| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9825242718446603, | |
| "eval_steps": 500, | |
| "global_step": 192, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6655, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8229, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2e-05, | |
| "loss": 1.871, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9893, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0291, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2e-05, | |
| "loss": 1.983, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0649, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3775, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6722, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7044, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 2e-05, | |
| "loss": 1.77, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0366, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0772, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1213, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1209, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2e-05, | |
| "loss": 2.567, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6556, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8708, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0389, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9722, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 2e-05, | |
| "loss": 2.075, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0757, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0656, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4261, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7365, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8049, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9021, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9696, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1454, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9359, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2e-05, | |
| "loss": 2.2369, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2e-05, | |
| "loss": 2.309, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7485, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 2e-05, | |
| "loss": 1.805, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2e-05, | |
| "loss": 1.777, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9951, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1456, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0073, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1608, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2e-05, | |
| "loss": 2.2249, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7468, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7292, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8926, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9109, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0223, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0283, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9571, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4003, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9499, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7059, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7516, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9586, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0152, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1286, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1614, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2e-05, | |
| "loss": 2.168, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8767, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8243, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9965, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9171, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9598, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8569, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0991, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1616, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7902, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7404, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9132, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9342, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0537, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0116, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0901, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1829, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8606, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6824, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8978, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8608, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9369, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8742, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0519, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1078, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8818, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7438, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8832, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7988, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0218, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 2e-05, | |
| "loss": 1.949, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9499, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1105, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8723, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2e-05, | |
| "loss": 1.696, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7281, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8753, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0551, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8918, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 2e-05, | |
| "loss": 1.883, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0562, | |
| "step": 192 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 192, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 3.523807718510592e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |