| { |
| "best_metric": 0.49116000533103943, |
| "best_model_checkpoint": "experts/mistralic-expert-6/checkpoint-800", |
| "epoch": 0.21932830705962988, |
| "eval_steps": 200, |
| "global_step": 800, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6994, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8302, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4811, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5604, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6531, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5385, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4783, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6369, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5879, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5295, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5714, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.8482, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.669, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5611, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4365, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3577, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5298, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6154, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5656, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.9267, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4734, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5643, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3862, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5253, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5414, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5773, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.378, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4835, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5724, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.499, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4688, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3042, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3976, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5073, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.515, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5217, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4707, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6927, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5852, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4088, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4264, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.736, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6521, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4341, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.382, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4791, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.7997, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3342, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5161, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.6684, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3646, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5941, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6108, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5637, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5081, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4254, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6792, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.617, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7451, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5197, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4747, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7081, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6431, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3876, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6011, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.351, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4798, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4795, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6627, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5638, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4911, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5154, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6197, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6435, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5543, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.409, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7722, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4525, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3458, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6547, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6557, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4964, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5434, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.7047, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6133, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.8532, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.59, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.6374, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6089, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6657, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4013, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4148, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5466, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4535, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5942, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7103, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5085, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7082, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5191, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5846, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4225, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5228, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3559, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4337, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4334, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4864, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.544, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4475, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4525, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.7092, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5131, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6282, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4774, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.8279, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.386, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.441, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3907, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4543, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5451, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3776, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6798, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.6153, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.474, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4728, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3922, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.465, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4588, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5072, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3563, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6652, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5841, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6905, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4146, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4482, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5262, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3658, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6413, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7497, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5407, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4794, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.7326, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4932, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5621, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4034, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.522, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5849, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3201, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5649, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5546, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5961, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5517, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3701, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.529, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5147, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4859, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5261, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6766, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4343, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5389, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6087, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4613, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6619, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.8939, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.518, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3756, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4739, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5381, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.581, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6024, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4306, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5632, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5424, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6596, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5767, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5615, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6309, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5179, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6533, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4672, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5559, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4247, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.471, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4948, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4501, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.7633, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3766, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3518, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6723, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5156, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4782, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5309, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4495, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.544, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4744, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.6445, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4907, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.05, |
| "eval_loss": 0.515356183052063, |
| "eval_runtime": 151.1205, |
| "eval_samples_per_second": 6.617, |
| "eval_steps_per_second": 3.309, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.05, |
| "mmlu_eval_accuracy": 0.6052520883458705, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.96, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.37, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.3635515544184194, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5532, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5743, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3798, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5082, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5915, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4404, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6578, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5879, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5568, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5765, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5987, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4298, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4531, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4482, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5302, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4468, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5594, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.7247, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6556, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4433, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6161, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4172, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5306, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6016, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4863, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4464, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5199, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4004, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.486, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6632, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5232, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5558, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5069, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6969, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5236, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5968, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4871, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6781, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5094, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4382, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4039, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5455, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.408, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3586, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4394, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5684, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4283, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.667, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.6325, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4068, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5148, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4078, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4243, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5693, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5459, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3619, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2712, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3858, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5059, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7991, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3928, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3715, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4377, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4248, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3999, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.7594, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.536, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3598, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4725, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4488, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4896, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5649, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 1.0017, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.493, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5391, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4265, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3358, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3607, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6027, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4673, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4951, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4463, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6422, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8212, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5748, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.335, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.433, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6139, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5622, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3532, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4295, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3865, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4471, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6755, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.8007, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5947, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6367, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5283, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5305, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3633, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7916, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.491, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3126, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.6352, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5478, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4332, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4126, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.7758, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4927, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4095, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4575, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5186, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4996, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5662, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4463, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5485, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4455, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5777, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4782, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6597, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.6666, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3679, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4848, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4259, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4203, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4307, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7558, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4322, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5617, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7348, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5441, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4523, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7244, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3556, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5941, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3755, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4598, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4438, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.7113, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5306, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5786, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8635, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4195, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.8005, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4222, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7129, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5694, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5516, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6283, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5407, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3771, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4404, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7389, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4725, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3962, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5303, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4354, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4631, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5122, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3922, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.413, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6075, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4615, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3987, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3374, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5237, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5015, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.377, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.407, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.7536, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6903, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.6078, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.419, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5108, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3322, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.766, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4698, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4709, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4441, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5529, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.9117, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3571, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6611, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5024, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4579, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4244, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5683, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4138, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7081, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4609, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3667, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4574, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.406, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5737, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4219, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5496, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "eval_loss": 0.504305899143219, |
| "eval_runtime": 150.9201, |
| "eval_samples_per_second": 6.626, |
| "eval_steps_per_second": 3.313, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "mmlu_eval_accuracy": 0.5962314830963265, |
| "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.37058823529411766, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.5198283232067646, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.7203, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6216, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3669, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4154, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4745, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5108, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3767, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3694, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4148, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.41, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.6553, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4297, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.361, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5838, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.558, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3952, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5075, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5347, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3776, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6025, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3818, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4756, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5455, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4308, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5763, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3749, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.8519, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5916, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.549, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.9338, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5916, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4029, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4253, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4892, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4926, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3972, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.49, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4434, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3448, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4215, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4378, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5748, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4101, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4322, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.521, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3289, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.665, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5222, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3904, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5349, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6914, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.489, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5752, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.472, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4699, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3475, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.568, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5797, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4502, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6758, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3968, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5402, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.561, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4116, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7404, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2737, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4121, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5407, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4883, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.443, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4444, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4305, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4499, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3891, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5665, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4387, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4408, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4695, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.7828, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4792, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.415, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6478, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3864, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3394, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3196, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.378, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5638, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4954, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3348, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5741, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4474, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4204, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5899, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4217, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4679, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5714, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5085, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6354, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5042, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3519, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.598, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.531, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4191, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4318, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7143, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5497, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4538, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7157, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6946, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.249, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5324, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3812, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.8261, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7036, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4518, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5628, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4108, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4973, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4102, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4631, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.7448, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4011, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5605, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5694, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3686, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3842, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4278, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6496, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3467, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4131, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4686, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5447, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4391, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4717, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4715, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5484, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.48, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.6417, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5614, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3814, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.486, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4974, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4136, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.2957, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.353, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.347, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3728, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5899, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5083, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3103, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4695, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4578, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5379, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4456, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4886, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4037, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4171, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.7318, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5345, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5906, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3113, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3796, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.491, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6367, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5304, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4946, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.357, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2836, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4036, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4874, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6752, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5289, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4872, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6771, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5498, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4336, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.51, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6332, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6271, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.354, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4489, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.337, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4045, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2503, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.37, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6919, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4756, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2652, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5119, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4497, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5773, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6006, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5529, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3411, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4447, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4417, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5952, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_loss": 0.4948292374610901, |
| "eval_runtime": 151.0377, |
| "eval_samples_per_second": 6.621, |
| "eval_steps_per_second": 3.31, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.16, |
| "mmlu_eval_accuracy": 0.5936753977956382, |
| "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.4, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.3888888888888889, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.2675891808056645, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.567, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3148, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4119, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.7136, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.619, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3971, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4503, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4632, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3975, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.433, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.8727, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6681, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3808, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6533, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6251, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.547, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4395, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4786, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5755, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.492, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4931, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3624, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.458, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3164, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4457, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6502, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6553, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5192, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6162, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6068, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.639, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5272, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6829, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6279, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3735, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6645, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3801, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4048, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6524, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5326, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5401, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5767, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4564, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4885, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5471, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5977, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5148, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.368, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4038, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3482, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5143, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3521, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5433, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.7291, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6005, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3673, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6243, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.371, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6469, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.6202, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4989, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5338, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4753, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5331, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4077, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4434, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3566, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4665, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4379, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3503, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4873, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4437, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3753, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6785, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4846, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5105, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7299, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7591, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7995, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4812, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7257, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.46, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4938, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4368, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5902, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3429, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3912, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4183, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5885, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4157, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3395, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5497, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5296, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.7148, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6782, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6738, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5682, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.463, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5401, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5509, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5193, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6301, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5211, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.6558, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.435, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3486, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.529, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3684, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.734, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4505, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5651, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3606, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.41, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4431, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3383, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4261, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4638, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.518, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4139, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4933, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3796, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3857, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4463, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3225, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.474, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.379, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4629, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4682, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5005, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5222, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.8757, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4481, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4092, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.366, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.69, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3852, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6261, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4176, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.7555, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5731, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4146, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4825, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3533, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.6229, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5118, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6279, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4771, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4081, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3866, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4032, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4952, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7429, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4653, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4293, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4435, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5746, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5019, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3924, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5731, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6708, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4854, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4314, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4307, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.6679, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3449, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4143, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5017, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4705, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4359, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4688, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5323, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3918, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4635, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.7891, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4755, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.303, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4352, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4531, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5802, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.549, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5076, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.499, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6112, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4021, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.465, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4729, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5338, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5243, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4216, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.646, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4429, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5621, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4695, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4384, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3778, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6291, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5068, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4221, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.22, |
| "eval_loss": 0.49116000533103943, |
| "eval_runtime": 151.0794, |
| "eval_samples_per_second": 6.619, |
| "eval_steps_per_second": 3.31, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.22, |
| "mmlu_eval_accuracy": 0.5989241164363682, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 1.0, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.92, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.6, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.3764705882352941, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.3888888888888889, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.4367162143281484, |
| "step": 800 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 10941, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 1.884474013536092e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
