| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 185, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005405405405405406, | |
| "grad_norm": 2.034959316253662, | |
| "learning_rate": 1e-05, | |
| "loss": 2.3274, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010810810810810811, | |
| "grad_norm": 1.8778449296951294, | |
| "learning_rate": 9.945945945945947e-06, | |
| "loss": 2.227, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.016216216216216217, | |
| "grad_norm": 1.820204734802246, | |
| "learning_rate": 9.891891891891893e-06, | |
| "loss": 2.3029, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.021621621621621623, | |
| "grad_norm": 1.6635195016860962, | |
| "learning_rate": 9.83783783783784e-06, | |
| "loss": 2.2474, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02702702702702703, | |
| "grad_norm": 1.362695336341858, | |
| "learning_rate": 9.783783783783785e-06, | |
| "loss": 2.0898, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.032432432432432434, | |
| "grad_norm": 1.105480670928955, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 2.0848, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03783783783783784, | |
| "grad_norm": 0.9252412915229797, | |
| "learning_rate": 9.675675675675676e-06, | |
| "loss": 2.0229, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.043243243243243246, | |
| "grad_norm": 0.7585290670394897, | |
| "learning_rate": 9.621621621621622e-06, | |
| "loss": 1.9767, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04864864864864865, | |
| "grad_norm": 0.7198324203491211, | |
| "learning_rate": 9.567567567567568e-06, | |
| "loss": 1.9532, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05405405405405406, | |
| "grad_norm": 0.7078152298927307, | |
| "learning_rate": 9.513513513513514e-06, | |
| "loss": 1.845, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05945945945945946, | |
| "grad_norm": 0.7482191920280457, | |
| "learning_rate": 9.45945945945946e-06, | |
| "loss": 1.9299, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06486486486486487, | |
| "grad_norm": 0.7480123043060303, | |
| "learning_rate": 9.405405405405407e-06, | |
| "loss": 1.8255, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07027027027027027, | |
| "grad_norm": 0.8021811842918396, | |
| "learning_rate": 9.351351351351353e-06, | |
| "loss": 1.8584, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07567567567567568, | |
| "grad_norm": 0.8102870583534241, | |
| "learning_rate": 9.297297297297299e-06, | |
| "loss": 1.8065, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08108108108108109, | |
| "grad_norm": 0.8265206813812256, | |
| "learning_rate": 9.243243243243243e-06, | |
| "loss": 1.8241, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08648648648648649, | |
| "grad_norm": 0.7550373673439026, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 1.7726, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0918918918918919, | |
| "grad_norm": 0.6808019876480103, | |
| "learning_rate": 9.135135135135136e-06, | |
| "loss": 1.7703, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0972972972972973, | |
| "grad_norm": 0.6290677189826965, | |
| "learning_rate": 9.081081081081082e-06, | |
| "loss": 1.7586, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10270270270270271, | |
| "grad_norm": 0.6062494516372681, | |
| "learning_rate": 9.027027027027028e-06, | |
| "loss": 1.7713, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.10810810810810811, | |
| "grad_norm": 0.5270338654518127, | |
| "learning_rate": 8.972972972972974e-06, | |
| "loss": 1.652, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11351351351351352, | |
| "grad_norm": 0.5446780920028687, | |
| "learning_rate": 8.91891891891892e-06, | |
| "loss": 1.6587, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11891891891891893, | |
| "grad_norm": 0.500572681427002, | |
| "learning_rate": 8.864864864864866e-06, | |
| "loss": 1.6509, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12432432432432433, | |
| "grad_norm": 0.5181313157081604, | |
| "learning_rate": 8.810810810810811e-06, | |
| "loss": 1.5852, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12972972972972974, | |
| "grad_norm": 0.5200425386428833, | |
| "learning_rate": 8.756756756756759e-06, | |
| "loss": 1.5981, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 0.4856228232383728, | |
| "learning_rate": 8.702702702702703e-06, | |
| "loss": 1.562, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14054054054054055, | |
| "grad_norm": 0.47276848554611206, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 1.6072, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.14594594594594595, | |
| "grad_norm": 0.4729558229446411, | |
| "learning_rate": 8.594594594594595e-06, | |
| "loss": 1.6082, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15135135135135136, | |
| "grad_norm": 0.4218498468399048, | |
| "learning_rate": 8.540540540540542e-06, | |
| "loss": 1.5092, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.15675675675675677, | |
| "grad_norm": 0.4382428526878357, | |
| "learning_rate": 8.486486486486488e-06, | |
| "loss": 1.5741, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16216216216216217, | |
| "grad_norm": 0.4070360064506531, | |
| "learning_rate": 8.432432432432434e-06, | |
| "loss": 1.5187, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.16756756756756758, | |
| "grad_norm": 0.41073259711265564, | |
| "learning_rate": 8.378378378378378e-06, | |
| "loss": 1.4882, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17297297297297298, | |
| "grad_norm": 0.4228561222553253, | |
| "learning_rate": 8.324324324324326e-06, | |
| "loss": 1.4688, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1783783783783784, | |
| "grad_norm": 0.39718520641326904, | |
| "learning_rate": 8.27027027027027e-06, | |
| "loss": 1.4849, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1837837837837838, | |
| "grad_norm": 0.4296083152294159, | |
| "learning_rate": 8.216216216216217e-06, | |
| "loss": 1.4872, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1891891891891892, | |
| "grad_norm": 0.4223209321498871, | |
| "learning_rate": 8.162162162162163e-06, | |
| "loss": 1.4476, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1945945945945946, | |
| "grad_norm": 0.40885603427886963, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 1.4787, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.42571428418159485, | |
| "learning_rate": 8.054054054054055e-06, | |
| "loss": 1.4734, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.20540540540540542, | |
| "grad_norm": 0.3889505863189697, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4245, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.21081081081081082, | |
| "grad_norm": 0.39451542496681213, | |
| "learning_rate": 7.945945945945946e-06, | |
| "loss": 1.4509, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.21621621621621623, | |
| "grad_norm": 0.3710978329181671, | |
| "learning_rate": 7.891891891891894e-06, | |
| "loss": 1.4155, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22162162162162163, | |
| "grad_norm": 0.35685208439826965, | |
| "learning_rate": 7.837837837837838e-06, | |
| "loss": 1.4081, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.22702702702702704, | |
| "grad_norm": 0.33981138467788696, | |
| "learning_rate": 7.783783783783784e-06, | |
| "loss": 1.3643, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.23243243243243245, | |
| "grad_norm": 0.3623262941837311, | |
| "learning_rate": 7.72972972972973e-06, | |
| "loss": 1.4029, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.23783783783783785, | |
| "grad_norm": 0.3363046646118164, | |
| "learning_rate": 7.675675675675676e-06, | |
| "loss": 1.3669, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.24324324324324326, | |
| "grad_norm": 0.33891770243644714, | |
| "learning_rate": 7.621621621621622e-06, | |
| "loss": 1.3635, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.24864864864864866, | |
| "grad_norm": 0.33656665682792664, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 1.352, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.25405405405405407, | |
| "grad_norm": 0.3358979821205139, | |
| "learning_rate": 7.513513513513514e-06, | |
| "loss": 1.3606, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2594594594594595, | |
| "grad_norm": 0.32272347807884216, | |
| "learning_rate": 7.45945945945946e-06, | |
| "loss": 1.342, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2648648648648649, | |
| "grad_norm": 0.3023531436920166, | |
| "learning_rate": 7.4054054054054055e-06, | |
| "loss": 1.3267, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 0.3095267713069916, | |
| "learning_rate": 7.3513513513513525e-06, | |
| "loss": 1.342, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2756756756756757, | |
| "grad_norm": 0.3033177852630615, | |
| "learning_rate": 7.297297297297298e-06, | |
| "loss": 1.3611, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2810810810810811, | |
| "grad_norm": 0.29411327838897705, | |
| "learning_rate": 7.243243243243244e-06, | |
| "loss": 1.3114, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2864864864864865, | |
| "grad_norm": 0.28862670063972473, | |
| "learning_rate": 7.189189189189189e-06, | |
| "loss": 1.3045, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2918918918918919, | |
| "grad_norm": 0.33967703580856323, | |
| "learning_rate": 7.135135135135136e-06, | |
| "loss": 1.2879, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2972972972972973, | |
| "grad_norm": 0.3063770532608032, | |
| "learning_rate": 7.0810810810810815e-06, | |
| "loss": 1.3237, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3027027027027027, | |
| "grad_norm": 0.2903819978237152, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 1.3088, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3081081081081081, | |
| "grad_norm": 0.2815365791320801, | |
| "learning_rate": 6.972972972972973e-06, | |
| "loss": 1.3184, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.31351351351351353, | |
| "grad_norm": 0.3042902946472168, | |
| "learning_rate": 6.91891891891892e-06, | |
| "loss": 1.3182, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.31891891891891894, | |
| "grad_norm": 0.27602043747901917, | |
| "learning_rate": 6.864864864864865e-06, | |
| "loss": 1.3051, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.32432432432432434, | |
| "grad_norm": 0.30109262466430664, | |
| "learning_rate": 6.810810810810811e-06, | |
| "loss": 1.3268, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.32972972972972975, | |
| "grad_norm": 0.27311018109321594, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 1.305, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.33513513513513515, | |
| "grad_norm": 0.2715848684310913, | |
| "learning_rate": 6.702702702702704e-06, | |
| "loss": 1.3034, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.34054054054054056, | |
| "grad_norm": 0.2581041157245636, | |
| "learning_rate": 6.648648648648649e-06, | |
| "loss": 1.2567, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.34594594594594597, | |
| "grad_norm": 0.26454904675483704, | |
| "learning_rate": 6.594594594594595e-06, | |
| "loss": 1.2792, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.35135135135135137, | |
| "grad_norm": 0.27279508113861084, | |
| "learning_rate": 6.540540540540541e-06, | |
| "loss": 1.2745, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.3567567567567568, | |
| "grad_norm": 0.27239367365837097, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 1.2947, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3621621621621622, | |
| "grad_norm": 0.2796631455421448, | |
| "learning_rate": 6.432432432432433e-06, | |
| "loss": 1.2705, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3675675675675676, | |
| "grad_norm": 0.266652911901474, | |
| "learning_rate": 6.378378378378379e-06, | |
| "loss": 1.2441, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.372972972972973, | |
| "grad_norm": 0.26300087571144104, | |
| "learning_rate": 6.324324324324325e-06, | |
| "loss": 1.259, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3783783783783784, | |
| "grad_norm": 0.3145560622215271, | |
| "learning_rate": 6.270270270270271e-06, | |
| "loss": 1.3417, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3837837837837838, | |
| "grad_norm": 0.2778513729572296, | |
| "learning_rate": 6.2162162162162164e-06, | |
| "loss": 1.2198, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3891891891891892, | |
| "grad_norm": 0.25896984338760376, | |
| "learning_rate": 6.162162162162163e-06, | |
| "loss": 1.231, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3945945945945946, | |
| "grad_norm": 0.2606508433818817, | |
| "learning_rate": 6.108108108108109e-06, | |
| "loss": 1.2799, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.2623414695262909, | |
| "learning_rate": 6.054054054054055e-06, | |
| "loss": 1.2269, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 0.26297539472579956, | |
| "learning_rate": 6e-06, | |
| "loss": 1.2751, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.41081081081081083, | |
| "grad_norm": 0.24481281638145447, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 1.2047, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.41621621621621624, | |
| "grad_norm": 0.2644273340702057, | |
| "learning_rate": 5.8918918918918924e-06, | |
| "loss": 1.2507, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.42162162162162165, | |
| "grad_norm": 0.2676522731781006, | |
| "learning_rate": 5.837837837837839e-06, | |
| "loss": 1.2374, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.42702702702702705, | |
| "grad_norm": 0.2541520297527313, | |
| "learning_rate": 5.783783783783784e-06, | |
| "loss": 1.2225, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.43243243243243246, | |
| "grad_norm": 0.252860426902771, | |
| "learning_rate": 5.729729729729731e-06, | |
| "loss": 1.225, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.43783783783783786, | |
| "grad_norm": 0.24671323597431183, | |
| "learning_rate": 5.675675675675676e-06, | |
| "loss": 1.1988, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.44324324324324327, | |
| "grad_norm": 0.25774669647216797, | |
| "learning_rate": 5.621621621621622e-06, | |
| "loss": 1.1917, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4486486486486487, | |
| "grad_norm": 0.2582065165042877, | |
| "learning_rate": 5.567567567567568e-06, | |
| "loss": 1.2049, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4540540540540541, | |
| "grad_norm": 0.2658644914627075, | |
| "learning_rate": 5.513513513513515e-06, | |
| "loss": 1.1864, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4594594594594595, | |
| "grad_norm": 0.2440362274646759, | |
| "learning_rate": 5.45945945945946e-06, | |
| "loss": 1.1772, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4648648648648649, | |
| "grad_norm": 0.2492218315601349, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 1.2137, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.4702702702702703, | |
| "grad_norm": 0.2513432204723358, | |
| "learning_rate": 5.351351351351351e-06, | |
| "loss": 1.2077, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4756756756756757, | |
| "grad_norm": 0.252913236618042, | |
| "learning_rate": 5.297297297297298e-06, | |
| "loss": 1.215, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.4810810810810811, | |
| "grad_norm": 0.24399417638778687, | |
| "learning_rate": 5.243243243243244e-06, | |
| "loss": 1.1977, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.4864864864864865, | |
| "grad_norm": 0.2553783059120178, | |
| "learning_rate": 5.18918918918919e-06, | |
| "loss": 1.2004, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4918918918918919, | |
| "grad_norm": 0.24357211589813232, | |
| "learning_rate": 5.135135135135135e-06, | |
| "loss": 1.1741, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4972972972972973, | |
| "grad_norm": 0.2634775638580322, | |
| "learning_rate": 5.081081081081082e-06, | |
| "loss": 1.1964, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5027027027027027, | |
| "grad_norm": 0.24633990228176117, | |
| "learning_rate": 5.027027027027027e-06, | |
| "loss": 1.1541, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5081081081081081, | |
| "grad_norm": 0.2492210417985916, | |
| "learning_rate": 4.9729729729729735e-06, | |
| "loss": 1.1663, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5135135135135135, | |
| "grad_norm": 0.24858078360557556, | |
| "learning_rate": 4.91891891891892e-06, | |
| "loss": 1.1684, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.518918918918919, | |
| "grad_norm": 0.24967968463897705, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 1.1695, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5243243243243243, | |
| "grad_norm": 0.24852266907691956, | |
| "learning_rate": 4.810810810810811e-06, | |
| "loss": 1.1529, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5297297297297298, | |
| "grad_norm": 0.24503152072429657, | |
| "learning_rate": 4.756756756756757e-06, | |
| "loss": 1.1689, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5351351351351351, | |
| "grad_norm": 0.25913184881210327, | |
| "learning_rate": 4.702702702702703e-06, | |
| "loss": 1.1888, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 0.23641978204250336, | |
| "learning_rate": 4.6486486486486495e-06, | |
| "loss": 1.1676, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5459459459459459, | |
| "grad_norm": 0.25850117206573486, | |
| "learning_rate": 4.594594594594596e-06, | |
| "loss": 1.1609, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5513513513513514, | |
| "grad_norm": 0.24710091948509216, | |
| "learning_rate": 4.540540540540541e-06, | |
| "loss": 1.1522, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5567567567567567, | |
| "grad_norm": 0.2579835057258606, | |
| "learning_rate": 4.486486486486487e-06, | |
| "loss": 1.1872, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5621621621621622, | |
| "grad_norm": 0.26577505469322205, | |
| "learning_rate": 4.432432432432433e-06, | |
| "loss": 1.1775, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5675675675675675, | |
| "grad_norm": 0.289943665266037, | |
| "learning_rate": 4.378378378378379e-06, | |
| "loss": 1.1849, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.572972972972973, | |
| "grad_norm": 0.29339632391929626, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 1.1607, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5783783783783784, | |
| "grad_norm": 0.24826812744140625, | |
| "learning_rate": 4.270270270270271e-06, | |
| "loss": 1.1834, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5837837837837838, | |
| "grad_norm": 0.24059158563613892, | |
| "learning_rate": 4.216216216216217e-06, | |
| "loss": 1.1611, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5891891891891892, | |
| "grad_norm": 0.2687944769859314, | |
| "learning_rate": 4.162162162162163e-06, | |
| "loss": 1.1515, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5945945945945946, | |
| "grad_norm": 0.27312615513801575, | |
| "learning_rate": 4.108108108108108e-06, | |
| "loss": 1.1971, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.27208036184310913, | |
| "learning_rate": 4.0540540540540545e-06, | |
| "loss": 1.1789, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6054054054054054, | |
| "grad_norm": 0.2820265591144562, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.1981, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6108108108108108, | |
| "grad_norm": 0.26071199774742126, | |
| "learning_rate": 3.945945945945947e-06, | |
| "loss": 1.1936, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6162162162162163, | |
| "grad_norm": 0.2583164870738983, | |
| "learning_rate": 3.891891891891892e-06, | |
| "loss": 1.1672, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6216216216216216, | |
| "grad_norm": 0.27570709586143494, | |
| "learning_rate": 3.837837837837838e-06, | |
| "loss": 1.2096, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6270270270270271, | |
| "grad_norm": 0.2592144012451172, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 1.169, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6324324324324324, | |
| "grad_norm": 0.26106804609298706, | |
| "learning_rate": 3.72972972972973e-06, | |
| "loss": 1.1912, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6378378378378379, | |
| "grad_norm": 0.25909966230392456, | |
| "learning_rate": 3.6756756756756763e-06, | |
| "loss": 1.1462, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6432432432432432, | |
| "grad_norm": 0.28418397903442383, | |
| "learning_rate": 3.621621621621622e-06, | |
| "loss": 1.1584, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6486486486486487, | |
| "grad_norm": 0.26269835233688354, | |
| "learning_rate": 3.567567567567568e-06, | |
| "loss": 1.1556, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.654054054054054, | |
| "grad_norm": 0.2741927206516266, | |
| "learning_rate": 3.513513513513514e-06, | |
| "loss": 1.1681, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6594594594594595, | |
| "grad_norm": 0.2586127519607544, | |
| "learning_rate": 3.45945945945946e-06, | |
| "loss": 1.1531, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6648648648648648, | |
| "grad_norm": 0.25531938672065735, | |
| "learning_rate": 3.4054054054054057e-06, | |
| "loss": 1.1636, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6702702702702703, | |
| "grad_norm": 0.2960667610168457, | |
| "learning_rate": 3.351351351351352e-06, | |
| "loss": 1.1818, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 0.31269460916519165, | |
| "learning_rate": 3.2972972972972976e-06, | |
| "loss": 1.2105, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6810810810810811, | |
| "grad_norm": 0.2489175945520401, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 1.1327, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.6864864864864865, | |
| "grad_norm": 0.27137380838394165, | |
| "learning_rate": 3.1891891891891894e-06, | |
| "loss": 1.1882, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6918918918918919, | |
| "grad_norm": 0.27693238854408264, | |
| "learning_rate": 3.1351351351351356e-06, | |
| "loss": 1.1574, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6972972972972973, | |
| "grad_norm": 0.26204511523246765, | |
| "learning_rate": 3.0810810810810817e-06, | |
| "loss": 1.1535, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7027027027027027, | |
| "grad_norm": 0.2716626822948456, | |
| "learning_rate": 3.0270270270270274e-06, | |
| "loss": 1.1572, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7081081081081081, | |
| "grad_norm": 0.2737910747528076, | |
| "learning_rate": 2.9729729729729736e-06, | |
| "loss": 1.1453, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7135135135135136, | |
| "grad_norm": 0.2846289873123169, | |
| "learning_rate": 2.9189189189189193e-06, | |
| "loss": 1.1371, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7189189189189189, | |
| "grad_norm": 0.2652347683906555, | |
| "learning_rate": 2.8648648648648654e-06, | |
| "loss": 1.1399, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7243243243243244, | |
| "grad_norm": 0.26175656914711, | |
| "learning_rate": 2.810810810810811e-06, | |
| "loss": 1.1691, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7297297297297297, | |
| "grad_norm": 0.274801641702652, | |
| "learning_rate": 2.7567567567567573e-06, | |
| "loss": 1.1563, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7351351351351352, | |
| "grad_norm": 0.2784542143344879, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 1.178, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7405405405405405, | |
| "grad_norm": 0.26470088958740234, | |
| "learning_rate": 2.648648648648649e-06, | |
| "loss": 1.1688, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.745945945945946, | |
| "grad_norm": 0.27295202016830444, | |
| "learning_rate": 2.594594594594595e-06, | |
| "loss": 1.1403, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7513513513513513, | |
| "grad_norm": 0.2557908892631531, | |
| "learning_rate": 2.540540540540541e-06, | |
| "loss": 1.1573, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7567567567567568, | |
| "grad_norm": 0.27424272894859314, | |
| "learning_rate": 2.4864864864864867e-06, | |
| "loss": 1.1423, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7621621621621621, | |
| "grad_norm": 0.2811051607131958, | |
| "learning_rate": 2.432432432432433e-06, | |
| "loss": 1.1549, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7675675675675676, | |
| "grad_norm": 0.25805169343948364, | |
| "learning_rate": 2.3783783783783786e-06, | |
| "loss": 1.1486, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.772972972972973, | |
| "grad_norm": 0.26995599269866943, | |
| "learning_rate": 2.3243243243243247e-06, | |
| "loss": 1.1357, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.7783783783783784, | |
| "grad_norm": 0.2542300224304199, | |
| "learning_rate": 2.2702702702702705e-06, | |
| "loss": 1.1208, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7837837837837838, | |
| "grad_norm": 0.24706785380840302, | |
| "learning_rate": 2.2162162162162166e-06, | |
| "loss": 1.1324, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7891891891891892, | |
| "grad_norm": 0.2644682824611664, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 1.121, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.7945945945945946, | |
| "grad_norm": 0.2919802665710449, | |
| "learning_rate": 2.1081081081081085e-06, | |
| "loss": 1.1779, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.2681489586830139, | |
| "learning_rate": 2.054054054054054e-06, | |
| "loss": 1.1651, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8054054054054054, | |
| "grad_norm": 0.24865786731243134, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1272, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 0.26049530506134033, | |
| "learning_rate": 1.945945945945946e-06, | |
| "loss": 1.1299, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8162162162162162, | |
| "grad_norm": 0.2903294265270233, | |
| "learning_rate": 1.8918918918918922e-06, | |
| "loss": 1.1751, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8216216216216217, | |
| "grad_norm": 0.274706095457077, | |
| "learning_rate": 1.8378378378378381e-06, | |
| "loss": 1.1493, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.827027027027027, | |
| "grad_norm": 0.2549731433391571, | |
| "learning_rate": 1.783783783783784e-06, | |
| "loss": 1.1345, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8324324324324325, | |
| "grad_norm": 0.26216205954551697, | |
| "learning_rate": 1.72972972972973e-06, | |
| "loss": 1.1316, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8378378378378378, | |
| "grad_norm": 0.26119425892829895, | |
| "learning_rate": 1.675675675675676e-06, | |
| "loss": 1.132, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8432432432432433, | |
| "grad_norm": 0.26794594526290894, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 1.1532, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8486486486486486, | |
| "grad_norm": 0.2842087745666504, | |
| "learning_rate": 1.5675675675675678e-06, | |
| "loss": 1.0995, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8540540540540541, | |
| "grad_norm": 0.26724207401275635, | |
| "learning_rate": 1.5135135135135137e-06, | |
| "loss": 1.102, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8594594594594595, | |
| "grad_norm": 0.2525201737880707, | |
| "learning_rate": 1.4594594594594596e-06, | |
| "loss": 1.095, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8648648648648649, | |
| "grad_norm": 0.2626565098762512, | |
| "learning_rate": 1.4054054054054056e-06, | |
| "loss": 1.1435, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8702702702702703, | |
| "grad_norm": 0.2600706219673157, | |
| "learning_rate": 1.3513513513513515e-06, | |
| "loss": 1.1316, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.8756756756756757, | |
| "grad_norm": 0.2661210596561432, | |
| "learning_rate": 1.2972972972972974e-06, | |
| "loss": 1.1423, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.8810810810810811, | |
| "grad_norm": 0.2888488173484802, | |
| "learning_rate": 1.2432432432432434e-06, | |
| "loss": 1.1384, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.8864864864864865, | |
| "grad_norm": 0.25049468874931335, | |
| "learning_rate": 1.1891891891891893e-06, | |
| "loss": 1.1257, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.8918918918918919, | |
| "grad_norm": 0.2596016526222229, | |
| "learning_rate": 1.1351351351351352e-06, | |
| "loss": 1.139, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8972972972972973, | |
| "grad_norm": 0.26817384362220764, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 1.1399, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9027027027027027, | |
| "grad_norm": 0.27717313170433044, | |
| "learning_rate": 1.027027027027027e-06, | |
| "loss": 1.1356, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9081081081081082, | |
| "grad_norm": 0.27815675735473633, | |
| "learning_rate": 9.72972972972973e-07, | |
| "loss": 1.0882, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9135135135135135, | |
| "grad_norm": 0.26827895641326904, | |
| "learning_rate": 9.189189189189191e-07, | |
| "loss": 1.1382, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.918918918918919, | |
| "grad_norm": 0.26916950941085815, | |
| "learning_rate": 8.64864864864865e-07, | |
| "loss": 1.1633, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9243243243243243, | |
| "grad_norm": 0.2528124451637268, | |
| "learning_rate": 8.108108108108109e-07, | |
| "loss": 1.13, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9297297297297298, | |
| "grad_norm": 0.2664795219898224, | |
| "learning_rate": 7.567567567567569e-07, | |
| "loss": 1.1121, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9351351351351351, | |
| "grad_norm": 0.257194846868515, | |
| "learning_rate": 7.027027027027028e-07, | |
| "loss": 1.1416, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9405405405405406, | |
| "grad_norm": 0.2620390057563782, | |
| "learning_rate": 6.486486486486487e-07, | |
| "loss": 1.1354, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 0.25638052821159363, | |
| "learning_rate": 5.945945945945947e-07, | |
| "loss": 1.15, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9513513513513514, | |
| "grad_norm": 0.25654318928718567, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 1.114, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9567567567567568, | |
| "grad_norm": 0.3269580006599426, | |
| "learning_rate": 4.864864864864865e-07, | |
| "loss": 1.1254, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9621621621621622, | |
| "grad_norm": 0.26131343841552734, | |
| "learning_rate": 4.324324324324325e-07, | |
| "loss": 1.0901, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.9675675675675676, | |
| "grad_norm": 0.2583537697792053, | |
| "learning_rate": 3.7837837837837843e-07, | |
| "loss": 1.1467, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.972972972972973, | |
| "grad_norm": 0.2696709930896759, | |
| "learning_rate": 3.2432432432432436e-07, | |
| "loss": 1.1198, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9783783783783784, | |
| "grad_norm": 0.2895946502685547, | |
| "learning_rate": 2.702702702702703e-07, | |
| "loss": 1.1572, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.9837837837837838, | |
| "grad_norm": 0.25485262274742126, | |
| "learning_rate": 2.1621621621621625e-07, | |
| "loss": 1.152, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.9891891891891892, | |
| "grad_norm": 0.262500524520874, | |
| "learning_rate": 1.6216216216216218e-07, | |
| "loss": 1.1578, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.9945945945945946, | |
| "grad_norm": 0.26408693194389343, | |
| "learning_rate": 1.0810810810810812e-07, | |
| "loss": 1.1164, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2597890794277191, | |
| "learning_rate": 5.405405405405406e-08, | |
| "loss": 1.1132, | |
| "step": 185 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 185, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.3279459925898035e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |