| { | |
| "best_metric": 2.5794575214385986, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.017543859649123, | |
| "eval_steps": 50, | |
| "global_step": 129, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.023391812865497075, | |
| "grad_norm": 2.8899965286254883, | |
| "learning_rate": 1e-05, | |
| "loss": 3.621, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.023391812865497075, | |
| "eval_loss": 4.880038738250732, | |
| "eval_runtime": 3.2301, | |
| "eval_samples_per_second": 22.291, | |
| "eval_steps_per_second": 5.573, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04678362573099415, | |
| "grad_norm": 3.9275195598602295, | |
| "learning_rate": 2e-05, | |
| "loss": 3.9065, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 5.344070911407471, | |
| "learning_rate": 3e-05, | |
| "loss": 4.0645, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0935672514619883, | |
| "grad_norm": 4.742477893829346, | |
| "learning_rate": 4e-05, | |
| "loss": 3.9164, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.11695906432748537, | |
| "grad_norm": 4.960448265075684, | |
| "learning_rate": 5e-05, | |
| "loss": 4.1087, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 5.740211009979248, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9335, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.16374269005847952, | |
| "grad_norm": 7.367751598358154, | |
| "learning_rate": 7e-05, | |
| "loss": 3.8536, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1871345029239766, | |
| "grad_norm": 6.333174228668213, | |
| "learning_rate": 8e-05, | |
| "loss": 3.8627, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 7.679598808288574, | |
| "learning_rate": 9e-05, | |
| "loss": 3.4869, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.23391812865497075, | |
| "grad_norm": 8.830903053283691, | |
| "learning_rate": 0.0001, | |
| "loss": 3.8319, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2573099415204678, | |
| "grad_norm": 5.146438121795654, | |
| "learning_rate": 9.998257709344246e-05, | |
| "loss": 3.4504, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 3.7169127464294434, | |
| "learning_rate": 9.993032051607669e-05, | |
| "loss": 3.3142, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.30409356725146197, | |
| "grad_norm": 3.373077869415283, | |
| "learning_rate": 9.984326668636131e-05, | |
| "loss": 3.1032, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.32748538011695905, | |
| "grad_norm": 3.1468732357025146, | |
| "learning_rate": 9.972147627352592e-05, | |
| "loss": 3.0006, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 3.2604331970214844, | |
| "learning_rate": 9.956503415528984e-05, | |
| "loss": 3.1792, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3742690058479532, | |
| "grad_norm": 3.397855758666992, | |
| "learning_rate": 9.937404935870938e-05, | |
| "loss": 3.0764, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.39766081871345027, | |
| "grad_norm": 3.596156120300293, | |
| "learning_rate": 9.91486549841951e-05, | |
| "loss": 2.7991, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 4.534525394439697, | |
| "learning_rate": 9.888900811275204e-05, | |
| "loss": 2.8614, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 5.425872802734375, | |
| "learning_rate": 9.859528969650738e-05, | |
| "loss": 2.9683, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.4678362573099415, | |
| "grad_norm": 10.605550765991211, | |
| "learning_rate": 9.826770443260193e-05, | |
| "loss": 3.8376, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 3.0964808464050293, | |
| "learning_rate": 9.79064806205334e-05, | |
| "loss": 3.1105, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.5146198830409356, | |
| "grad_norm": 2.7125508785247803, | |
| "learning_rate": 9.751187000305076e-05, | |
| "loss": 2.934, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.5380116959064327, | |
| "grad_norm": 2.679877758026123, | |
| "learning_rate": 9.708414759071059e-05, | |
| "loss": 2.8647, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 2.6572399139404297, | |
| "learning_rate": 9.662361147021779e-05, | |
| "loss": 2.7279, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5847953216374269, | |
| "grad_norm": 2.8284733295440674, | |
| "learning_rate": 9.613058259668416e-05, | |
| "loss": 2.7198, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6081871345029239, | |
| "grad_norm": 3.222668409347534, | |
| "learning_rate": 9.56054045699494e-05, | |
| "loss": 2.9284, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 3.3537814617156982, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 2.6431, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.6549707602339181, | |
| "grad_norm": 3.6838319301605225, | |
| "learning_rate": 9.446008722749905e-05, | |
| "loss": 2.7465, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6783625730994152, | |
| "grad_norm": 4.8847432136535645, | |
| "learning_rate": 9.384074610206495e-05, | |
| "loss": 2.8716, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 6.4051103591918945, | |
| "learning_rate": 9.319085164772082e-05, | |
| "loss": 2.7997, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7251461988304093, | |
| "grad_norm": 2.051379680633545, | |
| "learning_rate": 9.251085678648072e-05, | |
| "loss": 2.8797, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.7485380116959064, | |
| "grad_norm": 2.1589481830596924, | |
| "learning_rate": 9.180123541782171e-05, | |
| "loss": 2.5857, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 2.3289027214050293, | |
| "learning_rate": 9.106248208841569e-05, | |
| "loss": 2.5558, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.7953216374269005, | |
| "grad_norm": 2.580897569656372, | |
| "learning_rate": 9.029511164747175e-05, | |
| "loss": 2.5283, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.8187134502923976, | |
| "grad_norm": 2.79526948928833, | |
| "learning_rate": 8.949965888792941e-05, | |
| "loss": 2.4669, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 2.9586293697357178, | |
| "learning_rate": 8.867667817375266e-05, | |
| "loss": 2.4342, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.8654970760233918, | |
| "grad_norm": 3.565655469894409, | |
| "learning_rate": 8.78267430535848e-05, | |
| "loss": 2.5569, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 3.735215425491333, | |
| "learning_rate": 8.695044586103296e-05, | |
| "loss": 2.6734, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 4.107497692108154, | |
| "learning_rate": 8.604839730186125e-05, | |
| "loss": 2.3497, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.935672514619883, | |
| "grad_norm": 6.719999313354492, | |
| "learning_rate": 8.512122602837993e-05, | |
| "loss": 2.793, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.9590643274853801, | |
| "grad_norm": 1.973287582397461, | |
| "learning_rate": 8.416957820132742e-05, | |
| "loss": 2.8444, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 2.6278090476989746, | |
| "learning_rate": 8.319411703955042e-05, | |
| "loss": 2.3696, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.0058479532163742, | |
| "grad_norm": 4.79380464553833, | |
| "learning_rate": 8.219552235779578e-05, | |
| "loss": 3.6136, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.0292397660818713, | |
| "grad_norm": 1.4787054061889648, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 2.1233, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 1.8972117900848389, | |
| "learning_rate": 8.013173181896283e-05, | |
| "loss": 2.4294, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0760233918128654, | |
| "grad_norm": 2.160202980041504, | |
| "learning_rate": 7.9067974251073e-05, | |
| "loss": 2.2788, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0994152046783625, | |
| "grad_norm": 2.476989269256592, | |
| "learning_rate": 7.79839587392154e-05, | |
| "loss": 2.1933, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.1228070175438596, | |
| "grad_norm": 2.577780246734619, | |
| "learning_rate": 7.688044075142887e-05, | |
| "loss": 1.8599, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.1461988304093567, | |
| "grad_norm": 3.1708245277404785, | |
| "learning_rate": 7.57581893473448e-05, | |
| "loss": 1.9908, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.1695906432748537, | |
| "grad_norm": 3.2842891216278076, | |
| "learning_rate": 7.461798664221711e-05, | |
| "loss": 1.838, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.1695906432748537, | |
| "eval_loss": 2.667975425720215, | |
| "eval_runtime": 3.2693, | |
| "eval_samples_per_second": 22.023, | |
| "eval_steps_per_second": 5.506, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.1929824561403508, | |
| "grad_norm": 3.982672691345215, | |
| "learning_rate": 7.346062726185332e-05, | |
| "loss": 1.9851, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.2163742690058479, | |
| "grad_norm": 4.517981052398682, | |
| "learning_rate": 7.228691778882693e-05, | |
| "loss": 1.7932, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.239766081871345, | |
| "grad_norm": 4.8002166748046875, | |
| "learning_rate": 7.109767620035689e-05, | |
| "loss": 2.3657, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 2.333965301513672, | |
| "learning_rate": 6.989373129824604e-05, | |
| "loss": 2.1879, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.286549707602339, | |
| "grad_norm": 2.5241196155548096, | |
| "learning_rate": 6.867592213127558e-05, | |
| "loss": 2.2052, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.3099415204678362, | |
| "grad_norm": 2.5986971855163574, | |
| "learning_rate": 6.744509741045835e-05, | |
| "loss": 2.0875, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 2.735356569290161, | |
| "learning_rate": 6.62021149175583e-05, | |
| "loss": 2.2641, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.3567251461988303, | |
| "grad_norm": 2.861424684524536, | |
| "learning_rate": 6.494784090728852e-05, | |
| "loss": 2.1462, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.3801169590643274, | |
| "grad_norm": 2.953516960144043, | |
| "learning_rate": 6.368314950360415e-05, | |
| "loss": 1.9254, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 3.429384231567383, | |
| "learning_rate": 6.240892209051121e-05, | |
| "loss": 2.0428, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.4269005847953216, | |
| "grad_norm": 3.717503070831299, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 1.6045, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.4502923976608186, | |
| "grad_norm": 4.62086820602417, | |
| "learning_rate": 5.983541738224141e-05, | |
| "loss": 1.5816, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 5.065677165985107, | |
| "learning_rate": 5.853793360434687e-05, | |
| "loss": 2.1671, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.4970760233918128, | |
| "grad_norm": 3.0961179733276367, | |
| "learning_rate": 5.7234499601677026e-05, | |
| "loss": 2.3001, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.52046783625731, | |
| "grad_norm": 2.8138766288757324, | |
| "learning_rate": 5.5926023758585146e-05, | |
| "loss": 2.272, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.543859649122807, | |
| "grad_norm": 2.5831410884857178, | |
| "learning_rate": 5.4613417973165106e-05, | |
| "loss": 1.9027, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.5672514619883042, | |
| "grad_norm": 3.276082754135132, | |
| "learning_rate": 5.329759702173477e-05, | |
| "loss": 2.3347, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.590643274853801, | |
| "grad_norm": 3.1773629188537598, | |
| "learning_rate": 5.197947792131348e-05, | |
| "loss": 1.9598, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.6140350877192984, | |
| "grad_norm": 3.4641201496124268, | |
| "learning_rate": 5.0659979290537954e-05, | |
| "loss": 1.9397, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.6374269005847952, | |
| "grad_norm": 3.880099058151245, | |
| "learning_rate": 4.934002070946206e-05, | |
| "loss": 1.8497, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.6608187134502925, | |
| "grad_norm": 3.90553879737854, | |
| "learning_rate": 4.802052207868654e-05, | |
| "loss": 1.692, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 5.151843070983887, | |
| "learning_rate": 4.670240297826523e-05, | |
| "loss": 1.7849, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.7076023391812867, | |
| "grad_norm": 4.96315336227417, | |
| "learning_rate": 4.5386582026834906e-05, | |
| "loss": 1.9887, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.7309941520467835, | |
| "grad_norm": 1.8645200729370117, | |
| "learning_rate": 4.407397624141487e-05, | |
| "loss": 2.2243, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 2.2499165534973145, | |
| "learning_rate": 4.276550039832299e-05, | |
| "loss": 2.24, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 2.531435251235962, | |
| "learning_rate": 4.146206639565312e-05, | |
| "loss": 2.167, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.801169590643275, | |
| "grad_norm": 2.621704578399658, | |
| "learning_rate": 4.01645826177586e-05, | |
| "loss": 2.0048, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.8245614035087718, | |
| "grad_norm": 3.041447401046753, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 1.9858, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.8479532163742691, | |
| "grad_norm": 3.520282745361328, | |
| "learning_rate": 3.759107790948882e-05, | |
| "loss": 1.9349, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.871345029239766, | |
| "grad_norm": 3.5151638984680176, | |
| "learning_rate": 3.631685049639586e-05, | |
| "loss": 1.6562, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 4.325654983520508, | |
| "learning_rate": 3.505215909271149e-05, | |
| "loss": 1.9125, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.9181286549707601, | |
| "grad_norm": 5.122260570526123, | |
| "learning_rate": 3.379788508244171e-05, | |
| "loss": 1.6561, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.9415204678362574, | |
| "grad_norm": 5.443163871765137, | |
| "learning_rate": 3.255490258954167e-05, | |
| "loss": 2.5903, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.9649122807017543, | |
| "grad_norm": 2.3221733570098877, | |
| "learning_rate": 3.132407786872442e-05, | |
| "loss": 1.9197, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.9883040935672516, | |
| "grad_norm": 4.910060405731201, | |
| "learning_rate": 3.0106268701753965e-05, | |
| "loss": 2.2014, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.0116959064327484, | |
| "grad_norm": 3.926137685775757, | |
| "learning_rate": 2.8902323799643116e-05, | |
| "loss": 2.227, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 1.8208245038986206, | |
| "learning_rate": 2.771308221117309e-05, | |
| "loss": 1.8213, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.0584795321637426, | |
| "grad_norm": 2.0533692836761475, | |
| "learning_rate": 2.6539372738146695e-05, | |
| "loss": 1.6875, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.08187134502924, | |
| "grad_norm": 2.5335185527801514, | |
| "learning_rate": 2.5382013357782893e-05, | |
| "loss": 1.7146, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 2.658461093902588, | |
| "learning_rate": 2.4241810652655196e-05, | |
| "loss": 1.6028, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.128654970760234, | |
| "grad_norm": 2.784259796142578, | |
| "learning_rate": 2.3119559248571128e-05, | |
| "loss": 1.5149, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.152046783625731, | |
| "grad_norm": 3.2446987628936768, | |
| "learning_rate": 2.2016041260784605e-05, | |
| "loss": 1.2587, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 3.726407527923584, | |
| "learning_rate": 2.0932025748927013e-05, | |
| "loss": 1.286, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.198830409356725, | |
| "grad_norm": 4.698680400848389, | |
| "learning_rate": 1.9868268181037185e-05, | |
| "loss": 1.2943, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 4.862926483154297, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.9368, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 4.17417573928833, | |
| "learning_rate": 1.7804477642204222e-05, | |
| "loss": 1.8199, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.2690058479532165, | |
| "grad_norm": 3.123291254043579, | |
| "learning_rate": 1.6805882960449594e-05, | |
| "loss": 2.1358, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.2923976608187133, | |
| "grad_norm": 3.0750467777252197, | |
| "learning_rate": 1.5830421798672568e-05, | |
| "loss": 1.6731, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 3.460376262664795, | |
| "learning_rate": 1.4878773971620074e-05, | |
| "loss": 1.7389, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.3391812865497075, | |
| "grad_norm": 3.5447030067443848, | |
| "learning_rate": 1.3951602698138771e-05, | |
| "loss": 1.5398, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.3391812865497075, | |
| "eval_loss": 2.5794575214385986, | |
| "eval_runtime": 3.2727, | |
| "eval_samples_per_second": 22.0, | |
| "eval_steps_per_second": 5.5, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.3625730994152048, | |
| "grad_norm": 3.9561784267425537, | |
| "learning_rate": 1.3049554138967051e-05, | |
| "loss": 1.5536, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 3.936734914779663, | |
| "learning_rate": 1.2173256946415212e-05, | |
| "loss": 1.4171, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.409356725146199, | |
| "grad_norm": 4.206720352172852, | |
| "learning_rate": 1.1323321826247346e-05, | |
| "loss": 1.2999, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.4327485380116958, | |
| "grad_norm": 5.046553611755371, | |
| "learning_rate": 1.0500341112070605e-05, | |
| "loss": 1.0988, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 5.955111026763916, | |
| "learning_rate": 9.704888352528258e-06, | |
| "loss": 0.9185, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.47953216374269, | |
| "grad_norm": 3.576009511947632, | |
| "learning_rate": 8.93751791158432e-06, | |
| "loss": 1.9256, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.502923976608187, | |
| "grad_norm": 2.3604319095611572, | |
| "learning_rate": 8.198764582178303e-06, | |
| "loss": 1.9458, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 2.6111466884613037, | |
| "learning_rate": 7.489143213519301e-06, | |
| "loss": 1.7967, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.5497076023391814, | |
| "grad_norm": 3.0367538928985596, | |
| "learning_rate": 6.809148352279182e-06, | |
| "loss": 1.4832, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.573099415204678, | |
| "grad_norm": 3.416107416152954, | |
| "learning_rate": 6.159253897935069e-06, | |
| "loss": 1.6191, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 3.648995876312256, | |
| "learning_rate": 5.539912772500944e-06, | |
| "loss": 1.4337, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.6198830409356724, | |
| "grad_norm": 3.623403310775757, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 1.1803, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.6432748538011697, | |
| "grad_norm": 4.428551197052002, | |
| "learning_rate": 4.394595430050613e-06, | |
| "loss": 1.1911, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 4.538227558135986, | |
| "learning_rate": 3.8694174033158555e-06, | |
| "loss": 1.0355, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.690058479532164, | |
| "grad_norm": 5.722489833831787, | |
| "learning_rate": 3.376388529782215e-06, | |
| "loss": 0.9762, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.7134502923976607, | |
| "grad_norm": 3.4204437732696533, | |
| "learning_rate": 2.9158524092894214e-06, | |
| "loss": 1.7561, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 2.083275079727173, | |
| "learning_rate": 2.488129996949251e-06, | |
| "loss": 1.8624, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.760233918128655, | |
| "grad_norm": 2.574385404586792, | |
| "learning_rate": 2.093519379466602e-06, | |
| "loss": 1.7507, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.783625730994152, | |
| "grad_norm": 2.773837089538574, | |
| "learning_rate": 1.7322955673980678e-06, | |
| "loss": 1.6317, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 3.0657918453216553, | |
| "learning_rate": 1.404710303492618e-06, | |
| "loss": 1.7106, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.8304093567251463, | |
| "grad_norm": 3.1810173988342285, | |
| "learning_rate": 1.1109918872479642e-06, | |
| "loss": 1.4181, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.853801169590643, | |
| "grad_norm": 3.865628719329834, | |
| "learning_rate": 8.513450158049108e-07, | |
| "loss": 1.4733, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 3.952064275741577, | |
| "learning_rate": 6.259506412906402e-07, | |
| "loss": 1.1925, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.9005847953216373, | |
| "grad_norm": 4.404904842376709, | |
| "learning_rate": 4.349658447101612e-07, | |
| "loss": 1.0374, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.9239766081871346, | |
| "grad_norm": 5.374385356903076, | |
| "learning_rate": 2.785237264740781e-07, | |
| "loss": 0.8663, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 4.283792018890381, | |
| "learning_rate": 1.567333136387017e-07, | |
| "loss": 1.8765, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.9707602339181287, | |
| "grad_norm": 3.530892848968506, | |
| "learning_rate": 6.967948392331835e-08, | |
| "loss": 1.4727, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.9941520467836256, | |
| "grad_norm": 6.096131324768066, | |
| "learning_rate": 1.742290655755707e-08, | |
| "loss": 1.1021, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.017543859649123, | |
| "grad_norm": 3.1924569606781006, | |
| "learning_rate": 0.0, | |
| "loss": 2.1947, | |
| "step": 129 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 129, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 5.386784338870272e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |