| { | |
| "best_metric": 1.3203791379928589, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.15088645794039984, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007544322897019992, | |
| "grad_norm": 28.610990524291992, | |
| "learning_rate": 5e-06, | |
| "loss": 8.4345, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007544322897019992, | |
| "eval_loss": 2.208278179168701, | |
| "eval_runtime": 220.3339, | |
| "eval_samples_per_second": 10.135, | |
| "eval_steps_per_second": 5.07, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0015088645794039985, | |
| "grad_norm": 29.75434112548828, | |
| "learning_rate": 1e-05, | |
| "loss": 7.8978, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002263296869105998, | |
| "grad_norm": 25.781906127929688, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.1028, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.003017729158807997, | |
| "grad_norm": 21.908246994018555, | |
| "learning_rate": 2e-05, | |
| "loss": 7.9355, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.003772161448509996, | |
| "grad_norm": 21.276395797729492, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.2463, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004526593738211996, | |
| "grad_norm": 18.05831527709961, | |
| "learning_rate": 3e-05, | |
| "loss": 7.0793, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005281026027913994, | |
| "grad_norm": 18.404077529907227, | |
| "learning_rate": 3.5e-05, | |
| "loss": 6.6955, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006035458317615994, | |
| "grad_norm": 14.49231243133545, | |
| "learning_rate": 4e-05, | |
| "loss": 6.241, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0067898906073179935, | |
| "grad_norm": 13.323002815246582, | |
| "learning_rate": 4.5e-05, | |
| "loss": 6.7582, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.007544322897019992, | |
| "grad_norm": 13.41299819946289, | |
| "learning_rate": 5e-05, | |
| "loss": 6.5987, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008298755186721992, | |
| "grad_norm": 12.793521881103516, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 6.5348, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009053187476423991, | |
| "grad_norm": 12.212225914001465, | |
| "learning_rate": 6e-05, | |
| "loss": 5.9619, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.00980761976612599, | |
| "grad_norm": 15.029327392578125, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 6.5328, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.010562052055827989, | |
| "grad_norm": 41.65517807006836, | |
| "learning_rate": 7e-05, | |
| "loss": 6.5128, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.011316484345529988, | |
| "grad_norm": 26.0521240234375, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 6.2811, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.012070916635231988, | |
| "grad_norm": 16.089492797851562, | |
| "learning_rate": 8e-05, | |
| "loss": 6.7146, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.012825348924933987, | |
| "grad_norm": 11.127829551696777, | |
| "learning_rate": 8.5e-05, | |
| "loss": 6.6048, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.013579781214635987, | |
| "grad_norm": 12.313581466674805, | |
| "learning_rate": 9e-05, | |
| "loss": 6.6018, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.014334213504337986, | |
| "grad_norm": 9.649194717407227, | |
| "learning_rate": 9.5e-05, | |
| "loss": 6.4406, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.015088645794039984, | |
| "grad_norm": 9.973852157592773, | |
| "learning_rate": 0.0001, | |
| "loss": 5.958, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.015843078083741986, | |
| "grad_norm": 10.235247611999512, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 6.0049, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.016597510373443983, | |
| "grad_norm": 12.643736839294434, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 5.9654, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01735194266314598, | |
| "grad_norm": 11.938157081604004, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 5.579, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.018106374952847983, | |
| "grad_norm": 11.99202823638916, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 6.014, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01886080724254998, | |
| "grad_norm": 12.833913803100586, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 6.1423, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01961523953225198, | |
| "grad_norm": 9.916985511779785, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 5.8949, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02036967182195398, | |
| "grad_norm": 11.102204322814941, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 6.1882, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.021124104111655977, | |
| "grad_norm": 10.992071151733398, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 6.3971, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.02187853640135798, | |
| "grad_norm": 10.18704605102539, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 6.0592, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.022632968691059976, | |
| "grad_norm": 11.332548141479492, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 6.082, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.023387400980761978, | |
| "grad_norm": 9.355793952941895, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 5.669, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.024141833270463976, | |
| "grad_norm": 9.461305618286133, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 5.5116, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.024896265560165973, | |
| "grad_norm": 10.504962921142578, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 6.328, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.025650697849867975, | |
| "grad_norm": 9.149413108825684, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 5.3682, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.026405130139569973, | |
| "grad_norm": 9.902061462402344, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 5.6849, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.027159562429271974, | |
| "grad_norm": 9.989099502563477, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 5.3639, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.02791399471897397, | |
| "grad_norm": 9.415989875793457, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 5.2341, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.028668427008675973, | |
| "grad_norm": 11.008522987365723, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 5.0794, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.02942285929837797, | |
| "grad_norm": 10.653331756591797, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 6.2048, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03017729158807997, | |
| "grad_norm": 14.365373611450195, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 5.6984, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03093172387778197, | |
| "grad_norm": 11.19192886352539, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 5.2766, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03168615616748397, | |
| "grad_norm": 10.388138771057129, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 5.033, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03244058845718597, | |
| "grad_norm": 9.952187538146973, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 4.9101, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.03319502074688797, | |
| "grad_norm": 10.397631645202637, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 5.0858, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.033949453036589965, | |
| "grad_norm": 10.087145805358887, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 4.4193, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03470388532629196, | |
| "grad_norm": 11.928916931152344, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 4.2049, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03545831761599397, | |
| "grad_norm": 13.15577220916748, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.8644, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.036212749905695965, | |
| "grad_norm": 13.862105369567871, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 3.52, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03696718219539796, | |
| "grad_norm": 9.312602996826172, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 1.3459, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03772161448509996, | |
| "grad_norm": 15.924839973449707, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 2.7023, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03772161448509996, | |
| "eval_loss": 1.6293420791625977, | |
| "eval_runtime": 222.4612, | |
| "eval_samples_per_second": 10.038, | |
| "eval_steps_per_second": 5.021, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03847604677480196, | |
| "grad_norm": 14.912610054016113, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 6.8462, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03923047906450396, | |
| "grad_norm": 12.471588134765625, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 7.0599, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.03998491135420596, | |
| "grad_norm": 8.87014389038086, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 6.3873, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04073934364390796, | |
| "grad_norm": 8.259026527404785, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 5.9529, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04149377593360996, | |
| "grad_norm": 10.203092575073242, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 6.6684, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.042248208223311955, | |
| "grad_norm": 8.597597122192383, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 6.3543, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.04300264051301396, | |
| "grad_norm": 8.163016319274902, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 5.8639, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.04375707280271596, | |
| "grad_norm": 8.327492713928223, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 6.0806, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.044511505092417955, | |
| "grad_norm": 7.924893856048584, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 6.2455, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.04526593738211995, | |
| "grad_norm": 8.130339622497559, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 5.4982, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04602036967182195, | |
| "grad_norm": 8.018017768859863, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 5.7014, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.046774801961523955, | |
| "grad_norm": 8.45346736907959, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 6.2539, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04752923425122595, | |
| "grad_norm": 10.911635398864746, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 6.3726, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04828366654092795, | |
| "grad_norm": 9.591697692871094, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 6.1596, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04903809883062995, | |
| "grad_norm": 7.932915210723877, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 6.1618, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04979253112033195, | |
| "grad_norm": 8.063417434692383, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 5.4566, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05054696341003395, | |
| "grad_norm": 8.73833179473877, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 6.3322, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05130139569973595, | |
| "grad_norm": 10.085183143615723, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 5.8554, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05205582798943795, | |
| "grad_norm": 10.401449203491211, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 6.3202, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.052810260279139945, | |
| "grad_norm": 8.073527336120605, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 5.6743, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05356469256884195, | |
| "grad_norm": 8.164782524108887, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 5.9845, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05431912485854395, | |
| "grad_norm": 7.460975170135498, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 6.0119, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.055073557148245945, | |
| "grad_norm": 8.160001754760742, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 5.8083, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.05582798943794794, | |
| "grad_norm": 7.511910438537598, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.9554, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.05658242172764994, | |
| "grad_norm": 8.06807804107666, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 5.4149, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.057336854017351946, | |
| "grad_norm": 8.23879623413086, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 6.0332, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.058091286307053944, | |
| "grad_norm": 8.210224151611328, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 5.5186, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.05884571859675594, | |
| "grad_norm": 8.879250526428223, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 5.9477, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05960015088645794, | |
| "grad_norm": 8.089262962341309, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 5.8029, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.06035458317615994, | |
| "grad_norm": 8.588730812072754, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 5.6917, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06110901546586194, | |
| "grad_norm": 7.952780723571777, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 5.4262, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06186344775556394, | |
| "grad_norm": 7.111525058746338, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 5.1552, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.06261788004526593, | |
| "grad_norm": 8.799844741821289, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 6.2327, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.06337231233496794, | |
| "grad_norm": 7.895553112030029, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 4.8941, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.06412674462466994, | |
| "grad_norm": 8.473020553588867, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 6.1927, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06488117691437194, | |
| "grad_norm": 8.47846508026123, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 5.1994, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.06563560920407394, | |
| "grad_norm": 8.409833908081055, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 6.0418, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.06639004149377593, | |
| "grad_norm": 8.531917572021484, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 5.0468, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.06714447378347793, | |
| "grad_norm": 8.19450569152832, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 5.3869, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.06789890607317993, | |
| "grad_norm": 8.973502159118652, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 5.2092, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06865333836288193, | |
| "grad_norm": 8.892091751098633, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 5.4139, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.06940777065258392, | |
| "grad_norm": 8.022160530090332, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 4.7068, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07016220294228594, | |
| "grad_norm": 10.1045560836792, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 5.8565, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07091663523198793, | |
| "grad_norm": 9.677287101745605, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 5.1281, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.07167106752168993, | |
| "grad_norm": 10.185721397399902, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 5.3468, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07242549981139193, | |
| "grad_norm": 10.971282958984375, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 4.6916, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07317993210109393, | |
| "grad_norm": 11.636845588684082, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 4.4786, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.07393436439079593, | |
| "grad_norm": 10.711810111999512, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 3.0341, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.07468879668049792, | |
| "grad_norm": 8.94869613647461, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 2.3939, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.07544322897019992, | |
| "grad_norm": 8.477395057678223, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 1.6288, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07544322897019992, | |
| "eval_loss": 1.5527470111846924, | |
| "eval_runtime": 222.1555, | |
| "eval_samples_per_second": 10.052, | |
| "eval_steps_per_second": 5.028, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07619766125990192, | |
| "grad_norm": 12.187577247619629, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 6.6371, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.07695209354960392, | |
| "grad_norm": 11.621190071105957, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 7.1557, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.07770652583930593, | |
| "grad_norm": 9.95458698272705, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 6.7466, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.07846095812900793, | |
| "grad_norm": 7.798550605773926, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 6.3035, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.07921539041870992, | |
| "grad_norm": 7.551530838012695, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 6.4349, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07996982270841192, | |
| "grad_norm": 7.4671630859375, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 6.3822, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08072425499811392, | |
| "grad_norm": 7.680654525756836, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 5.8368, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08147868728781592, | |
| "grad_norm": 7.623833656311035, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 6.2462, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08223311957751792, | |
| "grad_norm": 8.119511604309082, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 5.8799, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.08298755186721991, | |
| "grad_norm": 8.163238525390625, | |
| "learning_rate": 5e-05, | |
| "loss": 6.4398, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08374198415692191, | |
| "grad_norm": 7.675838947296143, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 6.1381, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.08449641644662391, | |
| "grad_norm": 8.161046981811523, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 6.1432, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.08525084873632592, | |
| "grad_norm": 8.056014060974121, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 5.9772, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.08600528102602792, | |
| "grad_norm": 7.2145185470581055, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 5.7573, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.08675971331572992, | |
| "grad_norm": 7.514006614685059, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 5.9614, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.08751414560543191, | |
| "grad_norm": 7.510403633117676, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 5.3342, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.08826857789513391, | |
| "grad_norm": 8.066449165344238, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 5.5604, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.08902301018483591, | |
| "grad_norm": 6.986632347106934, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 5.5458, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.08977744247453791, | |
| "grad_norm": 7.432390213012695, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 5.9329, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0905318747642399, | |
| "grad_norm": 7.102813243865967, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 5.7666, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0912863070539419, | |
| "grad_norm": 7.3202714920043945, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 5.5461, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.0920407393436439, | |
| "grad_norm": 7.094704627990723, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 5.6711, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.09279517163334591, | |
| "grad_norm": 6.997950077056885, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 6.0052, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.09354960392304791, | |
| "grad_norm": 7.559538841247559, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 4.6827, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.09430403621274991, | |
| "grad_norm": 7.945469379425049, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 5.421, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.0950584685024519, | |
| "grad_norm": 7.207925796508789, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 5.95, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0958129007921539, | |
| "grad_norm": 7.7100725173950195, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 6.2309, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0965673330818559, | |
| "grad_norm": 8.336740493774414, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.8438, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.0973217653715579, | |
| "grad_norm": 7.9306511878967285, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 6.0361, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0980761976612599, | |
| "grad_norm": 8.01220417022705, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 6.1294, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0988306299509619, | |
| "grad_norm": 7.395640850067139, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 5.6617, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.0995850622406639, | |
| "grad_norm": 7.897842884063721, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 5.7985, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.1003394945303659, | |
| "grad_norm": 8.189974784851074, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 4.9965, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.1010939268200679, | |
| "grad_norm": 8.065415382385254, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 4.9086, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1018483591097699, | |
| "grad_norm": 8.694807052612305, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 4.776, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1026027913994719, | |
| "grad_norm": 7.551018714904785, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 5.4173, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1033572236891739, | |
| "grad_norm": 7.498192310333252, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 5.1536, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1041116559788759, | |
| "grad_norm": 7.734801769256592, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 4.447, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.10486608826857789, | |
| "grad_norm": 8.053444862365723, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 5.1973, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.10562052055827989, | |
| "grad_norm": 8.864779472351074, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 5.1369, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.10637495284798189, | |
| "grad_norm": 8.096254348754883, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 4.3606, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1071293851376839, | |
| "grad_norm": 8.422588348388672, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 5.2434, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1078838174273859, | |
| "grad_norm": 9.554398536682129, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 4.9714, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.1086382497170879, | |
| "grad_norm": 9.492046356201172, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 5.4207, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1093926820067899, | |
| "grad_norm": 9.344889640808105, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 4.7644, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11014711429649189, | |
| "grad_norm": 9.320569038391113, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 4.5768, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.11090154658619389, | |
| "grad_norm": 9.480978965759277, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 3.6032, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.11165597887589589, | |
| "grad_norm": 9.053571701049805, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 3.3363, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.11241041116559788, | |
| "grad_norm": 7.955657958984375, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 2.2038, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.11316484345529988, | |
| "grad_norm": 10.95832347869873, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 2.481, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11316484345529988, | |
| "eval_loss": 1.3595361709594727, | |
| "eval_runtime": 222.2461, | |
| "eval_samples_per_second": 10.047, | |
| "eval_steps_per_second": 5.026, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11391927574500188, | |
| "grad_norm": 7.0187482833862305, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 5.7281, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.11467370803470389, | |
| "grad_norm": 8.375664710998535, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 6.2015, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.11542814032440589, | |
| "grad_norm": 8.071150779724121, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 6.2623, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.11618257261410789, | |
| "grad_norm": 7.718661308288574, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 6.0401, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.11693700490380989, | |
| "grad_norm": 8.218561172485352, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 6.4259, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.11769143719351188, | |
| "grad_norm": 7.594738960266113, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 6.5689, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.11844586948321388, | |
| "grad_norm": 6.898587703704834, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 5.5474, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.11920030177291588, | |
| "grad_norm": 7.060200214385986, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 5.9418, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.11995473406261788, | |
| "grad_norm": 7.08625602722168, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 5.9037, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.12070916635231987, | |
| "grad_norm": 6.942751407623291, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 6.0381, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12146359864202187, | |
| "grad_norm": 7.1135406494140625, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 6.1544, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.12221803093172388, | |
| "grad_norm": 6.448284149169922, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 5.8762, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.12297246322142588, | |
| "grad_norm": 7.199798107147217, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 6.246, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.12372689551112788, | |
| "grad_norm": 7.024713039398193, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 6.028, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.12448132780082988, | |
| "grad_norm": 6.762689113616943, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 5.8399, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.12523576009053186, | |
| "grad_norm": 6.662613391876221, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 6.0572, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.12599019238023387, | |
| "grad_norm": 7.291405200958252, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 5.7887, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.12674462466993588, | |
| "grad_norm": 8.169452667236328, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 6.4472, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.12749905695963787, | |
| "grad_norm": 7.192326068878174, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 5.6836, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.12825348924933988, | |
| "grad_norm": 7.195889949798584, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 5.4774, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12900792153904186, | |
| "grad_norm": 7.8230204582214355, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 5.5761, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.12976235382874388, | |
| "grad_norm": 7.158181667327881, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 6.0725, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.13051678611844586, | |
| "grad_norm": 7.191005706787109, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 6.3648, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.13127121840814787, | |
| "grad_norm": 7.722646236419678, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 5.8579, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.13202565069784986, | |
| "grad_norm": 7.786721229553223, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 6.0884, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.13278008298755187, | |
| "grad_norm": 7.266101360321045, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 5.5713, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.13353451527725388, | |
| "grad_norm": 8.33176326751709, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 6.0155, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.13428894756695586, | |
| "grad_norm": 7.038465976715088, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 5.3413, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.13504337985665787, | |
| "grad_norm": 7.81002140045166, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 5.9929, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.13579781214635986, | |
| "grad_norm": 7.73304557800293, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 6.365, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13655224443606187, | |
| "grad_norm": 7.7701544761657715, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 5.4176, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.13730667672576385, | |
| "grad_norm": 6.981159687042236, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 4.9839, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.13806110901546587, | |
| "grad_norm": 7.628045558929443, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 5.467, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.13881554130516785, | |
| "grad_norm": 7.262263298034668, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 5.3167, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.13956997359486986, | |
| "grad_norm": 8.176560401916504, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 5.5774, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.14032440588457187, | |
| "grad_norm": 8.238152503967285, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 5.6847, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.14107883817427386, | |
| "grad_norm": 7.836312770843506, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 5.6294, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.14183327046397587, | |
| "grad_norm": 8.08753490447998, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 5.3877, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.14258770275367785, | |
| "grad_norm": 7.177961349487305, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 4.799, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.14334213504337986, | |
| "grad_norm": 7.347300052642822, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 4.6541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.14409656733308185, | |
| "grad_norm": 8.560412406921387, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 5.5527, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.14485099962278386, | |
| "grad_norm": 8.744789123535156, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 5.1754, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.14560543191248584, | |
| "grad_norm": 8.127669334411621, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 4.7211, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.14635986420218786, | |
| "grad_norm": 8.923094749450684, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 5.0364, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.14711429649188984, | |
| "grad_norm": 9.52410888671875, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 4.7301, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.14786872878159185, | |
| "grad_norm": 9.446085929870605, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 4.8496, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.14862316107129386, | |
| "grad_norm": 9.303170204162598, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 4.8698, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.14937759336099585, | |
| "grad_norm": 10.187568664550781, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 3.2888, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.15013202565069786, | |
| "grad_norm": 7.18109655380249, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 2.0294, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.15088645794039984, | |
| "grad_norm": 7.182096004486084, | |
| "learning_rate": 0.0, | |
| "loss": 1.7007, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15088645794039984, | |
| "eval_loss": 1.3203791379928589, | |
| "eval_runtime": 222.0788, | |
| "eval_samples_per_second": 10.055, | |
| "eval_steps_per_second": 5.03, | |
| "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": 2.865573512997765e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |