad_norm": 16.722646713256836, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 3.8913, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6815286624203821, | |
| "grad_norm": 17.876073837280273, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 3.6247, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 18.381223678588867, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 3.9273, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6942675159235668, | |
| "grad_norm": 17.400468826293945, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 3.4908, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7006369426751592, | |
| "grad_norm": 21.45941925048828, | |
| "learning_rate": 5e-05, | |
| "loss": 4.4969, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7070063694267515, | |
| "grad_norm": 21.00533103942871, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 4.1499, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7133757961783439, | |
| "grad_norm": 20.375274658203125, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 3.4835, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7197452229299363, | |
| "grad_norm": 21.039329528808594, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 2.9916, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 25.608665466308594, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 3.5051, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.732484076433121, | |
| "grad_norm": 23.432601928710938, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 2.8175, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7388535031847133, | |
| "grad_norm": 31.607587814331055, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 4.0832, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7452229299363057, | |
| "grad_norm": 47.30448913574219, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 3.5419, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7515923566878981, | |
| "grad_norm": 4.965266227722168, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 3.0148, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7579617834394905, | |
| "grad_norm": 4.364875793457031, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 2.9562, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 4.950608253479004, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 3.453, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7707006369426752, | |
| "grad_norm": 4.8531951904296875, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 2.7735, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7770700636942676, | |
| "grad_norm": 5.037564754486084, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.94, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7834394904458599, | |
| "grad_norm": 5.288941383361816, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 3.0112, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7898089171974523, | |
| "grad_norm": 5.76696252822876, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 3.1363, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 5.585211753845215, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 3.3864, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 6.322639465332031, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 3.5615, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8089171974522293, | |
| "grad_norm": 6.185570240020752, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 3.1405, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8152866242038217, | |
| "grad_norm": 6.54400634765625, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 3.6258, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.821656050955414, | |
| "grad_norm": 6.367386341094971, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 3.153, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8280254777070064, | |
| "grad_norm": 6.623294830322266, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 3.0124, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8343949044585988, | |
| "grad_norm": 6.886430263519287, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 3.1905, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 7.273221969604492, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 2.9941, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8471337579617835, | |
| "grad_norm": 8.015644073486328, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 3.4856, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8535031847133758, | |
| "grad_norm": 8.67126750946045, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 3.5122, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8598726114649682, | |
| "grad_norm": 9.060256958007812, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 3.7418, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8662420382165605, | |
| "grad_norm": 8.686426162719727, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 3.003, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8726114649681529, | |
| "grad_norm": 8.711852073669434, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 3.1511, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 11.177677154541016, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 3.7168, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8853503184713376, | |
| "grad_norm": 10.784727096557617, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 3.0916, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 11.460144996643066, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 3.2102, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8980891719745223, | |
| "grad_norm": 12.439859390258789, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 3.6256, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9044585987261147, | |
| "grad_norm": 13.51606273651123, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 4.201, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.910828025477707, | |
| "grad_norm": 17.250572204589844, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 4.7407, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 16.19012451171875, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 4.0064, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9235668789808917, | |
| "grad_norm": 16.520713806152344, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 3.7402, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9299363057324841, | |
| "grad_norm": 17.06348991394043, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 4.0871, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9363057324840764, | |
| "grad_norm": 16.918140411376953, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 4.0366, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9426751592356688, | |
| "grad_norm": 17.086406707763672, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 3.6202, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9490445859872612, | |
| "grad_norm": 16.144107818603516, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 3.0974, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 18.168882369995117, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 3.2189, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "eval_loss": 0.856124222278595, | |
| "eval_runtime": 22.9621, | |
| "eval_samples_per_second": 11.541, | |
| "eval_steps_per_second": 5.792, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9617834394904459, | |
| "grad_norm": 23.63081932067871, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 4.073, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9681528662420382, | |
| "grad_norm": 19.112306594848633, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 3.6236, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9745222929936306, | |
| "grad_norm": 31.241661071777344, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 3.0811, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9808917197452229, | |
| "grad_norm": 22.64638328552246, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.9199, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9872611464968153, | |
| "grad_norm": 27.255001068115234, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 2.6116, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 66.21800231933594, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 4.2625, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 24.99510383605957, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 4.1692, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0063694267515924, | |
| "grad_norm": 3.7077722549438477, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.7945, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0127388535031847, | |
| "grad_norm": 3.5477285385131836, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 2.4028, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.019108280254777, | |
| "grad_norm": 4.283918857574463, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 3.1117, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0254777070063694, | |
| "grad_norm": 3.8144969940185547, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.6418, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 4.030559539794922, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 2.377, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0382165605095541, | |
| "grad_norm": 4.541848659515381, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 2.5533, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0445859872611465, | |
| "grad_norm": 4.711845874786377, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.4903, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0509554140127388, | |
| "grad_norm": 4.770325183868408, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 2.6872, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0573248407643312, | |
| "grad_norm": 4.293312072753906, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 1.9581, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0636942675159236, | |
| "grad_norm": 5.336209297180176, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 2.0512, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 5.297292232513428, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.6065, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0764331210191083, | |
| "grad_norm": 4.493692874908447, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 1.8749, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0828025477707006, | |
| "grad_norm": 5.324576377868652, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.1537, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.089171974522293, | |
| "grad_norm": 5.945623874664307, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 2.3464, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.0955414012738853, | |
| "grad_norm": 5.990110397338867, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 2.1652, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1019108280254777, | |
| "grad_norm": 5.564594745635986, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 1.9364, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 6.395089149475098, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 2.3456, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 6.796019554138184, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 2.09, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1210191082802548, | |
| "grad_norm": 7.000392913818359, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.333, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.127388535031847, | |
| "grad_norm": 6.889556407928467, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 1.5809, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1337579617834395, | |
| "grad_norm": 7.113092422485352, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.0368, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1401273885350318, | |
| "grad_norm": 6.891910552978516, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 1.5373, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 10.173569679260254, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 1.9954, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1528662420382165, | |
| "grad_norm": 7.4136061668396, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 1.5246, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1592356687898089, | |
| "grad_norm": 8.997750282287598, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.0869, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1656050955414012, | |
| "grad_norm": 9.356908798217773, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 2.0526, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1719745222929936, | |
| "grad_norm": 10.576266288757324, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.0726, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.178343949044586, | |
| "grad_norm": 10.605173110961914, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 1.7944, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 9.885431289672852, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 1.732, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.1910828025477707, | |
| "grad_norm": 10.850882530212402, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 2.0103, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.197452229299363, | |
| "grad_norm": 10.830645561218262, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 1.9147, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2038216560509554, | |
| "grad_norm": 14.65571117401123, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.0657, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2101910828025477, | |
| "grad_norm": 10.46584701538086, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 1.7092, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.21656050955414, | |
| "grad_norm": 11.583234786987305, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 1.5889, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 19.2625732421875, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 1.8016, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2292993630573248, | |
| "grad_norm": 13.321444511413574, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 1.3804, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2356687898089171, | |
| "grad_norm": 13.012649536132812, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 1.2135, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2420382165605095, | |
| "grad_norm": 17.863269805908203, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 1.7409, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2484076433121019, | |
| "grad_norm": 30.971969604492188, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 1.8244, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2547770700636942, | |
| "grad_norm": 4.519101619720459, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 2.9923, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 4.492830753326416, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 2.6286, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.267515923566879, | |
| "grad_norm": 4.864869117736816, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 2.9949, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 4.6130547523498535, | |
| "learning_rate": 0.0, | |
| "loss": 2.3603, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "eval_loss": 0.8630052804946899, | |
| "eval_runtime": 22.948, | |
| "eval_samples_per_second": 11.548, | |
| "eval_steps_per_second": 5.796, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.04485205706539e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |