| { | |
| "best_metric": 0.6734878420829773, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.36363636363636365, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018181818181818182, | |
| "grad_norm": 138.68397521972656, | |
| "learning_rate": 5e-06, | |
| "loss": 14.5689, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0018181818181818182, | |
| "eval_loss": 4.257202625274658, | |
| "eval_runtime": 15.5048, | |
| "eval_samples_per_second": 59.788, | |
| "eval_steps_per_second": 29.926, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036363636363636364, | |
| "grad_norm": 122.03857421875, | |
| "learning_rate": 1e-05, | |
| "loss": 15.5416, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005454545454545455, | |
| "grad_norm": 139.68202209472656, | |
| "learning_rate": 1.5e-05, | |
| "loss": 16.1754, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007272727272727273, | |
| "grad_norm": 154.2438507080078, | |
| "learning_rate": 2e-05, | |
| "loss": 15.2385, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00909090909090909, | |
| "grad_norm": 146.93878173828125, | |
| "learning_rate": 2.5e-05, | |
| "loss": 15.5613, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01090909090909091, | |
| "grad_norm": 120.0749282836914, | |
| "learning_rate": 3e-05, | |
| "loss": 14.4842, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012727272727272728, | |
| "grad_norm": 126.85943603515625, | |
| "learning_rate": 3.5e-05, | |
| "loss": 13.4979, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014545454545454545, | |
| "grad_norm": 175.2422637939453, | |
| "learning_rate": 4e-05, | |
| "loss": 12.2678, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016363636363636365, | |
| "grad_norm": 129.5550537109375, | |
| "learning_rate": 4.5e-05, | |
| "loss": 11.7683, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01818181818181818, | |
| "grad_norm": 114.66398620605469, | |
| "learning_rate": 5e-05, | |
| "loss": 10.7645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 109.88460540771484, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 9.8579, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02181818181818182, | |
| "grad_norm": 122.863037109375, | |
| "learning_rate": 6e-05, | |
| "loss": 9.304, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023636363636363636, | |
| "grad_norm": 123.73020935058594, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 8.3192, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025454545454545455, | |
| "grad_norm": 147.00076293945312, | |
| "learning_rate": 7e-05, | |
| "loss": 7.907, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02727272727272727, | |
| "grad_norm": 338.5273132324219, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 8.9278, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02909090909090909, | |
| "grad_norm": 143.44017028808594, | |
| "learning_rate": 8e-05, | |
| "loss": 7.7997, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03090909090909091, | |
| "grad_norm": 175.75901794433594, | |
| "learning_rate": 8.5e-05, | |
| "loss": 7.0567, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03272727272727273, | |
| "grad_norm": 161.01739501953125, | |
| "learning_rate": 9e-05, | |
| "loss": 7.9452, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.034545454545454546, | |
| "grad_norm": 217.87303161621094, | |
| "learning_rate": 9.5e-05, | |
| "loss": 8.277, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03636363636363636, | |
| "grad_norm": 193.74627685546875, | |
| "learning_rate": 0.0001, | |
| "loss": 7.6536, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.038181818181818185, | |
| "grad_norm": 253.91064453125, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 6.6774, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 213.30625915527344, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 6.5032, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04181818181818182, | |
| "grad_norm": 161.59420776367188, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 6.4562, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04363636363636364, | |
| "grad_norm": 163.30369567871094, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 6.051, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 164.57496643066406, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 5.5738, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04727272727272727, | |
| "grad_norm": 137.8650360107422, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 5.9991, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04909090909090909, | |
| "grad_norm": 147.88990783691406, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 5.2916, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05090909090909091, | |
| "grad_norm": 175.40298461914062, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 5.123, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05272727272727273, | |
| "grad_norm": 143.73516845703125, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 4.696, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05454545454545454, | |
| "grad_norm": 162.42837524414062, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 4.4559, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.056363636363636366, | |
| "grad_norm": 138.40577697753906, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 5.1119, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05818181818181818, | |
| "grad_norm": 152.168212890625, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 4.7344, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 221.61441040039062, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 4.8725, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06181818181818182, | |
| "grad_norm": 134.0053253173828, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 4.1396, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06363636363636363, | |
| "grad_norm": 169.15768432617188, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 4.7024, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06545454545454546, | |
| "grad_norm": 295.5130310058594, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 4.3401, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06727272727272728, | |
| "grad_norm": 170.70733642578125, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 5.3419, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06909090909090909, | |
| "grad_norm": 150.62545776367188, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 4.5806, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07090909090909091, | |
| "grad_norm": 159.96763610839844, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 4.1185, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 145.9441680908203, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 4.1673, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07454545454545454, | |
| "grad_norm": 107.2767333984375, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 4.7757, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07636363636363637, | |
| "grad_norm": 204.49026489257812, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 4.7596, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07818181818181819, | |
| "grad_norm": 115.9788818359375, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 5.3676, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 70.77046966552734, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.8195, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08181818181818182, | |
| "grad_norm": 74.69246673583984, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 4.8097, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08363636363636363, | |
| "grad_norm": 90.72003173828125, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 5.028, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08545454545454545, | |
| "grad_norm": 85.53260803222656, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 4.1861, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08727272727272728, | |
| "grad_norm": 90.01811981201172, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 4.2923, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0890909090909091, | |
| "grad_norm": 103.90205383300781, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 4.0331, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 54.6790657043457, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 5.0269, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "eval_loss": 1.0066642761230469, | |
| "eval_runtime": 15.4751, | |
| "eval_samples_per_second": 59.903, | |
| "eval_steps_per_second": 29.984, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09272727272727273, | |
| "grad_norm": 134.6455535888672, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 3.991, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09454545454545454, | |
| "grad_norm": 106.82148742675781, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 3.4738, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09636363636363636, | |
| "grad_norm": 75.30558013916016, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 3.722, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09818181818181818, | |
| "grad_norm": 73.11234283447266, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 2.7794, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 47.96721267700195, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 3.1889, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10181818181818182, | |
| "grad_norm": 55.4221076965332, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 3.0504, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10363636363636364, | |
| "grad_norm": 85.66980743408203, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 3.4424, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10545454545454545, | |
| "grad_norm": 73.78186798095703, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 2.7928, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10727272727272727, | |
| "grad_norm": 60.2110481262207, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 3.0366, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10909090909090909, | |
| "grad_norm": 95.8529052734375, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 3.5818, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11090909090909092, | |
| "grad_norm": 89.5805892944336, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 2.9999, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11272727272727273, | |
| "grad_norm": 45.49700164794922, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 3.0906, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11454545454545455, | |
| "grad_norm": 66.79446411132812, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 2.9276, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11636363636363636, | |
| "grad_norm": 47.410247802734375, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 2.6777, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11818181818181818, | |
| "grad_norm": 46.62559509277344, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 2.6861, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 51.441925048828125, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 3.3482, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12181818181818181, | |
| "grad_norm": 96.47298431396484, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 2.9825, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12363636363636364, | |
| "grad_norm": 34.14326477050781, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.9371, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12545454545454546, | |
| "grad_norm": 57.666561126708984, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 4.0862, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12727272727272726, | |
| "grad_norm": 40.737274169921875, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 3.4629, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1290909090909091, | |
| "grad_norm": 30.08909034729004, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 3.2942, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13090909090909092, | |
| "grad_norm": 37.007659912109375, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 3.4122, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13272727272727272, | |
| "grad_norm": 40.06611251831055, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 3.4205, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13454545454545455, | |
| "grad_norm": 32.19481658935547, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.5059, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 37.26934814453125, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 3.2033, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13818181818181818, | |
| "grad_norm": 34.74386978149414, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 3.1631, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 30.365886688232422, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 2.9813, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14181818181818182, | |
| "grad_norm": 38.53085708618164, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 3.1473, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14363636363636365, | |
| "grad_norm": 58.952613830566406, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.3567, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 35.19995880126953, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.1655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14727272727272728, | |
| "grad_norm": 42.377830505371094, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.6093, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14909090909090908, | |
| "grad_norm": 36.57028579711914, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 3.458, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1509090909090909, | |
| "grad_norm": 51.59428405761719, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 3.4229, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15272727272727274, | |
| "grad_norm": 39.17350769042969, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.0197, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15454545454545454, | |
| "grad_norm": 41.6758918762207, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 3.0252, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15636363636363637, | |
| "grad_norm": 46.32060241699219, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.81, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15818181818181817, | |
| "grad_norm": 63.42094421386719, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 3.1065, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 40.31554412841797, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 3.1684, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1618181818181818, | |
| "grad_norm": 47.06768035888672, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 3.2292, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16363636363636364, | |
| "grad_norm": 44.603546142578125, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 3.0992, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16545454545454547, | |
| "grad_norm": 40.659542083740234, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 3.1235, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16727272727272727, | |
| "grad_norm": 51.51438522338867, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.1261, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1690909090909091, | |
| "grad_norm": 35.61064910888672, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 3.3095, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1709090909090909, | |
| "grad_norm": 33.06657409667969, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 3.1334, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17272727272727273, | |
| "grad_norm": 42.53789520263672, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 3.1923, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17454545454545456, | |
| "grad_norm": 63.94884490966797, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 3.2337, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17636363636363636, | |
| "grad_norm": 44.34423828125, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 3.6265, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1781818181818182, | |
| "grad_norm": 42.41252899169922, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 4.2179, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 39.532737731933594, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 3.535, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 41.67788314819336, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 4.1957, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "eval_loss": 0.7785072326660156, | |
| "eval_runtime": 15.6483, | |
| "eval_samples_per_second": 59.24, | |
| "eval_steps_per_second": 29.652, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18363636363636363, | |
| "grad_norm": 83.87976837158203, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 2.4931, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.18545454545454546, | |
| "grad_norm": 62.57492446899414, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 2.7229, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.18727272727272729, | |
| "grad_norm": 44.38495635986328, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 2.6211, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1890909090909091, | |
| "grad_norm": 54.712833404541016, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.1997, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.19090909090909092, | |
| "grad_norm": 49.44144058227539, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 2.1469, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19272727272727272, | |
| "grad_norm": 44.76526641845703, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 2.7389, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.19454545454545455, | |
| "grad_norm": 47.95205307006836, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 2.6509, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.19636363636363635, | |
| "grad_norm": 43.48308181762695, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 2.8425, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.19818181818181818, | |
| "grad_norm": 45.31011962890625, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 2.9956, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 47.43437957763672, | |
| "learning_rate": 5e-05, | |
| "loss": 2.7493, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2018181818181818, | |
| "grad_norm": 40.40293502807617, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 2.8543, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.20363636363636364, | |
| "grad_norm": 49.42204666137695, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 3.1348, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.20545454545454545, | |
| "grad_norm": 42.550025939941406, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 2.4712, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20727272727272728, | |
| "grad_norm": 39.86064147949219, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 2.49, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20909090909090908, | |
| "grad_norm": 39.12519454956055, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 2.7172, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2109090909090909, | |
| "grad_norm": 33.93404769897461, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 2.3934, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21272727272727274, | |
| "grad_norm": 42.115943908691406, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 2.597, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21454545454545454, | |
| "grad_norm": 61.48637771606445, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 2.6182, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.21636363636363637, | |
| "grad_norm": 37.781768798828125, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 2.5236, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 40.604156494140625, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 3.1361, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 43.32360076904297, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 3.1341, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.22181818181818183, | |
| "grad_norm": 40.844425201416016, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.4682, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.22363636363636363, | |
| "grad_norm": 80.41752624511719, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 2.5737, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.22545454545454546, | |
| "grad_norm": 48.227821350097656, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 3.3177, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 45.7944221496582, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 2.9183, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2290909090909091, | |
| "grad_norm": 51.166900634765625, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 2.6203, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2309090909090909, | |
| "grad_norm": 46.844322204589844, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 2.8397, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.23272727272727273, | |
| "grad_norm": 61.02446365356445, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.8248, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23454545454545456, | |
| "grad_norm": 65.25340270996094, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 3.2427, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.23636363636363636, | |
| "grad_norm": 60.50672912597656, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 3.0476, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2381818181818182, | |
| "grad_norm": 58.63713836669922, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 3.0828, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 69.12300872802734, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 3.5299, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24181818181818182, | |
| "grad_norm": 43.3862419128418, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 2.9089, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.24363636363636362, | |
| "grad_norm": 55.379859924316406, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.5253, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24545454545454545, | |
| "grad_norm": 46.44770812988281, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 2.6508, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24727272727272728, | |
| "grad_norm": 94.80418395996094, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 3.1445, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24909090909090909, | |
| "grad_norm": 45.59507369995117, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 2.6856, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2509090909090909, | |
| "grad_norm": 41.788909912109375, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 3.0028, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25272727272727274, | |
| "grad_norm": 48.725833892822266, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 3.396, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2545454545454545, | |
| "grad_norm": 41.74631881713867, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 3.4218, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25636363636363635, | |
| "grad_norm": 39.308189392089844, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 2.4569, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2581818181818182, | |
| "grad_norm": 48.53035354614258, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 3.0169, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 58.768245697021484, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 3.017, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.26181818181818184, | |
| "grad_norm": 42.100189208984375, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 2.8048, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2636363636363636, | |
| "grad_norm": 48.90083312988281, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 2.9452, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26545454545454544, | |
| "grad_norm": 57.28050231933594, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.3012, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2672727272727273, | |
| "grad_norm": 41.389217376708984, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 3.5768, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2690909090909091, | |
| "grad_norm": 40.396934509277344, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 3.341, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.27090909090909093, | |
| "grad_norm": 42.712406158447266, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 3.2837, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 51.320068359375, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 3.9642, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "eval_loss": 0.6896986961364746, | |
| "eval_runtime": 15.5088, | |
| "eval_samples_per_second": 59.772, | |
| "eval_steps_per_second": 29.918, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.27454545454545454, | |
| "grad_norm": 123.20899963378906, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 1.9542, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.27636363636363637, | |
| "grad_norm": 71.69274139404297, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.2125, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2781818181818182, | |
| "grad_norm": 104.9278564453125, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 1.8621, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 59.245018005371094, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.1506, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2818181818181818, | |
| "grad_norm": 123.07303619384766, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.8345, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28363636363636363, | |
| "grad_norm": 79.02330017089844, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 2.8691, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.28545454545454546, | |
| "grad_norm": 80.06781005859375, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 1.9592, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2872727272727273, | |
| "grad_norm": 73.02725219726562, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.6779, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.28909090909090907, | |
| "grad_norm": 50.68994140625, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 2.3196, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 72.57898712158203, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.3827, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2927272727272727, | |
| "grad_norm": 82.22683715820312, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.4989, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.29454545454545455, | |
| "grad_norm": 59.47288131713867, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 2.3806, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2963636363636364, | |
| "grad_norm": 55.69282531738281, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 2.0908, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.29818181818181816, | |
| "grad_norm": 49.40199661254883, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.0631, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 81.0521240234375, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 2.9297, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3018181818181818, | |
| "grad_norm": 47.986270904541016, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 2.662, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.30363636363636365, | |
| "grad_norm": 67.48284912109375, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 2.609, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3054545454545455, | |
| "grad_norm": 58.087398529052734, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.5202, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.30727272727272725, | |
| "grad_norm": 57.88316345214844, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 2.3587, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3090909090909091, | |
| "grad_norm": 100.48794555664062, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.3491, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3109090909090909, | |
| "grad_norm": 59.3438606262207, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 2.7131, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.31272727272727274, | |
| "grad_norm": 53.041725158691406, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 3.0228, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3145454545454546, | |
| "grad_norm": 45.357826232910156, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 2.0914, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.31636363636363635, | |
| "grad_norm": 41.874752044677734, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 1.9584, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 47.15371322631836, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 2.3451, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 42.890743255615234, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.4312, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.32181818181818184, | |
| "grad_norm": 115.22737121582031, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 2.4122, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3236363636363636, | |
| "grad_norm": 44.23305892944336, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.2942, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.32545454545454544, | |
| "grad_norm": 41.19794464111328, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 3.0775, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.32727272727272727, | |
| "grad_norm": 67.62647247314453, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 2.9131, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3290909090909091, | |
| "grad_norm": 42.33940124511719, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 2.1782, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.33090909090909093, | |
| "grad_norm": 54.356407165527344, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.4858, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3327272727272727, | |
| "grad_norm": 42.20392990112305, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 2.5662, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.33454545454545453, | |
| "grad_norm": 46.561893463134766, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.6082, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.33636363636363636, | |
| "grad_norm": 53.607418060302734, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 2.8982, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3381818181818182, | |
| "grad_norm": 49.72503662109375, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 2.9618, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 41.763607025146484, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 3.0604, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3418181818181818, | |
| "grad_norm": 48.85137939453125, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 3.3131, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34363636363636363, | |
| "grad_norm": 42.69316101074219, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.9401, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.34545454545454546, | |
| "grad_norm": 57.626651763916016, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.1762, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3472727272727273, | |
| "grad_norm": 44.62453842163086, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 2.9992, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.3490909090909091, | |
| "grad_norm": 48.011863708496094, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 2.984, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3509090909090909, | |
| "grad_norm": 50.390472412109375, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 2.8539, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3527272727272727, | |
| "grad_norm": 44.408843994140625, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.7273, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.35454545454545455, | |
| "grad_norm": 49.21343994140625, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 3.5664, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3563636363636364, | |
| "grad_norm": 48.34445571899414, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.0188, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.35818181818181816, | |
| "grad_norm": 33.7151985168457, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 3.038, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 40.66362762451172, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 2.9837, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3618181818181818, | |
| "grad_norm": 37.93846893310547, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 3.1249, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 36.60719299316406, | |
| "learning_rate": 0.0, | |
| "loss": 3.1517, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "eval_loss": 0.6734878420829773, | |
| "eval_runtime": 15.5531, | |
| "eval_samples_per_second": 59.602, | |
| "eval_steps_per_second": 29.833, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.49024359317504e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |