| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 1.0642749071121216, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.6085, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 1.2326161861419678, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.8001, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.036, | |
| "grad_norm": 0.9848614931106567, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 1.6124, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 1.1129230260849, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6241, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9483588933944702, | |
| "learning_rate": 0.00012727272727272728, | |
| "loss": 1.5575, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 0.7823343276977539, | |
| "learning_rate": 0.00015454545454545454, | |
| "loss": 1.3941, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.084, | |
| "grad_norm": 1.2419981956481934, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 1.4403, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 1.2519845962524414, | |
| "learning_rate": 0.00019999906887858078, | |
| "loss": 1.527, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.108, | |
| "grad_norm": 1.28350830078125, | |
| "learning_rate": 0.00019998510240408496, | |
| "loss": 1.2943, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.3078265190124512, | |
| "learning_rate": 0.00019995437844895334, | |
| "loss": 1.2846, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.132, | |
| "grad_norm": 1.412528395652771, | |
| "learning_rate": 0.00019990690216251396, | |
| "loss": 1.0703, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.135581612586975, | |
| "learning_rate": 0.00019984268150178167, | |
| "loss": 1.0773, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.156, | |
| "grad_norm": 1.4680742025375366, | |
| "learning_rate": 0.0001997617272301248, | |
| "loss": 0.8765, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 1.1661655902862549, | |
| "learning_rate": 0.00019966405291546094, | |
| "loss": 1.1527, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.564658284187317, | |
| "learning_rate": 0.00019954967492798333, | |
| "loss": 1.0761, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.2385212182998657, | |
| "learning_rate": 0.00019941861243741684, | |
| "loss": 1.1971, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.204, | |
| "grad_norm": 1.300706148147583, | |
| "learning_rate": 0.0001992708874098054, | |
| "loss": 1.1586, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 1.4915199279785156, | |
| "learning_rate": 0.00019910652460383035, | |
| "loss": 1.0416, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.228, | |
| "grad_norm": 1.3301211595535278, | |
| "learning_rate": 0.00019892555156666089, | |
| "loss": 0.9176, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.579613447189331, | |
| "learning_rate": 0.00019872799862933731, | |
| "loss": 0.924, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.252, | |
| "grad_norm": 1.756352424621582, | |
| "learning_rate": 0.0001985138989016874, | |
| "loss": 1.1, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 1.9087395668029785, | |
| "learning_rate": 0.00019828328826677726, | |
| "loss": 1.3137, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.276, | |
| "grad_norm": 1.5105749368667603, | |
| "learning_rate": 0.00019803620537489736, | |
| "loss": 0.971, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.9009131193161011, | |
| "learning_rate": 0.00019777269163708468, | |
| "loss": 0.918, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.5716756582260132, | |
| "learning_rate": 0.00019749279121818235, | |
| "loss": 0.9052, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 1.2776657342910767, | |
| "learning_rate": 0.00019719655102943753, | |
| "loss": 1.0727, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.324, | |
| "grad_norm": 1.7678760290145874, | |
| "learning_rate": 0.00019688402072063903, | |
| "loss": 1.0161, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 1.4070408344268799, | |
| "learning_rate": 0.00019655525267179625, | |
| "loss": 0.9853, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.348, | |
| "grad_norm": 1.3068615198135376, | |
| "learning_rate": 0.00019621030198436006, | |
| "loss": 0.8356, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.5878366231918335, | |
| "learning_rate": 0.000195849226471988, | |
| "loss": 0.8672, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.372, | |
| "grad_norm": 1.5617555379867554, | |
| "learning_rate": 0.00019547208665085457, | |
| "loss": 0.9721, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.2143884897232056, | |
| "learning_rate": 0.00019507894572950882, | |
| "loss": 0.8779, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.396, | |
| "grad_norm": 1.5331264734268188, | |
| "learning_rate": 0.0001946698695982806, | |
| "loss": 0.7773, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 1.795127272605896, | |
| "learning_rate": 0.00019424492681823733, | |
| "loss": 1.0404, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.2372801303863525, | |
| "learning_rate": 0.00019380418860969322, | |
| "loss": 1.1664, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.2694289684295654, | |
| "learning_rate": 0.00019334772884027266, | |
| "loss": 0.7488, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.444, | |
| "grad_norm": 1.663999319076538, | |
| "learning_rate": 0.00019287562401253022, | |
| "loss": 1.1851, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 1.3195521831512451, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 0.9345, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.468, | |
| "grad_norm": 1.3398083448410034, | |
| "learning_rate": 0.00019188479828957772, | |
| "loss": 0.8728, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.5292913913726807, | |
| "learning_rate": 0.00019136624345653558, | |
| "loss": 0.8164, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.492, | |
| "grad_norm": 1.5369495153427124, | |
| "learning_rate": 0.0001908323756616754, | |
| "loss": 0.9344, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 1.511875033378601, | |
| "learning_rate": 0.00019028328438111934, | |
| "loss": 0.7121, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.516, | |
| "grad_norm": 2.1549644470214844, | |
| "learning_rate": 0.00018971906164244232, | |
| "loss": 0.9891, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 2.0295217037200928, | |
| "learning_rate": 0.00018913980200924822, | |
| "loss": 1.0895, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.9123669862747192, | |
| "learning_rate": 0.000188545602565321, | |
| "loss": 0.798, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 1.4052222967147827, | |
| "learning_rate": 0.00018793656289835365, | |
| "loss": 0.8515, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.564, | |
| "grad_norm": 1.736244797706604, | |
| "learning_rate": 0.00018731278508325708, | |
| "loss": 1.1771, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.6685560941696167, | |
| "learning_rate": 0.0001866743736650526, | |
| "loss": 0.9137, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.588, | |
| "grad_norm": 1.3800030946731567, | |
| "learning_rate": 0.0001860214356413501, | |
| "loss": 0.8261, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.19209623336792, | |
| "learning_rate": 0.00018535408044441514, | |
| "loss": 1.0179, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.612, | |
| "grad_norm": 1.3164235353469849, | |
| "learning_rate": 0.00018467241992282843, | |
| "loss": 0.7551, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 1.8980730772018433, | |
| "learning_rate": 0.0001839765683227398, | |
| "loss": 0.939, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.636, | |
| "grad_norm": 1.7097257375717163, | |
| "learning_rate": 0.00018326664226872065, | |
| "loss": 0.854, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 1.4715948104858398, | |
| "learning_rate": 0.00018254276074421768, | |
| "loss": 0.8128, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.0332729816436768, | |
| "learning_rate": 0.0001818050450716113, | |
| "loss": 0.9127, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 1.5340715646743774, | |
| "learning_rate": 0.00018105361889188203, | |
| "loss": 0.6944, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.684, | |
| "grad_norm": 1.4474735260009766, | |
| "learning_rate": 0.00018028860814388827, | |
| "loss": 0.9063, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 1.3392386436462402, | |
| "learning_rate": 0.00017951014104325904, | |
| "loss": 1.1269, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708, | |
| "grad_norm": 1.6966164112091064, | |
| "learning_rate": 0.00017871834806090501, | |
| "loss": 0.7627, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.2823700904846191, | |
| "learning_rate": 0.00017791336190115165, | |
| "loss": 0.8398, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.732, | |
| "grad_norm": 1.6307556629180908, | |
| "learning_rate": 0.00017709531747949796, | |
| "loss": 0.931, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 1.6843342781066895, | |
| "learning_rate": 0.00017626435190000467, | |
| "loss": 0.8472, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.756, | |
| "grad_norm": 1.5309902429580688, | |
| "learning_rate": 0.00017542060443231572, | |
| "loss": 1.0447, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.315926194190979, | |
| "learning_rate": 0.00017456421648831655, | |
| "loss": 0.8866, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.4360320568084717, | |
| "learning_rate": 0.00017369533159843369, | |
| "loss": 0.8383, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 1.3954721689224243, | |
| "learning_rate": 0.00017281409538757883, | |
| "loss": 0.8642, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.804, | |
| "grad_norm": 1.0596987009048462, | |
| "learning_rate": 0.00017192065555074245, | |
| "loss": 1.0095, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 1.3004777431488037, | |
| "learning_rate": 0.00017101516182823986, | |
| "loss": 0.8043, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.828, | |
| "grad_norm": 1.6855559349060059, | |
| "learning_rate": 0.00017009776598061495, | |
| "loss": 0.6924, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.01497745513916, | |
| "learning_rate": 0.00016916862176320508, | |
| "loss": 0.9077, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.852, | |
| "grad_norm": 1.7603662014007568, | |
| "learning_rate": 0.00016822788490037177, | |
| "loss": 0.7555, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 1.3321644067764282, | |
| "learning_rate": 0.00016727571305940125, | |
| "loss": 0.8248, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.876, | |
| "grad_norm": 1.8733290433883667, | |
| "learning_rate": 0.00016631226582407952, | |
| "loss": 0.9551, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 1.51967453956604, | |
| "learning_rate": 0.00016533770466794622, | |
| "loss": 0.8852, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.008765935897827, | |
| "learning_rate": 0.00016435219292723147, | |
| "loss": 0.8557, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.6430020332336426, | |
| "learning_rate": 0.00016335589577348104, | |
| "loss": 0.9588, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.924, | |
| "grad_norm": 1.231882929801941, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.8806, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 1.5018281936645508, | |
| "learning_rate": 0.0001613316149232341, | |
| "loss": 0.8107, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.948, | |
| "grad_norm": 1.7227574586868286, | |
| "learning_rate": 0.00016030397049575203, | |
| "loss": 0.8624, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.324074387550354, | |
| "learning_rate": 0.0001592662191364017, | |
| "loss": 0.8795, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.972, | |
| "grad_norm": 1.803436279296875, | |
| "learning_rate": 0.00015821853477207708, | |
| "loss": 0.9413, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 1.5749870538711548, | |
| "learning_rate": 0.00015716109299444158, | |
| "loss": 0.7096, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.996, | |
| "grad_norm": 1.7215006351470947, | |
| "learning_rate": 0.00015609407103049896, | |
| "loss": 0.8372, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.418531894683838, | |
| "learning_rate": 0.0001550176477128899, | |
| "loss": 0.9418, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.624677062034607, | |
| "learning_rate": 0.00015393200344991995, | |
| "loss": 0.6523, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 1.8510764837265015, | |
| "learning_rate": 0.0001528373201953229, | |
| "loss": 0.6848, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.044, | |
| "grad_norm": 2.023068904876709, | |
| "learning_rate": 0.00015173378141776568, | |
| "loss": 0.6711, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 1.6575472354888916, | |
| "learning_rate": 0.00015062157207009874, | |
| "loss": 0.9591, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.068, | |
| "grad_norm": 1.551555871963501, | |
| "learning_rate": 0.00014950087855835815, | |
| "loss": 0.5599, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.8778928518295288, | |
| "learning_rate": 0.000148371888710524, | |
| "loss": 0.8402, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.092, | |
| "grad_norm": 1.701336145401001, | |
| "learning_rate": 0.00014723479174504037, | |
| "loss": 0.646, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 2.068992853164673, | |
| "learning_rate": 0.00014608977823910257, | |
| "loss": 0.8327, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.116, | |
| "grad_norm": 2.2500243186950684, | |
| "learning_rate": 0.00014493704009671613, | |
| "loss": 0.8004, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 1.7683219909667969, | |
| "learning_rate": 0.00014377677051653404, | |
| "loss": 0.7407, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 1.5629931688308716, | |
| "learning_rate": 0.00014260916395947656, | |
| "loss": 0.6372, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 1.4379006624221802, | |
| "learning_rate": 0.0001414344161161395, | |
| "loss": 0.714, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.164, | |
| "grad_norm": 2.0503599643707275, | |
| "learning_rate": 0.00014025272387399674, | |
| "loss": 0.7416, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 1.7053464651107788, | |
| "learning_rate": 0.0001390642852844019, | |
| "loss": 0.7885, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.188, | |
| "grad_norm": 1.7159677743911743, | |
| "learning_rate": 0.00013786929952939477, | |
| "loss": 0.7757, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.182704448699951, | |
| "learning_rate": 0.00013666796688831866, | |
| "loss": 0.7267, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.212, | |
| "grad_norm": 1.7983808517456055, | |
| "learning_rate": 0.00013546048870425356, | |
| "loss": 0.8256, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 2.5008952617645264, | |
| "learning_rate": 0.00013424706735027108, | |
| "loss": 0.8931, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.236, | |
| "grad_norm": 2.0425970554351807, | |
| "learning_rate": 0.00013302790619551674, | |
| "loss": 0.8057, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 1.6034411191940308, | |
| "learning_rate": 0.00013180320957112537, | |
| "loss": 0.5682, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.7696748971939087, | |
| "learning_rate": 0.0001305731827359753, | |
| "loss": 0.6621, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 1.8495662212371826, | |
| "learning_rate": 0.000129338031842287, | |
| "loss": 1.0297, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.284, | |
| "grad_norm": 1.9274413585662842, | |
| "learning_rate": 0.00012809796390107195, | |
| "loss": 0.698, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 1.388038158416748, | |
| "learning_rate": 0.0001268531867474377, | |
| "loss": 0.6074, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.308, | |
| "grad_norm": 1.9666837453842163, | |
| "learning_rate": 0.0001256039090057547, | |
| "loss": 0.6017, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 2.2773425579071045, | |
| "learning_rate": 0.00012435034005469107, | |
| "loss": 0.7342, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.332, | |
| "grad_norm": 2.059131383895874, | |
| "learning_rate": 0.0001230926899921206, | |
| "loss": 0.7482, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 1.4702714681625366, | |
| "learning_rate": 0.00012183116959991024, | |
| "loss": 0.6535, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3559999999999999, | |
| "grad_norm": 2.339047908782959, | |
| "learning_rate": 0.00012056599030859366, | |
| "loss": 0.6876, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 1.404402732849121, | |
| "learning_rate": 0.000119297364161935, | |
| "loss": 0.6558, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.5481153726577759, | |
| "learning_rate": 0.0001180255037813906, | |
| "loss": 0.8033, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 1.8469116687774658, | |
| "learning_rate": 0.00011675062233047364, | |
| "loss": 0.9062, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.404, | |
| "grad_norm": 2.170661211013794, | |
| "learning_rate": 0.00011547293347902812, | |
| "loss": 1.0204, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 1.6551048755645752, | |
| "learning_rate": 0.00011419265136741768, | |
| "loss": 0.7657, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.428, | |
| "grad_norm": 1.9395204782485962, | |
| "learning_rate": 0.00011290999057063569, | |
| "loss": 0.5953, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.674118995666504, | |
| "learning_rate": 0.00011162516606234277, | |
| "loss": 0.6788, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.452, | |
| "grad_norm": 2.0402939319610596, | |
| "learning_rate": 0.00011033839317883701, | |
| "loss": 0.6253, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 1.7363512516021729, | |
| "learning_rate": 0.0001090498875829638, | |
| "loss": 0.7973, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.476, | |
| "grad_norm": 3.42630672454834, | |
| "learning_rate": 0.00010775986522797063, | |
| "loss": 0.8157, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 2.1797847747802734, | |
| "learning_rate": 0.0001064685423213136, | |
| "loss": 0.6909, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.1669764518737793, | |
| "learning_rate": 0.00010517613528842097, | |
| "loss": 0.8111, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 1.4340484142303467, | |
| "learning_rate": 0.00010388286073642015, | |
| "loss": 0.8091, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.524, | |
| "grad_norm": 1.9211022853851318, | |
| "learning_rate": 0.00010258893541783476, | |
| "loss": 0.7045, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 1.3780417442321777, | |
| "learning_rate": 0.0001012945761942566, | |
| "loss": 0.6279, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.548, | |
| "grad_norm": 2.026918888092041, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8944, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.160315990447998, | |
| "learning_rate": 9.870542380574341e-05, | |
| "loss": 0.8724, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.572, | |
| "grad_norm": 1.6254347562789917, | |
| "learning_rate": 9.741106458216528e-05, | |
| "loss": 0.787, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 1.8320547342300415, | |
| "learning_rate": 9.61171392635799e-05, | |
| "loss": 0.8769, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.596, | |
| "grad_norm": 1.8300505876541138, | |
| "learning_rate": 9.482386471157904e-05, | |
| "loss": 0.6325, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 1.4335874319076538, | |
| "learning_rate": 9.353145767868638e-05, | |
| "loss": 0.6019, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 2.346412181854248, | |
| "learning_rate": 9.224013477202939e-05, | |
| "loss": 0.5956, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 1.917075276374817, | |
| "learning_rate": 9.095011241703623e-05, | |
| "loss": 0.8203, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6440000000000001, | |
| "grad_norm": 2.0511739253997803, | |
| "learning_rate": 8.9661606821163e-05, | |
| "loss": 0.7623, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 1.5712239742279053, | |
| "learning_rate": 8.837483393765724e-05, | |
| "loss": 0.7278, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6680000000000001, | |
| "grad_norm": 2.1062729358673096, | |
| "learning_rate": 8.709000942936433e-05, | |
| "loss": 0.6844, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 1.0471903085708618, | |
| "learning_rate": 8.580734863258237e-05, | |
| "loss": 0.6001, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.692, | |
| "grad_norm": 2.182645797729492, | |
| "learning_rate": 8.452706652097186e-05, | |
| "loss": 0.773, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 1.5720593929290771, | |
| "learning_rate": 8.324937766952638e-05, | |
| "loss": 0.5371, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.716, | |
| "grad_norm": 2.1978089809417725, | |
| "learning_rate": 8.197449621860943e-05, | |
| "loss": 0.8579, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 2.014761209487915, | |
| "learning_rate": 8.070263583806503e-05, | |
| "loss": 0.7591, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.8996471166610718, | |
| "learning_rate": 7.943400969140635e-05, | |
| "loss": 0.8224, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 1.631941556930542, | |
| "learning_rate": 7.816883040008978e-05, | |
| "loss": 0.5066, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.764, | |
| "grad_norm": 2.4093146324157715, | |
| "learning_rate": 7.690731000787948e-05, | |
| "loss": 0.7607, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 2.3090007305145264, | |
| "learning_rate": 7.56496599453089e-05, | |
| "loss": 0.7572, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.788, | |
| "grad_norm": 1.954633355140686, | |
| "learning_rate": 7.43960909942453e-05, | |
| "loss": 0.6932, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.2641658782958984, | |
| "learning_rate": 7.314681325256232e-05, | |
| "loss": 0.7305, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.812, | |
| "grad_norm": 1.7580173015594482, | |
| "learning_rate": 7.190203609892808e-05, | |
| "loss": 0.6618, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 2.495814085006714, | |
| "learning_rate": 7.066196815771302e-05, | |
| "loss": 0.6998, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8359999999999999, | |
| "grad_norm": 1.600066900253296, | |
| "learning_rate": 6.942681726402473e-05, | |
| "loss": 0.5911, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 1.9360573291778564, | |
| "learning_rate": 6.819679042887468e-05, | |
| "loss": 0.9434, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8599999999999999, | |
| "grad_norm": 2.6737253665924072, | |
| "learning_rate": 6.697209380448333e-05, | |
| "loss": 0.7037, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 2.3846089839935303, | |
| "learning_rate": 6.575293264972893e-05, | |
| "loss": 0.8254, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.884, | |
| "grad_norm": 2.3048312664031982, | |
| "learning_rate": 6.453951129574644e-05, | |
| "loss": 0.9766, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 1.872440218925476, | |
| "learning_rate": 6.333203311168135e-05, | |
| "loss": 0.8265, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.908, | |
| "grad_norm": 1.8152889013290405, | |
| "learning_rate": 6.213070047060524e-05, | |
| "loss": 0.8833, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.7105317115783691, | |
| "learning_rate": 6.093571471559811e-05, | |
| "loss": 0.6763, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.932, | |
| "grad_norm": 2.321948766708374, | |
| "learning_rate": 5.9747276126003257e-05, | |
| "loss": 0.6024, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 1.6151583194732666, | |
| "learning_rate": 5.856558388386055e-05, | |
| "loss": 0.5752, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.956, | |
| "grad_norm": 1.6309243440628052, | |
| "learning_rate": 5.739083604052351e-05, | |
| "loss": 0.4642, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 2.6804358959198, | |
| "learning_rate": 5.622322948346594e-05, | |
| "loss": 0.8228, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 2.1613543033599854, | |
| "learning_rate": 5.506295990328385e-05, | |
| "loss": 0.9623, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 1.6900639533996582, | |
| "learning_rate": 5.3910221760897464e-05, | |
| "loss": 0.7248, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.004, | |
| "grad_norm": 1.9322189092636108, | |
| "learning_rate": 5.276520825495963e-05, | |
| "loss": 0.719, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 1.8491179943084717, | |
| "learning_rate": 5.162811128947602e-05, | |
| "loss": 0.6374, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.028, | |
| "grad_norm": 1.827660322189331, | |
| "learning_rate": 5.0499121441641864e-05, | |
| "loss": 0.7009, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 1.9986926317214966, | |
| "learning_rate": 4.9378427929901306e-05, | |
| "loss": 0.693, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.052, | |
| "grad_norm": 2.270709753036499, | |
| "learning_rate": 4.826621858223431e-05, | |
| "loss": 0.5809, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 1.7829395532608032, | |
| "learning_rate": 4.7162679804677076e-05, | |
| "loss": 0.7653, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.076, | |
| "grad_norm": 1.772600531578064, | |
| "learning_rate": 4.606799655008009e-05, | |
| "loss": 0.4681, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.088, | |
| "grad_norm": 1.8009684085845947, | |
| "learning_rate": 4.498235228711012e-05, | |
| "loss": 0.4102, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.441601276397705, | |
| "learning_rate": 4.3905928969501056e-05, | |
| "loss": 0.6207, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 2.350313663482666, | |
| "learning_rate": 4.2838907005558406e-05, | |
| "loss": 0.659, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.124, | |
| "grad_norm": 2.4971227645874023, | |
| "learning_rate": 4.1781465227922957e-05, | |
| "loss": 0.6974, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.136, | |
| "grad_norm": 2.6324408054351807, | |
| "learning_rate": 4.0733780863598335e-05, | |
| "loss": 0.7108, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.148, | |
| "grad_norm": 2.413882255554199, | |
| "learning_rate": 3.9696029504247956e-05, | |
| "loss": 0.6228, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.309478282928467, | |
| "learning_rate": 3.866838507676592e-05, | |
| "loss": 0.6082, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.172, | |
| "grad_norm": 1.7304553985595703, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.482, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.184, | |
| "grad_norm": 1.6498790979385376, | |
| "learning_rate": 3.664410422651898e-05, | |
| "loss": 0.5673, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.196, | |
| "grad_norm": 2.485692262649536, | |
| "learning_rate": 3.5647807072768526e-05, | |
| "loss": 0.6241, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 2.3340401649475098, | |
| "learning_rate": 3.466229533205382e-05, | |
| "loss": 0.6536, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 2.2559235095977783, | |
| "learning_rate": 3.36877341759205e-05, | |
| "loss": 0.5349, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.232, | |
| "grad_norm": 2.2168805599212646, | |
| "learning_rate": 3.272428694059876e-05, | |
| "loss": 0.587, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.2439999999999998, | |
| "grad_norm": 2.802232503890991, | |
| "learning_rate": 3.177211509962826e-05, | |
| "loss": 0.7633, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 2.523928642272949, | |
| "learning_rate": 3.083137823679493e-05, | |
| "loss": 0.832, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.268, | |
| "grad_norm": 3.1273038387298584, | |
| "learning_rate": 2.9902234019385057e-05, | |
| "loss": 0.6549, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 1.7540401220321655, | |
| "learning_rate": 2.8984838171760143e-05, | |
| "loss": 0.5879, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.292, | |
| "grad_norm": 1.6333884000778198, | |
| "learning_rate": 2.8079344449257572e-05, | |
| "loss": 0.4111, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 2.337137222290039, | |
| "learning_rate": 2.7185904612421176e-05, | |
| "loss": 0.6164, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.316, | |
| "grad_norm": 2.0755574703216553, | |
| "learning_rate": 2.6304668401566335e-05, | |
| "loss": 0.5297, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.328, | |
| "grad_norm": 1.6737080812454224, | |
| "learning_rate": 2.5435783511683443e-05, | |
| "loss": 0.5128, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 1.6794757843017578, | |
| "learning_rate": 2.4579395567684283e-05, | |
| "loss": 0.5672, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 2.5523290634155273, | |
| "learning_rate": 2.373564809999532e-05, | |
| "loss": 0.604, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.364, | |
| "grad_norm": 2.5056543350219727, | |
| "learning_rate": 2.290468252050204e-05, | |
| "loss": 0.6708, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.376, | |
| "grad_norm": 2.3490312099456787, | |
| "learning_rate": 2.2086638098848356e-05, | |
| "loss": 0.5676, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.388, | |
| "grad_norm": 2.346902370452881, | |
| "learning_rate": 2.1281651939094992e-05, | |
| "loss": 0.4205, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.412750482559204, | |
| "learning_rate": 2.048985895674098e-05, | |
| "loss": 0.6508, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.412, | |
| "grad_norm": 1.9428837299346924, | |
| "learning_rate": 1.971139185611176e-05, | |
| "loss": 0.6276, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.424, | |
| "grad_norm": 2.1207127571105957, | |
| "learning_rate": 1.8946381108118006e-05, | |
| "loss": 0.5896, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.436, | |
| "grad_norm": 2.845691680908203, | |
| "learning_rate": 1.819495492838872e-05, | |
| "loss": 0.686, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 2.436755418777466, | |
| "learning_rate": 1.7457239255782333e-05, | |
| "loss": 0.5219, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 1.8439298868179321, | |
| "learning_rate": 1.6733357731279377e-05, | |
| "loss": 0.5294, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.472, | |
| "grad_norm": 2.1576616764068604, | |
| "learning_rate": 1.6023431677260214e-05, | |
| "loss": 0.642, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.484, | |
| "grad_norm": 2.6464619636535645, | |
| "learning_rate": 1.5327580077171587e-05, | |
| "loss": 0.72, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 2.5440258979797363, | |
| "learning_rate": 1.4645919555584885e-05, | |
| "loss": 0.721, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.508, | |
| "grad_norm": 2.1442267894744873, | |
| "learning_rate": 1.3978564358649927e-05, | |
| "loss": 0.6866, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 1.863466739654541, | |
| "learning_rate": 1.3325626334947383e-05, | |
| "loss": 0.6299, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.532, | |
| "grad_norm": 2.593524217605591, | |
| "learning_rate": 1.2687214916742918e-05, | |
| "loss": 0.8559, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 2.2633841037750244, | |
| "learning_rate": 1.206343710164638e-05, | |
| "loss": 0.7061, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.556, | |
| "grad_norm": 2.067728281021118, | |
| "learning_rate": 1.1454397434679021e-05, | |
| "loss": 0.5805, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.568, | |
| "grad_norm": 3.0227150917053223, | |
| "learning_rate": 1.0860197990751807e-05, | |
| "loss": 0.7688, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.5358314514160156, | |
| "learning_rate": 1.028093835755769e-05, | |
| "loss": 0.8175, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 2.311980962753296, | |
| "learning_rate": 9.716715618880668e-06, | |
| "loss": 0.7278, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.604, | |
| "grad_norm": 2.0393810272216797, | |
| "learning_rate": 9.1676243383246e-06, | |
| "loss": 0.5096, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.616, | |
| "grad_norm": 2.303673505783081, | |
| "learning_rate": 8.633756543464421e-06, | |
| "loss": 0.6962, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.628, | |
| "grad_norm": 2.4138078689575195, | |
| "learning_rate": 8.115201710422282e-06, | |
| "loss": 0.5913, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.1802985668182373, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 0.4857, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.652, | |
| "grad_norm": 2.260392427444458, | |
| "learning_rate": 7.124375987469767e-06, | |
| "loss": 0.5718, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.664, | |
| "grad_norm": 2.359084129333496, | |
| "learning_rate": 6.652271159727352e-06, | |
| "loss": 0.572, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.676, | |
| "grad_norm": 2.000070810317993, | |
| "learning_rate": 6.195811390306816e-06, | |
| "loss": 0.5156, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 1.6218239068984985, | |
| "learning_rate": 5.755073181762671e-06, | |
| "loss": 0.5135, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.580967903137207, | |
| "learning_rate": 5.3301304017194135e-06, | |
| "loss": 0.5636, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.7119999999999997, | |
| "grad_norm": 2.247042655944824, | |
| "learning_rate": 4.921054270491187e-06, | |
| "loss": 0.5567, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.724, | |
| "grad_norm": 1.9530930519104004, | |
| "learning_rate": 4.527913349145441e-06, | |
| "loss": 0.5014, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 2.009063482284546, | |
| "learning_rate": 4.150773528012008e-06, | |
| "loss": 0.5435, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.748, | |
| "grad_norm": 2.426750659942627, | |
| "learning_rate": 3.789698015639953e-06, | |
| "loss": 0.5175, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.8535356521606445, | |
| "learning_rate": 3.444747328203779e-06, | |
| "loss": 0.6014, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7720000000000002, | |
| "grad_norm": 2.2690775394439697, | |
| "learning_rate": 3.115979279360992e-06, | |
| "loss": 0.5924, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 2.555889844894409, | |
| "learning_rate": 2.8034489705624944e-06, | |
| "loss": 0.5563, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.7960000000000003, | |
| "grad_norm": 2.984099864959717, | |
| "learning_rate": 2.5072087818176382e-06, | |
| "loss": 0.6043, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.808, | |
| "grad_norm": 2.361802339553833, | |
| "learning_rate": 2.2273083629153147e-06, | |
| "loss": 0.6905, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.0989291667938232, | |
| "learning_rate": 1.963794625102655e-06, | |
| "loss": 0.5107, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 1.436015248298645, | |
| "learning_rate": 1.7167117332227533e-06, | |
| "loss": 0.4399, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.844, | |
| "grad_norm": 2.377495527267456, | |
| "learning_rate": 1.48610109831262e-06, | |
| "loss": 0.5663, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.856, | |
| "grad_norm": 2.165872097015381, | |
| "learning_rate": 1.2720013706627122e-06, | |
| "loss": 0.4201, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.868, | |
| "grad_norm": 1.8650953769683838, | |
| "learning_rate": 1.0744484333391368e-06, | |
| "loss": 0.6054, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 2.115549325942993, | |
| "learning_rate": 8.93475396169674e-07, | |
| "loss": 0.6261, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.892, | |
| "grad_norm": 2.3398571014404297, | |
| "learning_rate": 7.291125901946027e-07, | |
| "loss": 0.5593, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.904, | |
| "grad_norm": 2.0688586235046387, | |
| "learning_rate": 5.813875625831467e-07, | |
| "loss": 0.7174, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.916, | |
| "grad_norm": 2.1658966541290283, | |
| "learning_rate": 4.503250720166774e-07, | |
| "loss": 0.6091, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 1.7528609037399292, | |
| "learning_rate": 3.3594708453906775e-07, | |
| "loss": 0.4243, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.115555763244629, | |
| "learning_rate": 2.382727698752474e-07, | |
| "loss": 0.4401, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.952, | |
| "grad_norm": 2.410830020904541, | |
| "learning_rate": 1.5731849821833954e-07, | |
| "loss": 0.4738, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.964, | |
| "grad_norm": 2.3854079246520996, | |
| "learning_rate": 9.309783748606693e-08, | |
| "loss": 0.597, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 2.455721855163574, | |
| "learning_rate": 4.562155104665955e-08, | |
| "loss": 0.7317, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 2.988, | |
| "grad_norm": 2.517000436782837, | |
| "learning_rate": 1.4897595915053242e-08, | |
| "loss": 0.6519, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.7324557304382324, | |
| "learning_rate": 9.311214192120332e-10, | |
| "loss": 0.6816, | |
| "step": 750 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8318132819066880.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |