| { |
| "best_metric": 0.42375433444976807, |
| "best_model_checkpoint": "experts/mistralic-expert-10/checkpoint-200", |
| "epoch": 0.07547881875648646, |
| "eval_steps": 200, |
| "global_step": 200, |
| "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 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 7947, |
| "num_train_epochs": 3, |
| "save_steps": 200, |
| "total_flos": 2.9607227602305024e+16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
