| { |
| "best_metric": 0.3990638256072998, |
| "best_model_checkpoint": "experts/mistralic-expert-10/checkpoint-800", |
| "epoch": 0.30191527502594584, |
| "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": 0.4687, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4379, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4031, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6294, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3844, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4651, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4077, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5852, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6446, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4735, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.558, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3036, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4698, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4017, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3569, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4597, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5558, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5634, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5111, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3807, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4366, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3173, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5166, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.366, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5089, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4725, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4866, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4711, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3939, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.511, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3598, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3905, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4141, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4236, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5242, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4163, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.5281, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4195, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.546, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3423, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4877, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4921, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.501, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4801, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3017, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4624, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.387, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4138, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5153, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5187, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5212, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4381, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4194, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.341, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4823, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4429, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4283, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.436, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3699, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5406, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.5063, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3374, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4852, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 0.0002, |
| "loss": 0.4239, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4104, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4043, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.349, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4591, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4908, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5579, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3437, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5482, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4154, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5933, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3932, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4363, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4405, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3178, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3877, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5094, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4116, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4241, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3943, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4581, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4228, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3035, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4152, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.5081, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4523, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 0.0002, |
| "loss": 0.449, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.557, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.45, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4236, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4346, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3602, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5046, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5019, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4926, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4206, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3776, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5444, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.6434, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4859, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4262, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3652, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5247, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3739, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3502, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4463, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4473, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5138, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4755, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3593, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4006, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.656, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.5208, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 0.0002, |
| "loss": 0.397, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4043, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3414, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3716, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4867, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.399, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3143, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2965, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3126, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3965, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.367, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.434, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4158, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5169, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4491, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4397, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5547, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3094, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5779, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3973, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4148, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5891, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4738, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3647, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.4316, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3863, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3947, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3662, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.445, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.291, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3594, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4337, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3512, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4868, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.315, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4631, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.324, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5954, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4303, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3618, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3946, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2777, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.6499, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2976, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4635, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4091, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4171, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3077, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5025, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3307, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3656, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4754, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 0.0002, |
| "loss": 0.5165, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3207, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4359, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4249, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3493, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4893, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5003, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.373, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3351, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.35, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3593, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3318, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4867, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4948, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3646, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3675, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3392, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4022, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5323, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3873, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3812, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4918, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3578, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4789, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3591, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2973, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 0.0002, |
| "loss": 0.5745, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4953, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2947, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_loss": 0.42375433444976807, |
| "eval_runtime": 109.1395, |
| "eval_samples_per_second": 9.163, |
| "eval_steps_per_second": 4.581, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08, |
| "mmlu_eval_accuracy": 0.5992560475314558, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, |
| "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": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.6875, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "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": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "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": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.9230769230769231, |
| "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": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.33, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.3588869174451181, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4284, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5111, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.422, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4328, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4116, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.439, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.464, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3798, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3916, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2922, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.327, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.335, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3975, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2875, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4427, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5155, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3267, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4556, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4974, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4961, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.5042, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3421, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3941, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.439, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3627, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3963, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4182, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3124, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4372, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4958, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4551, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3301, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4211, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3774, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4033, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4546, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5565, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3413, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3859, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3295, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4185, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3187, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2488, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3094, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3372, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3628, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3314, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4823, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3295, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4211, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4578, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4093, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4194, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2662, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4565, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4378, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4021, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3133, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3951, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.331, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5553, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4799, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3909, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3802, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4322, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3896, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4574, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.353, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3028, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4246, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3687, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4603, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4508, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.5132, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3348, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.462, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3305, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.308, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4862, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5406, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3895, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4596, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4366, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3638, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4276, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5693, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4833, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4576, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5377, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5118, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.491, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3107, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4244, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4884, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4563, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5308, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.345, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3633, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4228, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4925, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5263, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3454, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4619, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3784, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4197, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4154, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4341, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5438, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3942, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4794, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4836, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2955, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3971, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3234, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5278, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4182, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4185, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4444, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3645, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3335, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3742, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3214, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.5141, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6287, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.443, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.4362, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3884, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3412, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 0.0002, |
| "loss": 0.6665, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4028, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4603, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4931, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3631, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4734, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6062, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3115, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4169, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4635, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4725, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.6336, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4462, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4844, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4525, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4264, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4249, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.321, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5316, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.426, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.5421, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.409, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3446, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3854, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2969, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4246, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5349, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4701, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.6272, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3358, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4276, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.706, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4013, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4539, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.267, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3943, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3842, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3887, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.387, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.5084, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3644, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4457, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4365, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4332, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3126, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.2798, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4693, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4337, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4098, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.3905, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.398, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4125, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 0.0002, |
| "loss": 0.4, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3486, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3711, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4269, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.629, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3242, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3457, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3969, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3255, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5122, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3144, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.413, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3475, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4273, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3019, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4705, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4843, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.15, |
| "eval_loss": 0.4129827618598938, |
| "eval_runtime": 109.106, |
| "eval_samples_per_second": 9.165, |
| "eval_steps_per_second": 4.583, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.15, |
| "mmlu_eval_accuracy": 0.5854928379085984, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.85, |
| "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": 0.9230769230769231, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "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": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.4235294117647059, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.7037037037037037, |
| "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": 0.8421052631578947, |
| "mmlu_loss": 1.5015099039666024, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4542, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3542, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4137, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.5194, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3564, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.3304, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4207, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4146, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4027, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 0.0002, |
| "loss": 0.4379, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4521, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3437, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3466, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3606, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4335, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3926, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4687, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4833, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.461, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5032, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.5697, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3396, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4629, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2868, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4346, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2989, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3497, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4435, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3108, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4014, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4556, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.2861, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3835, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.3564, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.6947, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 0.0002, |
| "loss": 0.4343, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.2947, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.463, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3497, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4365, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3461, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3665, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4413, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4485, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.2636, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3708, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3636, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4146, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.423, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4013, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.6993, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5807, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3912, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3059, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4265, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.4059, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3506, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.3441, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.5204, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.396, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 0.0002, |
| "loss": 0.2828, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.347, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3973, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3309, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5189, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4553, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4905, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3281, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.382, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.351, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.347, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4184, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3861, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4587, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.5241, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.2735, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4922, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4693, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3068, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4041, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.418, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4507, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3222, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4201, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3447, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3722, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.4043, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 0.0002, |
| "loss": 0.3941, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3056, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5213, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3598, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3764, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3994, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4477, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3355, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4877, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4088, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3827, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.5035, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4557, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3671, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3613, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.336, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.511, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.2907, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3733, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3576, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3107, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3299, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3523, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4687, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3858, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.3857, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 0.0002, |
| "loss": 0.4025, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4319, |
| "step": 517 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3826, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4093, |
| "step": 519 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3488, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4205, |
| "step": 521 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.2814, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.338, |
| "step": 523 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4434, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4188, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3765, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3246, |
| "step": 527 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4121, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3877, |
| "step": 529 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4116, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4427, |
| "step": 531 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5231, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3964, |
| "step": 533 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3822, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.472, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4829, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.461, |
| "step": 537 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.4329, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3283, |
| "step": 539 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.5144, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3756, |
| "step": 541 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.2996, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 0.0002, |
| "loss": 0.3402, |
| "step": 543 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.353, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.2992, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5111, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4229, |
| "step": 547 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3738, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5171, |
| "step": 549 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3569, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4096, |
| "step": 551 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3479, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3319, |
| "step": 553 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3309, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.2845, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4409, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5645, |
| "step": 557 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3732, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 559 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.2512, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4193, |
| "step": 561 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5557, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.5478, |
| "step": 563 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4448, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3717, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3647, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.2575, |
| "step": 567 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.3857, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 0.0002, |
| "loss": 0.4617, |
| "step": 569 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.382, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4205, |
| "step": 571 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3222, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3935, |
| "step": 573 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.6771, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3823, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3557, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3503, |
| "step": 577 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4304, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4304, |
| "step": 579 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3569, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4017, |
| "step": 581 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.2874, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3969, |
| "step": 583 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3979, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3535, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5506, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.4974, |
| "step": 587 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.465, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5877, |
| "step": 589 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3663, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3709, |
| "step": 591 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3023, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3651, |
| "step": 593 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.359, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.3255, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 0.0002, |
| "loss": 0.5587, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3171, |
| "step": 597 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3757, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4422, |
| "step": 599 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4832, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.23, |
| "eval_loss": 0.4104755222797394, |
| "eval_runtime": 109.1763, |
| "eval_samples_per_second": 9.159, |
| "eval_steps_per_second": 4.58, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.23, |
| "mmlu_eval_accuracy": 0.600983525298798, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.8125, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.7272727272727273, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.75, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "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": 0.6666666666666666, |
| "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": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "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": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "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": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, |
| "mmlu_eval_accuracy_professional_law": 0.4235294117647059, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.7727272727272727, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.3696848112827182, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3511, |
| "step": 601 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4113, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4181, |
| "step": 603 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3392, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3531, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3512, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3588, |
| "step": 607 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.474, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4339, |
| "step": 609 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4195, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4534, |
| "step": 611 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4335, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3466, |
| "step": 613 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4509, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3138, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4925, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3663, |
| "step": 617 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3389, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3356, |
| "step": 619 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3412, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.3135, |
| "step": 621 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 0.0002, |
| "loss": 0.4075, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.2261, |
| "step": 623 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5887, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3992, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.2996, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4591, |
| "step": 627 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4546, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3476, |
| "step": 629 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.2549, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3999, |
| "step": 631 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4251, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.519, |
| "step": 633 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.287, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4353, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3921, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4601, |
| "step": 637 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3948, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3871, |
| "step": 639 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3257, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4415, |
| "step": 641 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.5713, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.2481, |
| "step": 643 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.2991, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3199, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3172, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3125, |
| "step": 647 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.3865, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 0.0002, |
| "loss": 0.4205, |
| "step": 649 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4262, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4112, |
| "step": 651 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3032, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3647, |
| "step": 653 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3271, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4705, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3931, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.299, |
| "step": 657 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4335, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4908, |
| "step": 659 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3952, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4293, |
| "step": 661 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3744, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4025, |
| "step": 663 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3698, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3991, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4355, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.2599, |
| "step": 667 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4399, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3505, |
| "step": 669 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3391, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4293, |
| "step": 671 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3365, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.3957, |
| "step": 673 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.4188, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 0.0002, |
| "loss": 0.2761, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4095, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3873, |
| "step": 677 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.2936, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.383, |
| "step": 679 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.2176, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4014, |
| "step": 681 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3463, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3736, |
| "step": 683 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3639, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3512, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3511, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3433, |
| "step": 687 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3513, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4656, |
| "step": 689 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4801, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3189, |
| "step": 691 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.2969, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3249, |
| "step": 693 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4743, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3823, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3689, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.31, |
| "step": 697 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.504, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4326, |
| "step": 699 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4856, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.4571, |
| "step": 701 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 0.0002, |
| "loss": 0.3979, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3194, |
| "step": 703 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3403, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3805, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3984, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3481, |
| "step": 707 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3943, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4529, |
| "step": 709 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3888, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3332, |
| "step": 711 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4018, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3563, |
| "step": 713 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3706, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.5292, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3671, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.443, |
| "step": 717 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4454, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4584, |
| "step": 719 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4031, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3653, |
| "step": 721 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3394, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3004, |
| "step": 723 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.4038, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3076, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.3266, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6202, |
| "step": 727 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 0.0002, |
| "loss": 0.6375, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4881, |
| "step": 729 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4809, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3508, |
| "step": 731 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.2977, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3066, |
| "step": 733 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.2465, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4051, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5078, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3794, |
| "step": 737 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3655, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.41, |
| "step": 739 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.28, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.6678, |
| "step": 741 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3748, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4358, |
| "step": 743 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.2733, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3137, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3555, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5012, |
| "step": 747 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4232, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3369, |
| "step": 749 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5902, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4061, |
| "step": 751 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.4386, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.3959, |
| "step": 753 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.5717, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 0.0002, |
| "loss": 0.2406, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3643, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3262, |
| "step": 757 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3753, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4438, |
| "step": 759 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3201, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4942, |
| "step": 761 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3739, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3892, |
| "step": 763 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4161, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4042, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5629, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3621, |
| "step": 767 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3779, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3547, |
| "step": 769 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3077, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3511, |
| "step": 771 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.441, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5022, |
| "step": 773 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4117, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.2836, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5293, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.361, |
| "step": 777 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.5475, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4244, |
| "step": 779 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.3333, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 0.0002, |
| "loss": 0.4097, |
| "step": 781 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5082, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3192, |
| "step": 783 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5246, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.4406, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3497, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.4233, |
| "step": 787 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3693, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5512, |
| "step": 789 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3967, |
| "step": 791 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3856, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3286, |
| "step": 793 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.4823, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3191, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3015, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3961, |
| "step": 797 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.4248, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3794, |
| "step": 799 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3873, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.3, |
| "eval_loss": 0.3990638256072998, |
| "eval_runtime": 109.2651, |
| "eval_samples_per_second": 9.152, |
| "eval_steps_per_second": 4.576, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.3, |
| "mmlu_eval_accuracy": 0.604493892708381, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.75, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, |
| "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": 0.38461538461538464, |
| "mmlu_eval_accuracy_econometrics": 0.5833333333333334, |
| "mmlu_eval_accuracy_electrical_engineering": 0.6875, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.5, |
| "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": 0.6511627906976745, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "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": 0.6923076923076923, |
| "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": 0.6363636363636364, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.38, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.7037037037037037, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.5750128679736164, |
| "step": 800 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 7947, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 1.17826604566315e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
