| { |
| "best_metric": 0.38303717970848083, |
| "best_model_checkpoint": "experts/mistralic-expert-10/checkpoint-2600", |
| "epoch": 1.1321822813472968, |
| "eval_steps": 200, |
| "global_step": 3000, |
| "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 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.425, |
| "step": 801 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3845, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3733, |
| "step": 803 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.363, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3793, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.4721, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.5167, |
| "step": 807 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 0.0002, |
| "loss": 0.3505, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3491, |
| "step": 809 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4903, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3446, |
| "step": 811 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4438, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4073, |
| "step": 813 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.5825, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.2935, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3007, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4453, |
| "step": 817 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.615, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3628, |
| "step": 819 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3832, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4613, |
| "step": 821 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3757, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3734, |
| "step": 823 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3841, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3344, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3179, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.367, |
| "step": 827 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4955, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.8231, |
| "step": 829 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.2779, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4862, |
| "step": 831 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.2932, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.3601, |
| "step": 833 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 0.0002, |
| "loss": 0.4496, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4266, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4759, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6074, |
| "step": 837 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3159, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5017, |
| "step": 839 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4288, |
| "step": 841 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3236, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4802, |
| "step": 843 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4332, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3034, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.6039, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4708, |
| "step": 847 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3796, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3338, |
| "step": 849 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.2865, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5219, |
| "step": 851 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3998, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3432, |
| "step": 853 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3786, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3547, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.5127, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.2712, |
| "step": 857 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.3093, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.4039, |
| "step": 859 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.2482, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 0.0002, |
| "loss": 0.423, |
| "step": 861 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4174, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3853, |
| "step": 863 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4129, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.303, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.2835, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3552, |
| "step": 867 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4084, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.437, |
| "step": 869 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4588, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4266, |
| "step": 871 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3057, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3338, |
| "step": 873 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.429, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.2689, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3263, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3151, |
| "step": 877 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4315, |
| "step": 879 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.261, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.3729, |
| "step": 881 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4439, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4802, |
| "step": 883 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4456, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4012, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.2748, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 0.0002, |
| "loss": 0.4194, |
| "step": 887 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.335, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3426, |
| "step": 889 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3072, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.2908, |
| "step": 891 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3087, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.332, |
| "step": 893 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3327, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3398, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4792, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4321, |
| "step": 897 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3755, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4387, |
| "step": 899 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3636, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3296, |
| "step": 901 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3695, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3578, |
| "step": 903 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4312, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4787, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4067, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4523, |
| "step": 907 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4462, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3501, |
| "step": 909 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.4253, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3555, |
| "step": 911 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3581, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3853, |
| "step": 913 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 0.0002, |
| "loss": 0.3583, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3403, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.51, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.46, |
| "step": 917 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4004, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4262, |
| "step": 919 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3839, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3947, |
| "step": 921 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3703, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6606, |
| "step": 923 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4531, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4601, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4773, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4233, |
| "step": 927 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3726, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3259, |
| "step": 929 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3967, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3504, |
| "step": 931 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5648, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4654, |
| "step": 933 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5016, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4984, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.5063, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4057, |
| "step": 937 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.4189, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.3828, |
| "step": 939 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 0.0002, |
| "loss": 0.6808, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.2556, |
| "step": 941 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4098, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3492, |
| "step": 943 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3887, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4776, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3785, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.6116, |
| "step": 947 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3922, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5556, |
| "step": 949 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3638, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3341, |
| "step": 951 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.353, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3781, |
| "step": 953 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3228, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.381, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4504, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4482, |
| "step": 957 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3845, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4484, |
| "step": 959 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4075, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.3756, |
| "step": 961 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4157, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.461, |
| "step": 963 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.357, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4205, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.5231, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 0.0002, |
| "loss": 0.4422, |
| "step": 967 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4837, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4876, |
| "step": 969 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5502, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4151, |
| "step": 971 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3353, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5101, |
| "step": 973 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4383, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4221, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3016, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3894, |
| "step": 977 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4364, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3423, |
| "step": 979 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4575, |
| "step": 981 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4185, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 983 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4252, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4817, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.351, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.5212, |
| "step": 987 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.2992, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4371, |
| "step": 989 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4557, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4786, |
| "step": 991 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.4304, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 0.0002, |
| "loss": 0.2508, |
| "step": 993 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4242, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3587, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3439, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5542, |
| "step": 997 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4386, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 999 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5344, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.38, |
| "eval_loss": 0.3985953629016876, |
| "eval_runtime": 109.3022, |
| "eval_samples_per_second": 9.149, |
| "eval_steps_per_second": 4.574, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.38, |
| "mmlu_eval_accuracy": 0.5906332341994553, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "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": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "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": 0.5, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "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": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, |
| "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": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.45, |
| "mmlu_eval_accuracy_nutrition": 0.8181818181818182, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "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": 0.5833333333333334, |
| "mmlu_eval_accuracy_security_studies": 0.7037037037037037, |
| "mmlu_eval_accuracy_sociology": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.2684316527361035, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3472, |
| "step": 1001 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3951, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5367, |
| "step": 1003 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4711, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3632, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3761, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.353, |
| "step": 1007 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.3562, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.426, |
| "step": 1009 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.2956, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.2916, |
| "step": 1011 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5264, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6251, |
| "step": 1013 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.6728, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4962, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.5424, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.364, |
| "step": 1017 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4591, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.4597, |
| "step": 1019 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 0.0002, |
| "loss": 0.462, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3822, |
| "step": 1021 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3633, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4117, |
| "step": 1023 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.2984, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4655, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.467, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3636, |
| "step": 1027 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3043, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3143, |
| "step": 1029 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4934, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4891, |
| "step": 1031 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4264, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4359, |
| "step": 1033 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3438, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3003, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4805, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4374, |
| "step": 1037 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4357, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.3698, |
| "step": 1039 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.2758, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.5297, |
| "step": 1041 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4744, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.2926, |
| "step": 1043 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4878, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.2513, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 0.0002, |
| "loss": 0.4637, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5314, |
| "step": 1047 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5167, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.5169, |
| "step": 1049 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3331, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4777, |
| "step": 1051 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3089, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3226, |
| "step": 1053 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3865, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.415, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3107, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4346, |
| "step": 1057 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3714, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.414, |
| "step": 1059 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3913, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4632, |
| "step": 1061 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4461, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4552, |
| "step": 1063 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4575, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.395, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4418, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3859, |
| "step": 1067 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.2647, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4007, |
| "step": 1069 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3113, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3398, |
| "step": 1071 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.4337, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 0.0002, |
| "loss": 0.3307, |
| "step": 1073 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.4367, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5062, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3454, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.496, |
| "step": 1077 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.4928, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.2618, |
| "step": 1079 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3987, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3463, |
| "step": 1081 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3342, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3093, |
| "step": 1083 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3494, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3101, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3641, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.2831, |
| "step": 1087 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3778, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.2679, |
| "step": 1089 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3363, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3168, |
| "step": 1091 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3907, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.345, |
| "step": 1093 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3473, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.486, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.2929, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.2974, |
| "step": 1097 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.5737, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 0.0002, |
| "loss": 0.3385, |
| "step": 1099 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.2873, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4024, |
| "step": 1101 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.467, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5981, |
| "step": 1103 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3813, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4791, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4668, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5854, |
| "step": 1107 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3849, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4074, |
| "step": 1109 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.368, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.2722, |
| "step": 1111 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.2903, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4512, |
| "step": 1113 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3492, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4988, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.5053, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.454, |
| "step": 1117 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4813, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4281, |
| "step": 1119 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4667, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.494, |
| "step": 1121 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.4261, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3002, |
| "step": 1123 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3805, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.3057, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 0.0002, |
| "loss": 0.333, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3789, |
| "step": 1127 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3519, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.6106, |
| "step": 1129 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3852, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3817, |
| "step": 1131 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3334, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.4827, |
| "step": 1133 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3613, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.4734, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.4306, |
| "step": 1137 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.2529, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.2975, |
| "step": 1139 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.333, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3618, |
| "step": 1141 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.2847, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3323, |
| "step": 1143 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.5374, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3075, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.2617, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3223, |
| "step": 1147 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.433, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.2839, |
| "step": 1149 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.4591, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3216, |
| "step": 1151 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 0.0002, |
| "loss": 0.3637, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3696, |
| "step": 1153 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5672, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5244, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4046, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.5307, |
| "step": 1157 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3518, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3615, |
| "step": 1159 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.2892, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4484, |
| "step": 1161 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3547, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3376, |
| "step": 1163 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3083, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3545, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3231, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4428, |
| "step": 1167 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3406, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3504, |
| "step": 1169 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.394, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3407, |
| "step": 1171 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.2859, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3684, |
| "step": 1173 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4192, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3937, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.2995, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4853, |
| "step": 1177 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.3273, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 0.0002, |
| "loss": 0.4424, |
| "step": 1179 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.361, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4409, |
| "step": 1181 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4263, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4677, |
| "step": 1183 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.2679, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3583, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3348, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4628, |
| "step": 1187 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3279, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4118, |
| "step": 1189 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3114, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4237, |
| "step": 1191 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3609, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4261, |
| "step": 1193 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3747, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3264, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.499, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3874, |
| "step": 1197 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.402, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4942, |
| "step": 1199 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4471, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.45, |
| "eval_loss": 0.3973161280155182, |
| "eval_runtime": 109.3112, |
| "eval_samples_per_second": 9.148, |
| "eval_steps_per_second": 4.574, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.45, |
| "mmlu_eval_accuracy": 0.5878307641088379, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "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": 0.375, |
| "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": 0.45454545454545453, |
| "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": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "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": 0.45454545454545453, |
| "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": 0.7575757575757576, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.4117647058823529, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5833333333333334, |
| "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": 0.4444444444444444, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.3765496731816012, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4067, |
| "step": 1201 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.2672, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4171, |
| "step": 1203 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.3167, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 0.0002, |
| "loss": 0.4223, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3621, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.4782, |
| "step": 1207 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3241, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.4274, |
| "step": 1209 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3621, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.5274, |
| "step": 1211 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.4961, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.278, |
| "step": 1213 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3791, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3833, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3995, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.4709, |
| "step": 1217 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3735, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 1219 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3026, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 1221 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3003, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3901, |
| "step": 1223 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.382, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.6964, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.483, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.4353, |
| "step": 1227 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3116, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3651, |
| "step": 1229 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.2856, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3326, |
| "step": 1231 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 0.0002, |
| "loss": 0.3666, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3424, |
| "step": 1233 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.2386, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3193, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3373, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3223, |
| "step": 1237 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.4351, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.541, |
| "step": 1239 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.454, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3081, |
| "step": 1241 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.2999, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.345, |
| "step": 1243 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3299, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.4702, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5129, |
| "step": 1247 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.1806, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.2733, |
| "step": 1249 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3116, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3171, |
| "step": 1251 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.4545, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.3149, |
| "step": 1253 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.4079, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.4306, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.2925, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.2572, |
| "step": 1257 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 0.0002, |
| "loss": 0.5055, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.2929, |
| "step": 1259 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4425, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5234, |
| "step": 1261 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5057, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4988, |
| "step": 1263 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3877, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4465, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.2995, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4676, |
| "step": 1267 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4617, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4466, |
| "step": 1269 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3682, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.2637, |
| "step": 1271 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4405, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4056, |
| "step": 1273 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3295, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5838, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4957, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4556, |
| "step": 1277 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4455, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4701, |
| "step": 1279 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3817, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.5177, |
| "step": 1281 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3489, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3921, |
| "step": 1283 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.3708, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 0.0002, |
| "loss": 0.4552, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4568, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3945, |
| "step": 1287 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4025, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3903, |
| "step": 1289 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.2998, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3977, |
| "step": 1291 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3426, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4252, |
| "step": 1293 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4436, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4136, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.2734, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3641, |
| "step": 1297 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3653, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4079, |
| "step": 1299 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3217, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.4242, |
| "step": 1301 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3321, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3761, |
| "step": 1303 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3981, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3712, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.2812, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3644, |
| "step": 1307 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3719, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.3839, |
| "step": 1309 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.5035, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 0.0002, |
| "loss": 0.375, |
| "step": 1311 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4431, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3602, |
| "step": 1313 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4012, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3904, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3722, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3345, |
| "step": 1317 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.451, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3473, |
| "step": 1319 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3227, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.422, |
| "step": 1321 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3398, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4625, |
| "step": 1323 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4618, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.5057, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3704, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4962, |
| "step": 1327 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.6061, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4197, |
| "step": 1329 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4174, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3241, |
| "step": 1331 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3821, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3956, |
| "step": 1333 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3495, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3402, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3671, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.3832, |
| "step": 1337 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 0.0002, |
| "loss": 0.4076, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3412, |
| "step": 1339 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5289, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3185, |
| "step": 1341 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3007, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3142, |
| "step": 1343 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.384, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4195, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4198, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4044, |
| "step": 1347 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4306, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.2511, |
| "step": 1349 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4084, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4269, |
| "step": 1351 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3886, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.414, |
| "step": 1353 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4491, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3924, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3812, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3627, |
| "step": 1357 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3557, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4834, |
| "step": 1359 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4601, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5508, |
| "step": 1361 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.5223, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.3502, |
| "step": 1363 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 0.0002, |
| "loss": 0.4181, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.2477, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4421, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.2368, |
| "step": 1367 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.2845, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3117, |
| "step": 1369 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4895, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3828, |
| "step": 1371 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3144, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.497, |
| "step": 1373 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4081, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3645, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.327, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3926, |
| "step": 1377 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3008, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3308, |
| "step": 1379 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.494, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4915, |
| "step": 1381 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4259, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.4117, |
| "step": 1383 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3398, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3431, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.5041, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3945, |
| "step": 1387 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3664, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3797, |
| "step": 1389 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.428, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 0.0002, |
| "loss": 0.3813, |
| "step": 1391 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3467, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4689, |
| "step": 1393 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4307, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3812, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4326, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.355, |
| "step": 1397 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3454, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3435, |
| "step": 1399 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4541, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.53, |
| "eval_loss": 0.3919210135936737, |
| "eval_runtime": 109.3425, |
| "eval_samples_per_second": 9.146, |
| "eval_steps_per_second": 4.573, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.53, |
| "mmlu_eval_accuracy": 0.5943960342003739, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "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": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.6875, |
| "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": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.35, |
| "mmlu_eval_accuracy_nutrition": 0.7878787878787878, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_law": 0.38823529411764707, |
| "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": 0.8181818181818182, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.430495947249253, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3658, |
| "step": 1401 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3741, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.378, |
| "step": 1403 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4121, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3092, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3114, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.2581, |
| "step": 1407 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5029, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3106, |
| "step": 1409 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4683, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3414, |
| "step": 1411 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4253, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.2666, |
| "step": 1413 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3184, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.5808, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.3819, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 0.0002, |
| "loss": 0.4271, |
| "step": 1417 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4429, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5305, |
| "step": 1419 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4047, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4653, |
| "step": 1421 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.2859, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4155, |
| "step": 1423 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3999, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4395, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3888, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3586, |
| "step": 1427 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3716, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3936, |
| "step": 1429 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4108, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3877, |
| "step": 1431 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4893, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3578, |
| "step": 1433 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.2711, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.458, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3221, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.451, |
| "step": 1437 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4812, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.406, |
| "step": 1439 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4486, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3832, |
| "step": 1441 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.4196, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.5252, |
| "step": 1443 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 0.0002, |
| "loss": 0.3706, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4136, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.2706, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3341, |
| "step": 1447 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.324, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4763, |
| "step": 1449 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.5143, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4334, |
| "step": 1451 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3565, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3901, |
| "step": 1453 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4628, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4163, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4226, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4224, |
| "step": 1457 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.6355, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.2945, |
| "step": 1459 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.2085, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4438, |
| "step": 1461 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3362, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3187, |
| "step": 1463 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4986, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3524, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.354, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3271, |
| "step": 1467 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.4145, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.2718, |
| "step": 1469 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 0.0002, |
| "loss": 0.3021, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3734, |
| "step": 1471 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3642, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4624, |
| "step": 1473 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.2432, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5313, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3959, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.389, |
| "step": 1477 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4401, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.2572, |
| "step": 1479 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.41, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3675, |
| "step": 1481 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.2565, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4277, |
| "step": 1483 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3345, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4147, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.2469, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4311, |
| "step": 1487 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4395, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.5252, |
| "step": 1489 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3809, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4402, |
| "step": 1491 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4182, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3847, |
| "step": 1493 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3033, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.4108, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.475, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 0.0002, |
| "loss": 0.3047, |
| "step": 1497 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3097, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4509, |
| "step": 1499 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.2568, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3732, |
| "step": 1501 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4522, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 1503 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.2741, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.475, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.2935, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3507, |
| "step": 1507 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.2728, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4388, |
| "step": 1509 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4239, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4014, |
| "step": 1511 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3673, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4135, |
| "step": 1513 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3738, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3993, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3892, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4542, |
| "step": 1517 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3905, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4141, |
| "step": 1519 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3679, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.4694, |
| "step": 1521 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3103, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 0.0002, |
| "loss": 0.3809, |
| "step": 1523 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3723, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3751, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4919, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3574, |
| "step": 1527 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4483, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3586, |
| "step": 1529 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3358, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3223, |
| "step": 1531 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3234, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3683, |
| "step": 1533 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3324, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3982, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4331, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.2786, |
| "step": 1537 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3059, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4499, |
| "step": 1539 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3407, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4224, |
| "step": 1541 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4496, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.326, |
| "step": 1543 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3158, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3606, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3976, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4082, |
| "step": 1547 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.4189, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3236, |
| "step": 1549 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 0.0002, |
| "loss": 0.3878, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.4145, |
| "step": 1551 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3427, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3822, |
| "step": 1553 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.2967, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3972, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3402, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.4881, |
| "step": 1557 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5709, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3951, |
| "step": 1559 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3142, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3561, |
| "step": 1561 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.2552, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 1563 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3445, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.4087, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.2846, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.4352, |
| "step": 1567 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.245, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5761, |
| "step": 1569 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3988, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.2249, |
| "step": 1571 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3237, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.37, |
| "step": 1573 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.351, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.3701, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 0.0002, |
| "loss": 0.5041, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.362, |
| "step": 1577 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4153, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3849, |
| "step": 1579 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3509, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3699, |
| "step": 1581 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4186, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3581, |
| "step": 1583 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.349, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.46, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4459, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.2296, |
| "step": 1587 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3398, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4978, |
| "step": 1589 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3854, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5183, |
| "step": 1591 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4001, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3552, |
| "step": 1593 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.2493, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3472, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3289, |
| "step": 1597 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.4031, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.2777, |
| "step": 1599 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5774, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.6, |
| "eval_loss": 0.3895304203033447, |
| "eval_runtime": 109.2775, |
| "eval_samples_per_second": 9.151, |
| "eval_steps_per_second": 4.576, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.6, |
| "mmlu_eval_accuracy": 0.5869943163267687, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5714285714285714, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.6363636363636364, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "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": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "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": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163, |
| "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, |
| "mmlu_eval_accuracy_moral_scenarios": 0.4, |
| "mmlu_eval_accuracy_nutrition": 0.7272727272727273, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.4, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "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": 0.6111111111111112, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.2838659554175236, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.3699, |
| "step": 1601 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.2938, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 0.0002, |
| "loss": 0.5988, |
| "step": 1603 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3276, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3574, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.432, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3514, |
| "step": 1607 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.274, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3681, |
| "step": 1609 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.482, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5167, |
| "step": 1611 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3879, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5542, |
| "step": 1613 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.474, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3906, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.2942, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.5772, |
| "step": 1617 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.4884, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3636, |
| "step": 1619 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.327, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.2713, |
| "step": 1621 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3338, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6594, |
| "step": 1623 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.4218, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3861, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.6115, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3025, |
| "step": 1627 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.2809, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 0.0002, |
| "loss": 0.3767, |
| "step": 1629 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3247, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4439, |
| "step": 1631 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4216, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4993, |
| "step": 1633 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3122, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3715, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3244, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.5918, |
| "step": 1637 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4078, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3303, |
| "step": 1639 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.436, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3539, |
| "step": 1641 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.359, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4562, |
| "step": 1643 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4161, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.439, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.4917, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3726, |
| "step": 1647 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3898, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3957, |
| "step": 1649 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3719, |
| "step": 1651 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3424, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.2499, |
| "step": 1653 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3177, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.362, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 0.0002, |
| "loss": 0.3274, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4458, |
| "step": 1657 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4373, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4165, |
| "step": 1659 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3979, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4243, |
| "step": 1661 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.405, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4303, |
| "step": 1663 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4474, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4443, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3521, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.395, |
| "step": 1667 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3714, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3712, |
| "step": 1669 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4697, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3472, |
| "step": 1671 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3566, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.2428, |
| "step": 1673 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3856, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4522, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4025, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.2901, |
| "step": 1677 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.5288, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3925, |
| "step": 1679 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.3519, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4305, |
| "step": 1681 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 0.0002, |
| "loss": 0.4346, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3283, |
| "step": 1683 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3946, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3792, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.421, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3454, |
| "step": 1687 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4003, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4045, |
| "step": 1689 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.5354, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4258, |
| "step": 1691 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3847, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3322, |
| "step": 1693 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3082, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.6116, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4244, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4363, |
| "step": 1697 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.462, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3727, |
| "step": 1699 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4245, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4285, |
| "step": 1701 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3367, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.4064, |
| "step": 1703 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3957, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.2701, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3341, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3294, |
| "step": 1707 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.3126, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 0.0002, |
| "loss": 0.2662, |
| "step": 1709 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3646, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.2992, |
| "step": 1711 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.2863, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4255, |
| "step": 1713 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.358, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.2414, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3618, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3428, |
| "step": 1717 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.2993, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3672, |
| "step": 1719 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4532, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4125, |
| "step": 1721 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.2446, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3975, |
| "step": 1723 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3635, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3444, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3574, |
| "step": 1727 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3994, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.417, |
| "step": 1729 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3088, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4101, |
| "step": 1731 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.362, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3016, |
| "step": 1733 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.3502, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 0.0002, |
| "loss": 0.4299, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3193, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3758, |
| "step": 1737 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3268, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3823, |
| "step": 1739 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4436, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.6257, |
| "step": 1741 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.2789, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4746, |
| "step": 1743 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.266, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3342, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3474, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3157, |
| "step": 1747 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3684, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3161, |
| "step": 1749 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3891, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.2925, |
| "step": 1751 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4809, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3927, |
| "step": 1753 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3014, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4863, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3836, |
| "step": 1757 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.379, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.4615, |
| "step": 1759 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3501, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3033, |
| "step": 1761 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 0.0002, |
| "loss": 0.3509, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3571, |
| "step": 1763 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3623, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.619, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4163, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3481, |
| "step": 1767 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3853, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3461, |
| "step": 1769 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.384, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3164, |
| "step": 1771 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4145, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3118, |
| "step": 1773 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4347, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4596, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4147, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3541, |
| "step": 1777 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.5005, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3031, |
| "step": 1779 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.2922, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3097, |
| "step": 1781 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.337, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3128, |
| "step": 1783 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3294, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3482, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.4663, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.3276, |
| "step": 1787 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 0.0002, |
| "loss": 0.306, |
| "step": 1788 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.407, |
| "step": 1789 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3794, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4589, |
| "step": 1791 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4304, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.2741, |
| "step": 1793 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3716, |
| "step": 1794 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4256, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4697, |
| "step": 1796 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.2952, |
| "step": 1797 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5269, |
| "step": 1798 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5646, |
| "step": 1799 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4059, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.68, |
| "eval_loss": 0.39166855812072754, |
| "eval_runtime": 109.3139, |
| "eval_samples_per_second": 9.148, |
| "eval_steps_per_second": 4.574, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.68, |
| "mmlu_eval_accuracy": 0.5560851450931389, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.5, |
| "mmlu_eval_accuracy_astronomy": 0.5625, |
| "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.41379310344827586, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_physics": 0.2727272727272727, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.4375, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.5625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.75, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667, |
| "mmlu_eval_accuracy_international_law": 0.7692307692307693, |
| "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576, |
| "mmlu_eval_accuracy_moral_scenarios": 0.25, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.45714285714285713, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.35294117647058826, |
| "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.7727272727272727, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.7204176164949532, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.461, |
| "step": 1801 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3833, |
| "step": 1802 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4427, |
| "step": 1803 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3002, |
| "step": 1804 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3598, |
| "step": 1805 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3593, |
| "step": 1806 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3351, |
| "step": 1807 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3535, |
| "step": 1808 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.5737, |
| "step": 1809 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4914, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3136, |
| "step": 1811 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3889, |
| "step": 1812 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.4292, |
| "step": 1813 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3397, |
| "step": 1814 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 0.0002, |
| "loss": 0.3657, |
| "step": 1815 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3244, |
| "step": 1816 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5725, |
| "step": 1817 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.4424, |
| "step": 1818 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3545, |
| "step": 1819 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.4068, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3997, |
| "step": 1821 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5032, |
| "step": 1822 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.459, |
| "step": 1823 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3757, |
| "step": 1824 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3152, |
| "step": 1825 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3192, |
| "step": 1826 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.2966, |
| "step": 1827 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3819, |
| "step": 1828 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.4267, |
| "step": 1829 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3925, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.2969, |
| "step": 1831 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.2565, |
| "step": 1832 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3947, |
| "step": 1833 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3202, |
| "step": 1834 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3374, |
| "step": 1835 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3938, |
| "step": 1836 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.4021, |
| "step": 1837 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.3051, |
| "step": 1838 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.4243, |
| "step": 1839 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.5862, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 0.0002, |
| "loss": 0.297, |
| "step": 1841 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3822, |
| "step": 1842 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.292, |
| "step": 1843 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4845, |
| "step": 1844 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4623, |
| "step": 1845 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3601, |
| "step": 1846 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.423, |
| "step": 1847 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4713, |
| "step": 1848 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3791, |
| "step": 1849 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3058, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4273, |
| "step": 1851 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3721, |
| "step": 1852 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.454, |
| "step": 1853 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.2363, |
| "step": 1854 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3263, |
| "step": 1855 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4596, |
| "step": 1856 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.447, |
| "step": 1857 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4485, |
| "step": 1858 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3951, |
| "step": 1859 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4132, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.4679, |
| "step": 1861 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3163, |
| "step": 1862 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3127, |
| "step": 1863 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3591, |
| "step": 1864 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.2921, |
| "step": 1865 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3448, |
| "step": 1866 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.306, |
| "step": 1867 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 0.0002, |
| "loss": 0.3257, |
| "step": 1868 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2883, |
| "step": 1869 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.4315, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 1871 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.4628, |
| "step": 1872 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2322, |
| "step": 1873 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.332, |
| "step": 1874 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.4604, |
| "step": 1875 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3821, |
| "step": 1876 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.4336, |
| "step": 1877 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2622, |
| "step": 1878 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2976, |
| "step": 1879 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3707, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3412, |
| "step": 1881 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.4193, |
| "step": 1882 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2956, |
| "step": 1883 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2596, |
| "step": 1884 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3502, |
| "step": 1885 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2716, |
| "step": 1886 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2889, |
| "step": 1887 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3606, |
| "step": 1888 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2698, |
| "step": 1889 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.5677, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3544, |
| "step": 1891 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.2628, |
| "step": 1892 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 1893 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.0002, |
| "loss": 0.251, |
| "step": 1894 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3725, |
| "step": 1895 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3372, |
| "step": 1896 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3568, |
| "step": 1897 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.371, |
| "step": 1898 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5576, |
| "step": 1899 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3879, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 1901 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5556, |
| "step": 1902 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.633, |
| "step": 1903 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5748, |
| "step": 1904 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3446, |
| "step": 1905 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3372, |
| "step": 1906 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3771, |
| "step": 1907 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.388, |
| "step": 1908 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.4026, |
| "step": 1909 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3547, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.2864, |
| "step": 1911 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3319, |
| "step": 1912 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3788, |
| "step": 1913 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.4784, |
| "step": 1914 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5697, |
| "step": 1915 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.2473, |
| "step": 1916 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5273, |
| "step": 1917 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.2925, |
| "step": 1918 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.3852, |
| "step": 1919 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.4548, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 0.0002, |
| "loss": 0.5352, |
| "step": 1921 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4374, |
| "step": 1922 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3149, |
| "step": 1923 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.2606, |
| "step": 1924 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4221, |
| "step": 1925 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5093, |
| "step": 1926 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3893, |
| "step": 1927 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3928, |
| "step": 1928 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5786, |
| "step": 1929 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3188, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4177, |
| "step": 1931 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3714, |
| "step": 1932 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.351, |
| "step": 1933 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3633, |
| "step": 1934 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.367, |
| "step": 1935 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.343, |
| "step": 1936 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5017, |
| "step": 1937 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4612, |
| "step": 1938 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4591, |
| "step": 1939 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3154, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3919, |
| "step": 1941 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3806, |
| "step": 1942 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.5318, |
| "step": 1943 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.2507, |
| "step": 1944 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.4071, |
| "step": 1945 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.346, |
| "step": 1946 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 0.0002, |
| "loss": 0.3801, |
| "step": 1947 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2606, |
| "step": 1948 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2638, |
| "step": 1949 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3928, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5397, |
| "step": 1951 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3009, |
| "step": 1952 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4123, |
| "step": 1953 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2937, |
| "step": 1954 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.501, |
| "step": 1955 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4681, |
| "step": 1956 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4681, |
| "step": 1957 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4274, |
| "step": 1958 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5111, |
| "step": 1959 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.6188, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.489, |
| "step": 1961 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3385, |
| "step": 1962 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3856, |
| "step": 1963 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.518, |
| "step": 1964 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3944, |
| "step": 1965 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3873, |
| "step": 1966 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3657, |
| "step": 1967 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2595, |
| "step": 1968 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2723, |
| "step": 1969 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.5662, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.3468, |
| "step": 1971 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4133, |
| "step": 1972 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.2703, |
| "step": 1973 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 0.0002, |
| "loss": 0.4565, |
| "step": 1974 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3527, |
| "step": 1975 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5033, |
| "step": 1976 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.2786, |
| "step": 1977 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5598, |
| "step": 1978 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3407, |
| "step": 1979 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4159, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4061, |
| "step": 1981 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5353, |
| "step": 1982 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4982, |
| "step": 1983 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.2872, |
| "step": 1984 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5088, |
| "step": 1985 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4098, |
| "step": 1986 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4807, |
| "step": 1987 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4344, |
| "step": 1988 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.2698, |
| "step": 1989 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3883, |
| "step": 1991 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3446, |
| "step": 1992 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3468, |
| "step": 1993 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4361, |
| "step": 1994 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4865, |
| "step": 1995 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4295, |
| "step": 1996 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3997, |
| "step": 1997 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.4393, |
| "step": 1998 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.3501, |
| "step": 1999 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 0.0002, |
| "loss": 0.5102, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.75, |
| "eval_loss": 0.38687166571617126, |
| "eval_runtime": 109.2781, |
| "eval_samples_per_second": 9.151, |
| "eval_steps_per_second": 4.575, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.75, |
| "mmlu_eval_accuracy": 0.5761208271066307, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, |
| "mmlu_eval_accuracy_college_medicine": 0.5, |
| "mmlu_eval_accuracy_college_physics": 0.36363636363636365, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.4166666666666667, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, |
| "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, |
| "mmlu_eval_accuracy_global_facts": 0.5, |
| "mmlu_eval_accuracy_high_school_biology": 0.65625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.8181818181818182, |
| "mmlu_eval_accuracy_marketing": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, |
| "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576, |
| "mmlu_eval_accuracy_moral_scenarios": 0.26, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.34705882352941175, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.4202043774389412, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4309, |
| "step": 2001 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5225, |
| "step": 2002 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.249, |
| "step": 2003 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.364, |
| "step": 2004 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3648, |
| "step": 2005 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5522, |
| "step": 2006 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4413, |
| "step": 2007 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.2785, |
| "step": 2008 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4248, |
| "step": 2009 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3695, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3981, |
| "step": 2011 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4788, |
| "step": 2012 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3779, |
| "step": 2013 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4033, |
| "step": 2014 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4071, |
| "step": 2015 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.2724, |
| "step": 2016 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5079, |
| "step": 2017 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3718, |
| "step": 2018 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4945, |
| "step": 2019 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.2411, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3976, |
| "step": 2021 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.4309, |
| "step": 2022 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5805, |
| "step": 2023 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 2024 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.3098, |
| "step": 2025 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.352, |
| "step": 2026 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 0.0002, |
| "loss": 0.5283, |
| "step": 2027 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4583, |
| "step": 2028 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4028, |
| "step": 2029 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4902, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.2897, |
| "step": 2031 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3956, |
| "step": 2032 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3174, |
| "step": 2033 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3154, |
| "step": 2034 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5636, |
| "step": 2035 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3931, |
| "step": 2036 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.2885, |
| "step": 2037 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.5227, |
| "step": 2038 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3799, |
| "step": 2039 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3933, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.2778, |
| "step": 2041 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4688, |
| "step": 2042 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3564, |
| "step": 2043 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3667, |
| "step": 2044 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4693, |
| "step": 2045 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3708, |
| "step": 2046 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.2963, |
| "step": 2047 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.2706, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3936, |
| "step": 2049 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.4114, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3954, |
| "step": 2051 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3239, |
| "step": 2052 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 2053 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4331, |
| "step": 2054 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.345, |
| "step": 2055 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4375, |
| "step": 2056 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3903, |
| "step": 2057 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.2571, |
| "step": 2058 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4632, |
| "step": 2059 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3967, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5081, |
| "step": 2061 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4147, |
| "step": 2062 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3802, |
| "step": 2063 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.2943, |
| "step": 2064 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4239, |
| "step": 2065 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3984, |
| "step": 2066 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4164, |
| "step": 2067 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4025, |
| "step": 2068 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5059, |
| "step": 2069 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.258, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5789, |
| "step": 2071 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3489, |
| "step": 2072 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.5417, |
| "step": 2073 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.2975, |
| "step": 2074 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4202, |
| "step": 2075 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3251, |
| "step": 2076 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4218, |
| "step": 2077 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3517, |
| "step": 2078 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.3532, |
| "step": 2079 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 0.0002, |
| "loss": 0.4846, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3595, |
| "step": 2081 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3426, |
| "step": 2082 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4799, |
| "step": 2083 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4157, |
| "step": 2084 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3658, |
| "step": 2085 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4682, |
| "step": 2086 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3328, |
| "step": 2087 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3802, |
| "step": 2088 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.401, |
| "step": 2089 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3755, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.475, |
| "step": 2091 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4175, |
| "step": 2092 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3874, |
| "step": 2093 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4356, |
| "step": 2094 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3391, |
| "step": 2095 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3647, |
| "step": 2096 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3812, |
| "step": 2097 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4665, |
| "step": 2098 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4184, |
| "step": 2099 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4145, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.2649, |
| "step": 2101 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3053, |
| "step": 2102 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.3414, |
| "step": 2103 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4344, |
| "step": 2104 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.5297, |
| "step": 2105 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 0.0002, |
| "loss": 0.4709, |
| "step": 2106 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.2672, |
| "step": 2107 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.389, |
| "step": 2108 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.5068, |
| "step": 2109 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3128, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.386, |
| "step": 2111 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.418, |
| "step": 2112 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3515, |
| "step": 2113 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.363, |
| "step": 2114 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.2828, |
| "step": 2115 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.527, |
| "step": 2116 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3731, |
| "step": 2117 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3481, |
| "step": 2118 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3709, |
| "step": 2119 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4153, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4024, |
| "step": 2121 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3239, |
| "step": 2122 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3339, |
| "step": 2123 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3089, |
| "step": 2124 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3638, |
| "step": 2125 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4689, |
| "step": 2126 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4792, |
| "step": 2127 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3776, |
| "step": 2128 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.4096, |
| "step": 2129 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.393, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.6117, |
| "step": 2131 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.3704, |
| "step": 2132 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 0.0002, |
| "loss": 0.2536, |
| "step": 2133 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3941, |
| "step": 2134 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.2911, |
| "step": 2135 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.2801, |
| "step": 2136 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3888, |
| "step": 2137 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3835, |
| "step": 2138 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4371, |
| "step": 2139 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4879, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.2476, |
| "step": 2141 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3266, |
| "step": 2142 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.367, |
| "step": 2143 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3642, |
| "step": 2144 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.339, |
| "step": 2145 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.421, |
| "step": 2146 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.2816, |
| "step": 2147 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3254, |
| "step": 2148 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3502, |
| "step": 2149 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3683, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4288, |
| "step": 2151 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3377, |
| "step": 2152 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3397, |
| "step": 2153 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4889, |
| "step": 2154 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.2737, |
| "step": 2155 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4413, |
| "step": 2156 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4627, |
| "step": 2157 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.4673, |
| "step": 2158 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 0.0002, |
| "loss": 0.3384, |
| "step": 2159 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3156, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5177, |
| "step": 2161 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4097, |
| "step": 2162 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3213, |
| "step": 2163 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5659, |
| "step": 2164 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3978, |
| "step": 2165 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3835, |
| "step": 2166 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3393, |
| "step": 2167 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.34, |
| "step": 2168 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.2879, |
| "step": 2169 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3835, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3619, |
| "step": 2171 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5128, |
| "step": 2172 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4543, |
| "step": 2173 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.2672, |
| "step": 2174 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 2175 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4465, |
| "step": 2176 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.4687, |
| "step": 2177 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3366, |
| "step": 2178 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3243, |
| "step": 2179 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5249, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.372, |
| "step": 2181 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3483, |
| "step": 2182 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3361, |
| "step": 2183 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.5428, |
| "step": 2184 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.3573, |
| "step": 2185 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 0.0002, |
| "loss": 0.2723, |
| "step": 2186 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 2187 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4203, |
| "step": 2188 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.2706, |
| "step": 2189 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4164, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3946, |
| "step": 2191 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.2854, |
| "step": 2192 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3627, |
| "step": 2193 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3207, |
| "step": 2194 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3395, |
| "step": 2195 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3748, |
| "step": 2196 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3788, |
| "step": 2197 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3233, |
| "step": 2198 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3676, |
| "step": 2199 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3887, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.83, |
| "eval_loss": 0.38439005613327026, |
| "eval_runtime": 109.2939, |
| "eval_samples_per_second": 9.15, |
| "eval_steps_per_second": 4.575, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.83, |
| "mmlu_eval_accuracy": 0.5748054291718571, |
| "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, |
| "mmlu_eval_accuracy_anatomy": 0.35714285714285715, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.3, |
| "mmlu_eval_accuracy_high_school_biology": 0.625, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, |
| "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, |
| "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "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": 0.84, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, |
| "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.6764705882352942, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, |
| "mmlu_eval_accuracy_professional_law": 0.35294117647058826, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9545454545454546, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.8421052631578947, |
| "mmlu_loss": 1.4178901468026732, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4733, |
| "step": 2201 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4062, |
| "step": 2202 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3114, |
| "step": 2203 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4533, |
| "step": 2204 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.2862, |
| "step": 2205 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3976, |
| "step": 2206 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.4423, |
| "step": 2207 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3135, |
| "step": 2208 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.438, |
| "step": 2209 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3568, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.5287, |
| "step": 2211 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 0.0002, |
| "loss": 0.3898, |
| "step": 2212 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4428, |
| "step": 2213 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4172, |
| "step": 2214 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4741, |
| "step": 2215 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.395, |
| "step": 2216 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.384, |
| "step": 2217 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3952, |
| "step": 2218 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3902, |
| "step": 2219 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.372, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4101, |
| "step": 2221 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.5495, |
| "step": 2222 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3281, |
| "step": 2223 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4096, |
| "step": 2224 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3162, |
| "step": 2225 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4257, |
| "step": 2226 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4392, |
| "step": 2227 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3094, |
| "step": 2228 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.2851, |
| "step": 2229 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3607, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4014, |
| "step": 2231 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3487, |
| "step": 2232 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.456, |
| "step": 2233 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.4818, |
| "step": 2234 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.5032, |
| "step": 2235 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.371, |
| "step": 2236 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.3312, |
| "step": 2237 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.5657, |
| "step": 2238 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 0.0002, |
| "loss": 0.6022, |
| "step": 2239 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3075, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.5839, |
| "step": 2241 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4143, |
| "step": 2242 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4611, |
| "step": 2243 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.2838, |
| "step": 2244 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3226, |
| "step": 2245 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.2914, |
| "step": 2246 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3371, |
| "step": 2247 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3151, |
| "step": 2248 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3619, |
| "step": 2249 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3387, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4544, |
| "step": 2251 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4141, |
| "step": 2252 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4712, |
| "step": 2253 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.5178, |
| "step": 2254 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4036, |
| "step": 2255 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3911, |
| "step": 2256 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4072, |
| "step": 2257 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3007, |
| "step": 2258 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4318, |
| "step": 2259 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4445, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.2927, |
| "step": 2261 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.4548, |
| "step": 2262 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3339, |
| "step": 2263 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3769, |
| "step": 2264 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 0.0002, |
| "loss": 0.3929, |
| "step": 2265 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.2904, |
| "step": 2266 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3368, |
| "step": 2267 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4494, |
| "step": 2268 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4676, |
| "step": 2269 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.7326, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.322, |
| "step": 2271 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4107, |
| "step": 2272 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3418, |
| "step": 2273 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3139, |
| "step": 2274 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4646, |
| "step": 2275 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.6116, |
| "step": 2276 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3747, |
| "step": 2277 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4943, |
| "step": 2278 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3695, |
| "step": 2279 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4551, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4125, |
| "step": 2281 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4009, |
| "step": 2282 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3497, |
| "step": 2283 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.3549, |
| "step": 2284 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.2904, |
| "step": 2285 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4324, |
| "step": 2286 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.5577, |
| "step": 2287 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4284, |
| "step": 2288 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.4466, |
| "step": 2289 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.321, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.5164, |
| "step": 2291 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 0.0002, |
| "loss": 0.2949, |
| "step": 2292 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3043, |
| "step": 2293 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3467, |
| "step": 2294 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.6274, |
| "step": 2295 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.4842, |
| "step": 2296 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3895, |
| "step": 2297 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.247, |
| "step": 2298 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.44, |
| "step": 2299 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.285, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.2996, |
| "step": 2301 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3283, |
| "step": 2302 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.314, |
| "step": 2303 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3724, |
| "step": 2304 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3428, |
| "step": 2305 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3075, |
| "step": 2306 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.5215, |
| "step": 2307 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3552, |
| "step": 2308 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.4384, |
| "step": 2309 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.2418, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.4045, |
| "step": 2311 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3375, |
| "step": 2312 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3861, |
| "step": 2313 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3478, |
| "step": 2314 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.4169, |
| "step": 2315 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3833, |
| "step": 2316 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3786, |
| "step": 2317 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 0.0002, |
| "loss": 0.3264, |
| "step": 2318 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3801, |
| "step": 2319 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.2749, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.409, |
| "step": 2321 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3152, |
| "step": 2322 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3804, |
| "step": 2323 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.4292, |
| "step": 2324 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.4739, |
| "step": 2325 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3674, |
| "step": 2326 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3162, |
| "step": 2327 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.43, |
| "step": 2328 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5511, |
| "step": 2329 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.4271, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3951, |
| "step": 2331 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3982, |
| "step": 2332 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5203, |
| "step": 2333 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5344, |
| "step": 2334 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.2639, |
| "step": 2335 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.2864, |
| "step": 2336 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5598, |
| "step": 2337 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3732, |
| "step": 2338 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.414, |
| "step": 2339 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.5561, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3569, |
| "step": 2341 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.378, |
| "step": 2342 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.2984, |
| "step": 2343 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.4439, |
| "step": 2344 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 0.0002, |
| "loss": 0.3296, |
| "step": 2345 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3961, |
| "step": 2346 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3708, |
| "step": 2347 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3573, |
| "step": 2348 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3528, |
| "step": 2349 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3071, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.388, |
| "step": 2351 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 2352 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.4382, |
| "step": 2353 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3011, |
| "step": 2354 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3115, |
| "step": 2355 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3259, |
| "step": 2356 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.4709, |
| "step": 2357 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.4017, |
| "step": 2358 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.28, |
| "step": 2359 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3546, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.2827, |
| "step": 2361 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.4919, |
| "step": 2362 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.2716, |
| "step": 2363 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3859, |
| "step": 2364 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3418, |
| "step": 2365 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.4281, |
| "step": 2366 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3232, |
| "step": 2367 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.5031, |
| "step": 2368 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.465, |
| "step": 2369 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3826, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 0.0002, |
| "loss": 0.3709, |
| "step": 2371 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2194, |
| "step": 2372 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3256, |
| "step": 2373 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3842, |
| "step": 2374 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.5682, |
| "step": 2375 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.54, |
| "step": 2376 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2866, |
| "step": 2377 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.5901, |
| "step": 2378 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3731, |
| "step": 2379 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3329, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3889, |
| "step": 2381 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3428, |
| "step": 2382 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2896, |
| "step": 2383 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3676, |
| "step": 2384 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3106, |
| "step": 2385 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.41, |
| "step": 2386 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.4161, |
| "step": 2387 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3147, |
| "step": 2388 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.4987, |
| "step": 2389 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2835, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2644, |
| "step": 2391 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.349, |
| "step": 2392 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3534, |
| "step": 2393 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.5045, |
| "step": 2394 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.399, |
| "step": 2395 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.3701, |
| "step": 2396 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.4193, |
| "step": 2397 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 0.0002, |
| "loss": 0.2753, |
| "step": 2398 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3887, |
| "step": 2399 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3645, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.91, |
| "eval_loss": 0.38827234506607056, |
| "eval_runtime": 109.197, |
| "eval_samples_per_second": 9.158, |
| "eval_steps_per_second": 4.579, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.91, |
| "mmlu_eval_accuracy": 0.5877893898471896, |
| "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.7272727272727273, |
| "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, |
| "mmlu_eval_accuracy_college_medicine": 0.5, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.59375, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, |
| "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, |
| "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.7391304347826086, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, |
| "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, |
| "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576, |
| "mmlu_eval_accuracy_moral_scenarios": 0.28, |
| "mmlu_eval_accuracy_nutrition": 0.6666666666666666, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.4, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.6666666666666666, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.8947368421052632, |
| "mmlu_loss": 1.4799881345998525, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4175, |
| "step": 2401 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.366, |
| "step": 2402 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3297, |
| "step": 2403 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.452, |
| "step": 2404 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4331, |
| "step": 2405 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4413, |
| "step": 2406 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5263, |
| "step": 2407 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4252, |
| "step": 2408 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.269, |
| "step": 2409 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4083, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5603, |
| "step": 2411 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5307, |
| "step": 2412 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5182, |
| "step": 2413 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5611, |
| "step": 2414 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.5463, |
| "step": 2415 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3018, |
| "step": 2416 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.486, |
| "step": 2417 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3124, |
| "step": 2418 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3769, |
| "step": 2419 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3158, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.3975, |
| "step": 2421 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4261, |
| "step": 2422 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.2795, |
| "step": 2423 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 0.0002, |
| "loss": 0.4781, |
| "step": 2424 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3217, |
| "step": 2425 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.2763, |
| "step": 2426 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3485, |
| "step": 2427 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3441, |
| "step": 2428 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3678, |
| "step": 2429 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4558, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.2659, |
| "step": 2431 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.362, |
| "step": 2432 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3845, |
| "step": 2433 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4661, |
| "step": 2434 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4234, |
| "step": 2435 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3704, |
| "step": 2436 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5599, |
| "step": 2437 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5149, |
| "step": 2438 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.2346, |
| "step": 2439 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5292, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3288, |
| "step": 2441 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4239, |
| "step": 2442 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4601, |
| "step": 2443 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.2799, |
| "step": 2444 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4551, |
| "step": 2445 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4019, |
| "step": 2446 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.5058, |
| "step": 2447 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3413, |
| "step": 2448 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3242, |
| "step": 2449 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.4067, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 0.0002, |
| "loss": 0.3141, |
| "step": 2451 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.2926, |
| "step": 2452 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.371, |
| "step": 2453 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.2974, |
| "step": 2454 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.283, |
| "step": 2455 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3602, |
| "step": 2456 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.2553, |
| "step": 2457 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4967, |
| "step": 2458 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.2978, |
| "step": 2459 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3322, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3294, |
| "step": 2461 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3453, |
| "step": 2462 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3853, |
| "step": 2463 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4737, |
| "step": 2464 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4047, |
| "step": 2465 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.255, |
| "step": 2466 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3391, |
| "step": 2467 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3253, |
| "step": 2468 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3927, |
| "step": 2469 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3383, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4949, |
| "step": 2471 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3669, |
| "step": 2472 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4645, |
| "step": 2473 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3279, |
| "step": 2474 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.3675, |
| "step": 2475 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.4094, |
| "step": 2476 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 0.0002, |
| "loss": 0.335, |
| "step": 2477 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.233, |
| "step": 2478 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2573, |
| "step": 2479 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.25, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2578, |
| "step": 2481 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3913, |
| "step": 2482 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2654, |
| "step": 2483 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3471, |
| "step": 2484 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3928, |
| "step": 2485 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.328, |
| "step": 2486 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.4284, |
| "step": 2487 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.4382, |
| "step": 2488 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.296, |
| "step": 2489 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2685, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.4162, |
| "step": 2491 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.4309, |
| "step": 2492 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3143, |
| "step": 2493 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.5252, |
| "step": 2494 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.717, |
| "step": 2495 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2973, |
| "step": 2496 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3306, |
| "step": 2497 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3593, |
| "step": 2498 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3013, |
| "step": 2499 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3871, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.3837, |
| "step": 2501 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2698, |
| "step": 2502 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.2984, |
| "step": 2503 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 0.0002, |
| "loss": 0.4614, |
| "step": 2504 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4992, |
| "step": 2505 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3873, |
| "step": 2506 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.5034, |
| "step": 2507 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4938, |
| "step": 2508 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4141, |
| "step": 2509 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4087, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.254, |
| "step": 2511 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.2909, |
| "step": 2512 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.369, |
| "step": 2513 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3736, |
| "step": 2514 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 2515 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4367, |
| "step": 2516 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.4092, |
| "step": 2517 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3922, |
| "step": 2518 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.298, |
| "step": 2519 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.5166, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.36, |
| "step": 2521 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3477, |
| "step": 2522 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.5159, |
| "step": 2523 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.2759, |
| "step": 2524 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.2977, |
| "step": 2525 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.297, |
| "step": 2526 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3838, |
| "step": 2527 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.453, |
| "step": 2528 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.3895, |
| "step": 2529 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 0.0002, |
| "loss": 0.43, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3495, |
| "step": 2531 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.409, |
| "step": 2532 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3545, |
| "step": 2533 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3858, |
| "step": 2534 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.2848, |
| "step": 2535 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5221, |
| "step": 2536 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3817, |
| "step": 2537 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3993, |
| "step": 2538 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4836, |
| "step": 2539 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3156, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4209, |
| "step": 2541 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5286, |
| "step": 2542 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3666, |
| "step": 2543 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5576, |
| "step": 2544 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3293, |
| "step": 2545 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3885, |
| "step": 2546 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4399, |
| "step": 2547 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.5028, |
| "step": 2548 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3689, |
| "step": 2549 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3114, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4616, |
| "step": 2551 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3555, |
| "step": 2552 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4953, |
| "step": 2553 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3426, |
| "step": 2554 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.363, |
| "step": 2555 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.4908, |
| "step": 2556 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 0.0002, |
| "loss": 0.3993, |
| "step": 2557 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3924, |
| "step": 2558 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.5228, |
| "step": 2559 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4615, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4424, |
| "step": 2561 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.399, |
| "step": 2562 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3429, |
| "step": 2563 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3748, |
| "step": 2564 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4229, |
| "step": 2565 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.2808, |
| "step": 2566 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4284, |
| "step": 2567 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3534, |
| "step": 2568 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4359, |
| "step": 2569 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.497, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4427, |
| "step": 2571 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.432, |
| "step": 2572 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3364, |
| "step": 2573 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3063, |
| "step": 2574 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3597, |
| "step": 2575 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3732, |
| "step": 2576 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3926, |
| "step": 2577 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3631, |
| "step": 2578 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.424, |
| "step": 2579 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3406, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4381, |
| "step": 2581 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.3838, |
| "step": 2582 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 0.0002, |
| "loss": 0.4634, |
| "step": 2583 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3976, |
| "step": 2584 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3036, |
| "step": 2585 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3721, |
| "step": 2586 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4796, |
| "step": 2587 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3395, |
| "step": 2588 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4545, |
| "step": 2589 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3308, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.5343, |
| "step": 2591 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3698, |
| "step": 2592 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3936, |
| "step": 2593 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4874, |
| "step": 2594 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4573, |
| "step": 2595 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.5155, |
| "step": 2596 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.313, |
| "step": 2597 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.2951, |
| "step": 2598 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3221, |
| "step": 2599 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3647, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.98, |
| "eval_loss": 0.38303717970848083, |
| "eval_runtime": 109.2453, |
| "eval_samples_per_second": 9.154, |
| "eval_steps_per_second": 4.577, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.98, |
| "mmlu_eval_accuracy": 0.5749283809975633, |
| "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, |
| "mmlu_eval_accuracy_anatomy": 0.35714285714285715, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "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": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_physics": 0.45454545454545453, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "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": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, |
| "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, |
| "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.782608695652174, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "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": 0.7674418604651163, |
| "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, |
| "mmlu_eval_accuracy_moral_scenarios": 0.38, |
| "mmlu_eval_accuracy_nutrition": 0.696969696969697, |
| "mmlu_eval_accuracy_philosophy": 0.7352941176470589, |
| "mmlu_eval_accuracy_prehistory": 0.5714285714285714, |
| "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, |
| "mmlu_eval_accuracy_professional_law": 0.3764705882352941, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.6296296296296297, |
| "mmlu_eval_accuracy_sociology": 0.8636363636363636, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.5, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.6306076764515736, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4256, |
| "step": 2601 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3516, |
| "step": 2602 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3627, |
| "step": 2603 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.2871, |
| "step": 2604 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3737, |
| "step": 2605 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3189, |
| "step": 2606 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.5855, |
| "step": 2607 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3324, |
| "step": 2608 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.4824, |
| "step": 2609 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 0.0002, |
| "loss": 0.3564, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.5744, |
| "step": 2611 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.7767, |
| "step": 2612 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4538, |
| "step": 2613 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4027, |
| "step": 2614 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4214, |
| "step": 2615 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4277, |
| "step": 2616 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4669, |
| "step": 2617 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.494, |
| "step": 2618 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.2376, |
| "step": 2619 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4945, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4485, |
| "step": 2621 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4012, |
| "step": 2622 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4003, |
| "step": 2623 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3348, |
| "step": 2624 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3093, |
| "step": 2625 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4131, |
| "step": 2626 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3268, |
| "step": 2627 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4859, |
| "step": 2628 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3487, |
| "step": 2629 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.442, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3988, |
| "step": 2631 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.3076, |
| "step": 2632 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.39, |
| "step": 2633 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.414, |
| "step": 2634 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.409, |
| "step": 2635 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 0.0002, |
| "loss": 0.4814, |
| "step": 2636 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5269, |
| "step": 2637 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3708, |
| "step": 2638 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4898, |
| "step": 2639 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3501, |
| "step": 2640 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3574, |
| "step": 2641 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.5609, |
| "step": 2642 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.6113, |
| "step": 2643 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3769, |
| "step": 2644 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4144, |
| "step": 2645 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2995, |
| "step": 2646 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3899, |
| "step": 2647 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4546, |
| "step": 2648 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4, |
| "step": 2649 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.285, |
| "step": 2650 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2414, |
| "step": 2651 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2501, |
| "step": 2652 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2196, |
| "step": 2653 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2726, |
| "step": 2654 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.231, |
| "step": 2655 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.334, |
| "step": 2656 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.4264, |
| "step": 2657 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2843, |
| "step": 2658 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2883, |
| "step": 2659 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2554, |
| "step": 2660 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.2159, |
| "step": 2661 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 0.0002, |
| "loss": 0.3949, |
| "step": 2662 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2871, |
| "step": 2663 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2677, |
| "step": 2664 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3022, |
| "step": 2665 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3581, |
| "step": 2666 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 2667 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2871, |
| "step": 2668 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3681, |
| "step": 2669 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3434, |
| "step": 2670 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4263, |
| "step": 2671 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3252, |
| "step": 2672 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4105, |
| "step": 2673 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3723, |
| "step": 2674 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3481, |
| "step": 2675 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3001, |
| "step": 2676 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2805, |
| "step": 2677 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2431, |
| "step": 2678 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.3004, |
| "step": 2679 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4248, |
| "step": 2680 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2712, |
| "step": 2681 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2639, |
| "step": 2682 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4092, |
| "step": 2683 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2856, |
| "step": 2684 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4086, |
| "step": 2685 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2932, |
| "step": 2686 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2997, |
| "step": 2687 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.4117, |
| "step": 2688 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 0.0002, |
| "loss": 0.2806, |
| "step": 2689 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2779, |
| "step": 2690 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.282, |
| "step": 2691 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2972, |
| "step": 2692 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2692, |
| "step": 2693 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2669, |
| "step": 2694 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2806, |
| "step": 2695 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3212, |
| "step": 2696 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3337, |
| "step": 2697 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3569, |
| "step": 2698 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3191, |
| "step": 2699 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2757, |
| "step": 2700 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2498, |
| "step": 2701 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.226, |
| "step": 2702 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2739, |
| "step": 2703 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.312, |
| "step": 2704 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3233, |
| "step": 2705 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2295, |
| "step": 2706 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2532, |
| "step": 2707 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2618, |
| "step": 2708 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3992, |
| "step": 2709 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2854, |
| "step": 2710 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3489, |
| "step": 2711 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3874, |
| "step": 2712 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.263, |
| "step": 2713 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.3216, |
| "step": 2714 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 0.0002, |
| "loss": 0.2332, |
| "step": 2715 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.4758, |
| "step": 2716 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.331, |
| "step": 2717 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3426, |
| "step": 2718 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2859, |
| "step": 2719 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3842, |
| "step": 2720 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3333, |
| "step": 2721 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3451, |
| "step": 2722 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3668, |
| "step": 2723 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3427, |
| "step": 2724 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.276, |
| "step": 2725 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.238, |
| "step": 2726 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2865, |
| "step": 2727 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3766, |
| "step": 2728 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3955, |
| "step": 2729 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3085, |
| "step": 2730 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3089, |
| "step": 2731 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3204, |
| "step": 2732 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2277, |
| "step": 2733 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3654, |
| "step": 2734 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2761, |
| "step": 2735 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2538, |
| "step": 2736 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3303, |
| "step": 2737 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2479, |
| "step": 2738 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3122, |
| "step": 2739 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3183, |
| "step": 2740 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.3886, |
| "step": 2741 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 0.0002, |
| "loss": 0.2985, |
| "step": 2742 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3081, |
| "step": 2743 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2733, |
| "step": 2744 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3309, |
| "step": 2745 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3383, |
| "step": 2746 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3119, |
| "step": 2747 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.1899, |
| "step": 2748 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2661, |
| "step": 2749 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4145, |
| "step": 2750 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2715, |
| "step": 2751 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3557, |
| "step": 2752 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3699, |
| "step": 2753 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.242, |
| "step": 2754 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3456, |
| "step": 2755 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.4492, |
| "step": 2756 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3837, |
| "step": 2757 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2371, |
| "step": 2758 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2692, |
| "step": 2759 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.341, |
| "step": 2760 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.368, |
| "step": 2761 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3931, |
| "step": 2762 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2535, |
| "step": 2763 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3298, |
| "step": 2764 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3094, |
| "step": 2765 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.3492, |
| "step": 2766 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.405, |
| "step": 2767 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 0.0002, |
| "loss": 0.2497, |
| "step": 2768 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3936, |
| "step": 2769 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2061, |
| "step": 2770 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 2771 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.297, |
| "step": 2772 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2732, |
| "step": 2773 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.406, |
| "step": 2774 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3489, |
| "step": 2775 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.265, |
| "step": 2776 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3178, |
| "step": 2777 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2698, |
| "step": 2778 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2655, |
| "step": 2779 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3069, |
| "step": 2780 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3496, |
| "step": 2781 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3809, |
| "step": 2782 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2892, |
| "step": 2783 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3733, |
| "step": 2784 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2735, |
| "step": 2785 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2376, |
| "step": 2786 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2447, |
| "step": 2787 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.5068, |
| "step": 2788 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3448, |
| "step": 2789 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2551, |
| "step": 2790 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2932, |
| "step": 2791 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2685, |
| "step": 2792 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.3022, |
| "step": 2793 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.1912, |
| "step": 2794 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 0.0002, |
| "loss": 0.2919, |
| "step": 2795 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3043, |
| "step": 2796 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2692, |
| "step": 2797 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3578, |
| "step": 2798 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2203, |
| "step": 2799 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3575, |
| "step": 2800 |
| }, |
| { |
| "epoch": 1.06, |
| "eval_loss": 0.38800597190856934, |
| "eval_runtime": 109.2842, |
| "eval_samples_per_second": 9.15, |
| "eval_steps_per_second": 4.575, |
| "step": 2800 |
| }, |
| { |
| "epoch": 1.06, |
| "mmlu_eval_accuracy": 0.5875814356275961, |
| "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.6875, |
| "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.5, |
| "mmlu_eval_accuracy_college_chemistry": 0.375, |
| "mmlu_eval_accuracy_college_computer_science": 0.7272727272727273, |
| "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.5454545454545454, |
| "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.6875, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, |
| "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, |
| "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": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, |
| "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, |
| "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, |
| "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, |
| "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, |
| "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112, |
| "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, |
| "mmlu_eval_accuracy_management": 0.9090909090909091, |
| "mmlu_eval_accuracy_marketing": 0.88, |
| "mmlu_eval_accuracy_medical_genetics": 1.0, |
| "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, |
| "mmlu_eval_accuracy_moral_disputes": 0.5, |
| "mmlu_eval_accuracy_moral_scenarios": 0.45, |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, |
| "mmlu_eval_accuracy_philosophy": 0.7058823529411765, |
| "mmlu_eval_accuracy_prehistory": 0.5428571428571428, |
| "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, |
| "mmlu_eval_accuracy_professional_law": 0.40588235294117647, |
| "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, |
| "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, |
| "mmlu_eval_accuracy_public_relations": 0.4166666666666667, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, |
| "mmlu_eval_accuracy_virology": 0.5555555555555556, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.1781278766944576, |
| "step": 2800 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2323, |
| "step": 2801 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2335, |
| "step": 2802 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2922, |
| "step": 2803 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2546, |
| "step": 2804 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.307, |
| "step": 2805 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2169, |
| "step": 2806 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3367, |
| "step": 2807 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.193, |
| "step": 2808 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2419, |
| "step": 2809 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2932, |
| "step": 2810 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3043, |
| "step": 2811 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.3141, |
| "step": 2812 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4644, |
| "step": 2813 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.394, |
| "step": 2814 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.1683, |
| "step": 2815 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2328, |
| "step": 2816 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2948, |
| "step": 2817 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.345, |
| "step": 2818 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2388, |
| "step": 2819 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.2671, |
| "step": 2820 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 0.0002, |
| "loss": 0.4239, |
| "step": 2821 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2759, |
| "step": 2822 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.4436, |
| "step": 2823 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3438, |
| "step": 2824 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3013, |
| "step": 2825 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3897, |
| "step": 2826 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3117, |
| "step": 2827 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.356, |
| "step": 2828 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3203, |
| "step": 2829 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3263, |
| "step": 2830 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2972, |
| "step": 2831 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2773, |
| "step": 2832 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.1933, |
| "step": 2833 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.271, |
| "step": 2834 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3119, |
| "step": 2835 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2964, |
| "step": 2836 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.2103, |
| "step": 2837 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.342, |
| "step": 2838 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3252, |
| "step": 2839 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3068, |
| "step": 2840 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3078, |
| "step": 2841 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3497, |
| "step": 2842 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3023, |
| "step": 2843 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.192, |
| "step": 2844 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3317, |
| "step": 2845 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3274, |
| "step": 2846 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3762, |
| "step": 2847 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 0.0002, |
| "loss": 0.3169, |
| "step": 2848 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2757, |
| "step": 2849 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2847, |
| "step": 2850 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4609, |
| "step": 2851 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2406, |
| "step": 2852 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3228, |
| "step": 2853 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4058, |
| "step": 2854 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4722, |
| "step": 2855 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3736, |
| "step": 2856 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2614, |
| "step": 2857 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2086, |
| "step": 2858 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4341, |
| "step": 2859 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3702, |
| "step": 2860 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4001, |
| "step": 2861 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3323, |
| "step": 2862 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3294, |
| "step": 2863 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4903, |
| "step": 2864 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3946, |
| "step": 2865 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3521, |
| "step": 2866 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3517, |
| "step": 2867 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2101, |
| "step": 2868 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.3954, |
| "step": 2869 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2623, |
| "step": 2870 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.2688, |
| "step": 2871 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4493, |
| "step": 2872 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4535, |
| "step": 2873 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 0.0002, |
| "loss": 0.4178, |
| "step": 2874 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2571, |
| "step": 2875 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3075, |
| "step": 2876 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2683, |
| "step": 2877 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3076, |
| "step": 2878 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4161, |
| "step": 2879 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3623, |
| "step": 2880 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2654, |
| "step": 2881 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2897, |
| "step": 2882 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.269, |
| "step": 2883 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3755, |
| "step": 2884 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3471, |
| "step": 2885 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2919, |
| "step": 2886 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3532, |
| "step": 2887 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3141, |
| "step": 2888 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.5265, |
| "step": 2889 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.254, |
| "step": 2890 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3492, |
| "step": 2891 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3106, |
| "step": 2892 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3279, |
| "step": 2893 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3111, |
| "step": 2894 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2656, |
| "step": 2895 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3164, |
| "step": 2896 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.4187, |
| "step": 2897 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.308, |
| "step": 2898 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.2305, |
| "step": 2899 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.3586, |
| "step": 2900 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 0.0002, |
| "loss": 0.313, |
| "step": 2901 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4055, |
| "step": 2902 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3894, |
| "step": 2903 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3831, |
| "step": 2904 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4404, |
| "step": 2905 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2401, |
| "step": 2906 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2551, |
| "step": 2907 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4506, |
| "step": 2908 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2039, |
| "step": 2909 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2228, |
| "step": 2910 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2832, |
| "step": 2911 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.4078, |
| "step": 2912 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2692, |
| "step": 2913 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2432, |
| "step": 2914 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3321, |
| "step": 2915 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2188, |
| "step": 2916 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3575, |
| "step": 2917 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.1912, |
| "step": 2918 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2799, |
| "step": 2919 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3351, |
| "step": 2920 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3747, |
| "step": 2921 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2935, |
| "step": 2922 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2628, |
| "step": 2923 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3002, |
| "step": 2924 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2819, |
| "step": 2925 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.2781, |
| "step": 2926 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 0.0002, |
| "loss": 0.3801, |
| "step": 2927 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3979, |
| "step": 2928 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3066, |
| "step": 2929 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3936, |
| "step": 2930 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3342, |
| "step": 2931 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3349, |
| "step": 2932 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3188, |
| "step": 2933 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3038, |
| "step": 2934 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4298, |
| "step": 2935 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.376, |
| "step": 2936 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2475, |
| "step": 2937 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2411, |
| "step": 2938 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3738, |
| "step": 2939 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3013, |
| "step": 2940 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2054, |
| "step": 2941 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2717, |
| "step": 2942 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2846, |
| "step": 2943 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2674, |
| "step": 2944 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2663, |
| "step": 2945 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.4248, |
| "step": 2946 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3811, |
| "step": 2947 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3999, |
| "step": 2948 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2909, |
| "step": 2949 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3134, |
| "step": 2950 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.3653, |
| "step": 2951 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.2616, |
| "step": 2952 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.5204, |
| "step": 2953 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 0.0002, |
| "loss": 0.401, |
| "step": 2954 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3741, |
| "step": 2955 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2217, |
| "step": 2956 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2647, |
| "step": 2957 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2712, |
| "step": 2958 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3024, |
| "step": 2959 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3206, |
| "step": 2960 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2514, |
| "step": 2961 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3216, |
| "step": 2962 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3359, |
| "step": 2963 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2509, |
| "step": 2964 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.231, |
| "step": 2965 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.419, |
| "step": 2966 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3339, |
| "step": 2967 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.1886, |
| "step": 2968 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2665, |
| "step": 2969 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3003, |
| "step": 2970 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3198, |
| "step": 2971 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3534, |
| "step": 2972 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3669, |
| "step": 2973 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3065, |
| "step": 2974 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2236, |
| "step": 2975 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.346, |
| "step": 2976 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3511, |
| "step": 2977 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3114, |
| "step": 2978 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.2342, |
| "step": 2979 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 0.0002, |
| "loss": 0.3354, |
| "step": 2980 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3256, |
| "step": 2981 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.4458, |
| "step": 2982 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2968, |
| "step": 2983 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3186, |
| "step": 2984 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2455, |
| "step": 2985 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3452, |
| "step": 2986 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2324, |
| "step": 2987 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2485, |
| "step": 2988 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3671, |
| "step": 2989 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.333, |
| "step": 2990 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2334, |
| "step": 2991 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.256, |
| "step": 2992 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2344, |
| "step": 2993 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2412, |
| "step": 2994 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2087, |
| "step": 2995 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.3312, |
| "step": 2996 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2723, |
| "step": 2997 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.337, |
| "step": 2998 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2827, |
| "step": 2999 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 0.0002, |
| "loss": 0.2386, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.13, |
| "eval_loss": 0.3897075057029724, |
| "eval_runtime": 109.284, |
| "eval_samples_per_second": 9.15, |
| "eval_steps_per_second": 4.575, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.13, |
| "mmlu_eval_accuracy": 0.5749697229957464, |
| "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, |
| "mmlu_eval_accuracy_anatomy": 0.42857142857142855, |
| "mmlu_eval_accuracy_astronomy": 0.625, |
| "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, |
| "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, |
| "mmlu_eval_accuracy_college_biology": 0.5625, |
| "mmlu_eval_accuracy_college_chemistry": 0.25, |
| "mmlu_eval_accuracy_college_computer_science": 0.6363636363636364, |
| "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, |
| "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, |
| "mmlu_eval_accuracy_college_physics": 0.5454545454545454, |
| "mmlu_eval_accuracy_computer_security": 0.45454545454545453, |
| "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, |
| "mmlu_eval_accuracy_econometrics": 0.5, |
| "mmlu_eval_accuracy_electrical_engineering": 0.5625, |
| "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, |
| "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, |
| "mmlu_eval_accuracy_global_facts": 0.4, |
| "mmlu_eval_accuracy_high_school_biology": 0.53125, |
| "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, |
| "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, |
| "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666, |
| "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143, |
| "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, |
| "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, |
| "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, |
| "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": 0.8181818181818182, |
| "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, |
| "mmlu_eval_accuracy_human_aging": 0.6956521739130435, |
| "mmlu_eval_accuracy_human_sexuality": 0.5, |
| "mmlu_eval_accuracy_international_law": 0.8461538461538461, |
| "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": 0.88, |
| "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": 0.4, |
| "mmlu_eval_accuracy_nutrition": 0.6363636363636364, |
| "mmlu_eval_accuracy_philosophy": 0.7647058823529411, |
| "mmlu_eval_accuracy_prehistory": 0.5142857142857142, |
| "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, |
| "mmlu_eval_accuracy_professional_law": 0.43529411764705883, |
| "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, |
| "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, |
| "mmlu_eval_accuracy_public_relations": 0.5, |
| "mmlu_eval_accuracy_security_studies": 0.5555555555555556, |
| "mmlu_eval_accuracy_sociology": 0.9090909090909091, |
| "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, |
| "mmlu_eval_accuracy_virology": 0.6111111111111112, |
| "mmlu_eval_accuracy_world_religions": 0.7894736842105263, |
| "mmlu_loss": 1.4570590901546006, |
| "step": 3000 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 7947, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 4.414019681015562e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
