diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,5469 @@ +{ + "best_metric": 180.1415557861328, + "best_model_checkpoint": "./ckpts/TinyLlama_v1.1_2bit_int_three_times_data/checkpoint-900", + "epoch": 3.0, + "eval_steps": 150, + "global_step": 900, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 0.0, + "loss": 5773.8271, + "step": 1 + }, + { + "epoch": 0.01, + "learning_rate": 2e-05, + "loss": 6361.1123, + "step": 2 + }, + { + "epoch": 0.01, + "learning_rate": 2e-05, + "loss": 5510.5137, + "step": 3 + }, + { + "epoch": 0.01, + "learning_rate": 2e-05, + "loss": 2483.1978, + "step": 4 + }, + { + "epoch": 0.02, + "learning_rate": 2e-05, + "loss": 2523.3491, + "step": 5 + }, + { + "epoch": 0.02, + "learning_rate": 2e-05, + "loss": 2417.8442, + "step": 6 + }, + { + "epoch": 0.02, + "learning_rate": 2e-05, + "loss": 855.7811, + "step": 7 + }, + { + "epoch": 0.03, + "learning_rate": 2e-05, + "loss": 1127.734, + "step": 8 + }, + { + "epoch": 0.03, + "learning_rate": 2e-05, + "loss": 940.725, + "step": 9 + }, + { + "epoch": 0.03, + "learning_rate": 2e-05, + "loss": 783.8114, + "step": 10 + }, + { + "epoch": 0.04, + "learning_rate": 2e-05, + "loss": 720.3327, + "step": 11 + }, + { + "epoch": 0.04, + "learning_rate": 2e-05, + "loss": 653.4684, + "step": 12 + }, + { + "epoch": 0.04, + "learning_rate": 2e-05, + "loss": 565.9368, + "step": 13 + }, + { + "epoch": 0.05, + "learning_rate": 2e-05, + "loss": 776.7609, + "step": 14 + }, + { + "epoch": 0.05, + "learning_rate": 2e-05, + "loss": 986.0417, + "step": 15 + }, + { + "epoch": 0.05, + "learning_rate": 2e-05, + "loss": 590.2695, + "step": 16 + }, + { + "epoch": 0.06, + "learning_rate": 2e-05, + "loss": 485.4177, + "step": 17 + }, + { + "epoch": 0.06, + "learning_rate": 2e-05, + "loss": 493.4467, + "step": 18 + }, + { + "epoch": 0.06, + "learning_rate": 2e-05, + "loss": 421.5482, + "step": 19 + }, + { + "epoch": 0.07, + "learning_rate": 2e-05, + "loss": 475.7487, + "step": 20 + }, + { + "epoch": 0.07, + "learning_rate": 2e-05, + "loss": 395.1753, + "step": 21 + }, + { + "epoch": 0.07, + "learning_rate": 2e-05, + "loss": 399.0683, + "step": 22 + }, + { + "epoch": 0.08, + "learning_rate": 2e-05, + "loss": 449.912, + "step": 23 + }, + { + "epoch": 0.08, + "learning_rate": 2e-05, + "loss": 505.1134, + "step": 24 + }, + { + "epoch": 0.08, + "learning_rate": 2e-05, + "loss": 423.8591, + "step": 25 + }, + { + "epoch": 0.09, + "learning_rate": 2e-05, + "loss": 471.5508, + "step": 26 + }, + { + "epoch": 0.09, + "learning_rate": 2e-05, + "loss": 327.645, + "step": 27 + }, + { + "epoch": 0.09, + "learning_rate": 2e-05, + "loss": 402.1392, + "step": 28 + }, + { + "epoch": 0.1, + "learning_rate": 2e-05, + "loss": 353.8039, + "step": 29 + }, + { + "epoch": 0.1, + "learning_rate": 2e-05, + "loss": 366.3328, + "step": 30 + }, + { + "epoch": 0.1, + "learning_rate": 2e-05, + "loss": 395.2348, + "step": 31 + }, + { + "epoch": 0.11, + "learning_rate": 2e-05, + "loss": 370.4788, + "step": 32 + }, + { + "epoch": 0.11, + "learning_rate": 2e-05, + "loss": 389.9211, + "step": 33 + }, + { + "epoch": 0.11, + "learning_rate": 2e-05, + "loss": 338.4554, + "step": 34 + }, + { + "epoch": 0.12, + "learning_rate": 2e-05, + "loss": 416.246, + "step": 35 + }, + { + "epoch": 0.12, + "learning_rate": 2e-05, + "loss": 335.4657, + "step": 36 + }, + { + "epoch": 0.12, + "learning_rate": 2e-05, + "loss": 324.8686, + "step": 37 + }, + { + "epoch": 0.13, + "learning_rate": 2e-05, + "loss": 359.753, + "step": 38 + }, + { + "epoch": 0.13, + "learning_rate": 2e-05, + "loss": 404.8, + "step": 39 + }, + { + "epoch": 0.13, + "learning_rate": 2e-05, + "loss": 458.0917, + "step": 40 + }, + { + "epoch": 0.14, + "learning_rate": 2e-05, + "loss": 353.8201, + "step": 41 + }, + { + "epoch": 0.14, + "learning_rate": 2e-05, + "loss": 355.3937, + "step": 42 + }, + { + "epoch": 0.14, + "learning_rate": 2e-05, + "loss": 361.6961, + "step": 43 + }, + { + "epoch": 0.15, + "learning_rate": 2e-05, + "loss": 384.7435, + "step": 44 + }, + { + "epoch": 0.15, + "learning_rate": 2e-05, + "loss": 341.7866, + "step": 45 + }, + { + "epoch": 0.15, + "learning_rate": 2e-05, + "loss": 325.0611, + "step": 46 + }, + { + "epoch": 0.16, + "learning_rate": 2e-05, + "loss": 373.5464, + "step": 47 + }, + { + "epoch": 0.16, + "learning_rate": 2e-05, + "loss": 391.3689, + "step": 48 + }, + { + "epoch": 0.16, + "learning_rate": 2e-05, + "loss": 369.0639, + "step": 49 + }, + { + "epoch": 0.17, + "learning_rate": 2e-05, + "loss": 331.4485, + "step": 50 + }, + { + "epoch": 0.17, + "learning_rate": 2e-05, + "loss": 386.4754, + "step": 51 + }, + { + "epoch": 0.17, + "learning_rate": 2e-05, + "loss": 300.7639, + "step": 52 + }, + { + "epoch": 0.18, + "learning_rate": 2e-05, + "loss": 309.7045, + "step": 53 + }, + { + "epoch": 0.18, + "learning_rate": 2e-05, + "loss": 427.1003, + "step": 54 + }, + { + "epoch": 0.18, + "learning_rate": 2e-05, + "loss": 491.0206, + "step": 55 + }, + { + "epoch": 0.19, + "learning_rate": 2e-05, + "loss": 305.5411, + "step": 56 + }, + { + "epoch": 0.19, + "learning_rate": 2e-05, + "loss": 325.358, + "step": 57 + }, + { + "epoch": 0.19, + "learning_rate": 2e-05, + "loss": 383.5042, + "step": 58 + }, + { + "epoch": 0.2, + "learning_rate": 2e-05, + "loss": 495.7033, + "step": 59 + }, + { + "epoch": 0.2, + "learning_rate": 2e-05, + "loss": 359.8436, + "step": 60 + }, + { + "epoch": 0.2, + "learning_rate": 2e-05, + "loss": 335.7679, + "step": 61 + }, + { + "epoch": 0.21, + "learning_rate": 2e-05, + "loss": 333.6547, + "step": 62 + }, + { + "epoch": 0.21, + "learning_rate": 2e-05, + "loss": 369.515, + "step": 63 + }, + { + "epoch": 0.21, + "learning_rate": 2e-05, + "loss": 501.9354, + "step": 64 + }, + { + "epoch": 0.22, + "learning_rate": 2e-05, + "loss": 348.7731, + "step": 65 + }, + { + "epoch": 0.22, + "learning_rate": 2e-05, + "loss": 407.3716, + "step": 66 + }, + { + "epoch": 0.22, + "learning_rate": 2e-05, + "loss": 388.8562, + "step": 67 + }, + { + "epoch": 0.23, + "learning_rate": 2e-05, + "loss": 402.8853, + "step": 68 + }, + { + "epoch": 0.23, + "learning_rate": 2e-05, + "loss": 325.8807, + "step": 69 + }, + { + "epoch": 0.23, + "learning_rate": 2e-05, + "loss": 358.7309, + "step": 70 + }, + { + "epoch": 0.24, + "learning_rate": 2e-05, + "loss": 314.6354, + "step": 71 + }, + { + "epoch": 0.24, + "learning_rate": 2e-05, + "loss": 356.9031, + "step": 72 + }, + { + "epoch": 0.24, + "learning_rate": 2e-05, + "loss": 278.4463, + "step": 73 + }, + { + "epoch": 0.25, + "learning_rate": 2e-05, + "loss": 302.7334, + "step": 74 + }, + { + "epoch": 0.25, + "learning_rate": 2e-05, + "loss": 347.0969, + "step": 75 + }, + { + "epoch": 0.25, + "learning_rate": 2e-05, + "loss": 281.1858, + "step": 76 + }, + { + "epoch": 0.26, + "learning_rate": 2e-05, + "loss": 300.5947, + "step": 77 + }, + { + "epoch": 0.26, + "learning_rate": 2e-05, + "loss": 326.8865, + "step": 78 + }, + { + "epoch": 0.26, + "learning_rate": 2e-05, + "loss": 334.2254, + "step": 79 + }, + { + "epoch": 0.27, + "learning_rate": 2e-05, + "loss": 358.398, + "step": 80 + }, + { + "epoch": 0.27, + "learning_rate": 2e-05, + "loss": 380.9495, + "step": 81 + }, + { + "epoch": 0.27, + "learning_rate": 2e-05, + "loss": 252.83, + "step": 82 + }, + { + "epoch": 0.28, + "learning_rate": 2e-05, + "loss": 332.8982, + "step": 83 + }, + { + "epoch": 0.28, + "learning_rate": 2e-05, + "loss": 364.0591, + "step": 84 + }, + { + "epoch": 0.28, + "learning_rate": 2e-05, + "loss": 353.0194, + "step": 85 + }, + { + "epoch": 0.29, + "learning_rate": 2e-05, + "loss": 314.4102, + "step": 86 + }, + { + "epoch": 0.29, + "learning_rate": 2e-05, + "loss": 332.9731, + "step": 87 + }, + { + "epoch": 0.29, + "learning_rate": 2e-05, + "loss": 364.2653, + "step": 88 + }, + { + "epoch": 0.3, + "learning_rate": 2e-05, + "loss": 259.5952, + "step": 89 + }, + { + "epoch": 0.3, + "learning_rate": 2e-05, + "loss": 254.9917, + "step": 90 + }, + { + "epoch": 0.3, + "learning_rate": 2e-05, + "loss": 256.5221, + "step": 91 + }, + { + "epoch": 0.31, + "learning_rate": 2e-05, + "loss": 273.1299, + "step": 92 + }, + { + "epoch": 0.31, + "learning_rate": 2e-05, + "loss": 289.5511, + "step": 93 + }, + { + "epoch": 0.31, + "learning_rate": 2e-05, + "loss": 292.8752, + "step": 94 + }, + { + "epoch": 0.32, + "learning_rate": 2e-05, + "loss": 290.5632, + "step": 95 + }, + { + "epoch": 0.32, + "learning_rate": 2e-05, + "loss": 348.9591, + "step": 96 + }, + { + "epoch": 0.32, + "learning_rate": 2e-05, + "loss": 310.7304, + "step": 97 + }, + { + "epoch": 0.33, + "learning_rate": 2e-05, + "loss": 355.2693, + "step": 98 + }, + { + "epoch": 0.33, + "learning_rate": 2e-05, + "loss": 295.0119, + "step": 99 + }, + { + "epoch": 0.33, + "learning_rate": 2e-05, + "loss": 281.4653, + "step": 100 + }, + { + "epoch": 0.34, + "learning_rate": 2e-05, + "loss": 325.7113, + "step": 101 + }, + { + "epoch": 0.34, + "learning_rate": 2e-05, + "loss": 310.1772, + "step": 102 + }, + { + "epoch": 0.34, + "learning_rate": 2e-05, + "loss": 289.4645, + "step": 103 + }, + { + "epoch": 0.35, + "learning_rate": 2e-05, + "loss": 284.9719, + "step": 104 + }, + { + "epoch": 0.35, + "learning_rate": 2e-05, + "loss": 291.3521, + "step": 105 + }, + { + "epoch": 0.35, + "learning_rate": 2e-05, + "loss": 294.0998, + "step": 106 + }, + { + "epoch": 0.36, + "learning_rate": 2e-05, + "loss": 315.3473, + "step": 107 + }, + { + "epoch": 0.36, + "learning_rate": 2e-05, + "loss": 386.5405, + "step": 108 + }, + { + "epoch": 0.36, + "learning_rate": 2e-05, + "loss": 324.626, + "step": 109 + }, + { + "epoch": 0.37, + "learning_rate": 2e-05, + "loss": 309.3942, + "step": 110 + }, + { + "epoch": 0.37, + "learning_rate": 2e-05, + "loss": 332.9089, + "step": 111 + }, + { + "epoch": 0.37, + "learning_rate": 2e-05, + "loss": 318.4434, + "step": 112 + }, + { + "epoch": 0.38, + "learning_rate": 2e-05, + "loss": 279.5007, + "step": 113 + }, + { + "epoch": 0.38, + "learning_rate": 2e-05, + "loss": 299.7087, + "step": 114 + }, + { + "epoch": 0.38, + "learning_rate": 2e-05, + "loss": 339.1145, + "step": 115 + }, + { + "epoch": 0.39, + "learning_rate": 2e-05, + "loss": 313.114, + "step": 116 + }, + { + "epoch": 0.39, + "learning_rate": 2e-05, + "loss": 384.9408, + "step": 117 + }, + { + "epoch": 0.39, + "learning_rate": 2e-05, + "loss": 329.4101, + "step": 118 + }, + { + "epoch": 0.4, + "learning_rate": 2e-05, + "loss": 266.9434, + "step": 119 + }, + { + "epoch": 0.4, + "learning_rate": 2e-05, + "loss": 349.3153, + "step": 120 + }, + { + "epoch": 0.4, + "learning_rate": 2e-05, + "loss": 299.2152, + "step": 121 + }, + { + "epoch": 0.41, + "learning_rate": 2e-05, + "loss": 254.6638, + "step": 122 + }, + { + "epoch": 0.41, + "learning_rate": 2e-05, + "loss": 353.7803, + "step": 123 + }, + { + "epoch": 0.41, + "learning_rate": 2e-05, + "loss": 336.8068, + "step": 124 + }, + { + "epoch": 0.42, + "learning_rate": 2e-05, + "loss": 244.8727, + "step": 125 + }, + { + "epoch": 0.42, + "learning_rate": 2e-05, + "loss": 250.9102, + "step": 126 + }, + { + "epoch": 0.42, + "learning_rate": 2e-05, + "loss": 306.2723, + "step": 127 + }, + { + "epoch": 0.43, + "learning_rate": 2e-05, + "loss": 297.1634, + "step": 128 + }, + { + "epoch": 0.43, + "learning_rate": 2e-05, + "loss": 300.8284, + "step": 129 + }, + { + "epoch": 0.43, + "learning_rate": 2e-05, + "loss": 346.2424, + "step": 130 + }, + { + "epoch": 0.44, + "learning_rate": 2e-05, + "loss": 317.494, + "step": 131 + }, + { + "epoch": 0.44, + "learning_rate": 2e-05, + "loss": 278.0762, + "step": 132 + }, + { + "epoch": 0.44, + "learning_rate": 2e-05, + "loss": 387.448, + "step": 133 + }, + { + "epoch": 0.45, + "learning_rate": 2e-05, + "loss": 295.659, + "step": 134 + }, + { + "epoch": 0.45, + "learning_rate": 2e-05, + "loss": 248.9755, + "step": 135 + }, + { + "epoch": 0.45, + "learning_rate": 2e-05, + "loss": 276.8231, + "step": 136 + }, + { + "epoch": 0.46, + "learning_rate": 2e-05, + "loss": 240.1871, + "step": 137 + }, + { + "epoch": 0.46, + "learning_rate": 2e-05, + "loss": 286.2157, + "step": 138 + }, + { + "epoch": 0.46, + "learning_rate": 2e-05, + "loss": 264.3008, + "step": 139 + }, + { + "epoch": 0.47, + "learning_rate": 2e-05, + "loss": 275.0594, + "step": 140 + }, + { + "epoch": 0.47, + "learning_rate": 2e-05, + "loss": 318.6325, + "step": 141 + }, + { + "epoch": 0.47, + "learning_rate": 2e-05, + "loss": 343.2183, + "step": 142 + }, + { + "epoch": 0.48, + "learning_rate": 2e-05, + "loss": 286.0394, + "step": 143 + }, + { + "epoch": 0.48, + "learning_rate": 2e-05, + "loss": 388.476, + "step": 144 + }, + { + "epoch": 0.48, + "learning_rate": 2e-05, + "loss": 324.6236, + "step": 145 + }, + { + "epoch": 0.49, + "learning_rate": 2e-05, + "loss": 306.4463, + "step": 146 + }, + { + "epoch": 0.49, + "learning_rate": 2e-05, + "loss": 320.8977, + "step": 147 + }, + { + "epoch": 0.49, + "learning_rate": 2e-05, + "loss": 321.3353, + "step": 148 + }, + { + "epoch": 0.5, + "learning_rate": 2e-05, + "loss": 283.8933, + "step": 149 + }, + { + "epoch": 0.5, + "learning_rate": 2e-05, + "loss": 277.131, + "step": 150 + }, + { + "epoch": 0.5, + "eval_loss": 311.9327087402344, + "eval_runtime": 287.3484, + "eval_samples_per_second": 16.704, + "eval_steps_per_second": 1.044, + "step": 150 + }, + { + "epoch": 0.5, + "learning_rate": 2e-05, + "loss": 369.017, + "step": 151 + }, + { + "epoch": 0.51, + "learning_rate": 2e-05, + "loss": 333.5858, + "step": 152 + }, + { + "epoch": 0.51, + "learning_rate": 2e-05, + "loss": 260.8406, + "step": 153 + }, + { + "epoch": 0.51, + "learning_rate": 2e-05, + "loss": 406.8607, + "step": 154 + }, + { + "epoch": 0.52, + "learning_rate": 2e-05, + "loss": 294.3548, + "step": 155 + }, + { + "epoch": 0.52, + "learning_rate": 2e-05, + "loss": 234.9642, + "step": 156 + }, + { + "epoch": 0.52, + "learning_rate": 2e-05, + "loss": 355.0019, + "step": 157 + }, + { + "epoch": 0.53, + "learning_rate": 2e-05, + "loss": 214.566, + "step": 158 + }, + { + "epoch": 0.53, + "learning_rate": 2e-05, + "loss": 289.3258, + "step": 159 + }, + { + "epoch": 0.53, + "learning_rate": 2e-05, + "loss": 280.4387, + "step": 160 + }, + { + "epoch": 0.54, + "learning_rate": 2e-05, + "loss": 308.4476, + "step": 161 + }, + { + "epoch": 0.54, + "learning_rate": 2e-05, + "loss": 305.3412, + "step": 162 + }, + { + "epoch": 0.54, + "learning_rate": 2e-05, + "loss": 316.1368, + "step": 163 + }, + { + "epoch": 0.55, + "learning_rate": 2e-05, + "loss": 314.1208, + "step": 164 + }, + { + "epoch": 0.55, + "learning_rate": 2e-05, + "loss": 413.5666, + "step": 165 + }, + { + "epoch": 0.55, + "learning_rate": 2e-05, + "loss": 272.8596, + "step": 166 + }, + { + "epoch": 0.56, + "learning_rate": 2e-05, + "loss": 307.2043, + "step": 167 + }, + { + "epoch": 0.56, + "learning_rate": 2e-05, + "loss": 318.3997, + "step": 168 + }, + { + "epoch": 0.56, + "learning_rate": 2e-05, + "loss": 328.7892, + "step": 169 + }, + { + "epoch": 0.57, + "learning_rate": 2e-05, + "loss": 322.48, + "step": 170 + }, + { + "epoch": 0.57, + "learning_rate": 2e-05, + "loss": 296.6621, + "step": 171 + }, + { + "epoch": 0.57, + "learning_rate": 2e-05, + "loss": 316.1772, + "step": 172 + }, + { + "epoch": 0.58, + "learning_rate": 2e-05, + "loss": 240.0376, + "step": 173 + }, + { + "epoch": 0.58, + "learning_rate": 2e-05, + "loss": 295.6813, + "step": 174 + }, + { + "epoch": 0.58, + "learning_rate": 2e-05, + "loss": 318.4695, + "step": 175 + }, + { + "epoch": 0.59, + "learning_rate": 2e-05, + "loss": 307.9428, + "step": 176 + }, + { + "epoch": 0.59, + "learning_rate": 2e-05, + "loss": 265.1976, + "step": 177 + }, + { + "epoch": 0.59, + "learning_rate": 2e-05, + "loss": 327.1461, + "step": 178 + }, + { + "epoch": 0.6, + "learning_rate": 2e-05, + "loss": 288.329, + "step": 179 + }, + { + "epoch": 0.6, + "learning_rate": 2e-05, + "loss": 285.4697, + "step": 180 + }, + { + "epoch": 0.6, + "learning_rate": 2e-05, + "loss": 263.3636, + "step": 181 + }, + { + "epoch": 0.61, + "learning_rate": 2e-05, + "loss": 266.7616, + "step": 182 + }, + { + "epoch": 0.61, + "learning_rate": 2e-05, + "loss": 278.2103, + "step": 183 + }, + { + "epoch": 0.61, + "learning_rate": 2e-05, + "loss": 262.6075, + "step": 184 + }, + { + "epoch": 0.62, + "learning_rate": 2e-05, + "loss": 301.3991, + "step": 185 + }, + { + "epoch": 0.62, + "learning_rate": 2e-05, + "loss": 232.6239, + "step": 186 + }, + { + "epoch": 0.62, + "learning_rate": 2e-05, + "loss": 297.678, + "step": 187 + }, + { + "epoch": 0.63, + "learning_rate": 2e-05, + "loss": 221.4155, + "step": 188 + }, + { + "epoch": 0.63, + "learning_rate": 2e-05, + "loss": 322.8382, + "step": 189 + }, + { + "epoch": 0.63, + "learning_rate": 2e-05, + "loss": 291.5475, + "step": 190 + }, + { + "epoch": 0.64, + "learning_rate": 2e-05, + "loss": 243.8622, + "step": 191 + }, + { + "epoch": 0.64, + "learning_rate": 2e-05, + "loss": 303.3617, + "step": 192 + }, + { + "epoch": 0.64, + "learning_rate": 2e-05, + "loss": 240.6237, + "step": 193 + }, + { + "epoch": 0.65, + "learning_rate": 2e-05, + "loss": 224.548, + "step": 194 + }, + { + "epoch": 0.65, + "learning_rate": 2e-05, + "loss": 260.3085, + "step": 195 + }, + { + "epoch": 0.65, + "learning_rate": 2e-05, + "loss": 247.1713, + "step": 196 + }, + { + "epoch": 0.66, + "learning_rate": 2e-05, + "loss": 236.0542, + "step": 197 + }, + { + "epoch": 0.66, + "learning_rate": 2e-05, + "loss": 254.8948, + "step": 198 + }, + { + "epoch": 0.66, + "learning_rate": 2e-05, + "loss": 230.9995, + "step": 199 + }, + { + "epoch": 0.67, + "learning_rate": 2e-05, + "loss": 275.4061, + "step": 200 + }, + { + "epoch": 0.67, + "learning_rate": 2e-05, + "loss": 280.2545, + "step": 201 + }, + { + "epoch": 0.67, + "learning_rate": 2e-05, + "loss": 321.3388, + "step": 202 + }, + { + "epoch": 0.68, + "learning_rate": 2e-05, + "loss": 271.4129, + "step": 203 + }, + { + "epoch": 0.68, + "learning_rate": 2e-05, + "loss": 246.0401, + "step": 204 + }, + { + "epoch": 0.68, + "learning_rate": 2e-05, + "loss": 258.4705, + "step": 205 + }, + { + "epoch": 0.69, + "learning_rate": 2e-05, + "loss": 310.4536, + "step": 206 + }, + { + "epoch": 0.69, + "learning_rate": 2e-05, + "loss": 259.5285, + "step": 207 + }, + { + "epoch": 0.69, + "learning_rate": 2e-05, + "loss": 237.5303, + "step": 208 + }, + { + "epoch": 0.7, + "learning_rate": 2e-05, + "loss": 278.5979, + "step": 209 + }, + { + "epoch": 0.7, + "learning_rate": 2e-05, + "loss": 285.5738, + "step": 210 + }, + { + "epoch": 0.7, + "learning_rate": 2e-05, + "loss": 271.1486, + "step": 211 + }, + { + "epoch": 0.71, + "learning_rate": 2e-05, + "loss": 247.188, + "step": 212 + }, + { + "epoch": 0.71, + "learning_rate": 2e-05, + "loss": 234.2194, + "step": 213 + }, + { + "epoch": 0.71, + "learning_rate": 2e-05, + "loss": 301.1833, + "step": 214 + }, + { + "epoch": 0.72, + "learning_rate": 2e-05, + "loss": 267.9955, + "step": 215 + }, + { + "epoch": 0.72, + "learning_rate": 2e-05, + "loss": 288.0046, + "step": 216 + }, + { + "epoch": 0.72, + "learning_rate": 2e-05, + "loss": 263.2744, + "step": 217 + }, + { + "epoch": 0.73, + "learning_rate": 2e-05, + "loss": 236.09, + "step": 218 + }, + { + "epoch": 0.73, + "learning_rate": 2e-05, + "loss": 273.5681, + "step": 219 + }, + { + "epoch": 0.73, + "learning_rate": 2e-05, + "loss": 314.3279, + "step": 220 + }, + { + "epoch": 0.74, + "learning_rate": 2e-05, + "loss": 240.7439, + "step": 221 + }, + { + "epoch": 0.74, + "learning_rate": 2e-05, + "loss": 275.4979, + "step": 222 + }, + { + "epoch": 0.74, + "learning_rate": 2e-05, + "loss": 240.7771, + "step": 223 + }, + { + "epoch": 0.75, + "learning_rate": 2e-05, + "loss": 260.4418, + "step": 224 + }, + { + "epoch": 0.75, + "learning_rate": 2e-05, + "loss": 234.2265, + "step": 225 + }, + { + "epoch": 0.75, + "learning_rate": 2e-05, + "loss": 270.6098, + "step": 226 + }, + { + "epoch": 0.76, + "learning_rate": 2e-05, + "loss": 229.4481, + "step": 227 + }, + { + "epoch": 0.76, + "learning_rate": 2e-05, + "loss": 306.5601, + "step": 228 + }, + { + "epoch": 0.76, + "learning_rate": 2e-05, + "loss": 232.1583, + "step": 229 + }, + { + "epoch": 0.77, + "learning_rate": 2e-05, + "loss": 236.2021, + "step": 230 + }, + { + "epoch": 0.77, + "learning_rate": 2e-05, + "loss": 261.0574, + "step": 231 + }, + { + "epoch": 0.77, + "learning_rate": 2e-05, + "loss": 264.9504, + "step": 232 + }, + { + "epoch": 0.78, + "learning_rate": 2e-05, + "loss": 235.4366, + "step": 233 + }, + { + "epoch": 0.78, + "learning_rate": 2e-05, + "loss": 264.1496, + "step": 234 + }, + { + "epoch": 0.78, + "learning_rate": 2e-05, + "loss": 242.2036, + "step": 235 + }, + { + "epoch": 0.79, + "learning_rate": 2e-05, + "loss": 353.3135, + "step": 236 + }, + { + "epoch": 0.79, + "learning_rate": 2e-05, + "loss": 210.8655, + "step": 237 + }, + { + "epoch": 0.79, + "learning_rate": 2e-05, + "loss": 303.5106, + "step": 238 + }, + { + "epoch": 0.8, + "learning_rate": 2e-05, + "loss": 284.8805, + "step": 239 + }, + { + "epoch": 0.8, + "learning_rate": 2e-05, + "loss": 297.9016, + "step": 240 + }, + { + "epoch": 0.8, + "learning_rate": 2e-05, + "loss": 282.7299, + "step": 241 + }, + { + "epoch": 0.81, + "learning_rate": 2e-05, + "loss": 248.2567, + "step": 242 + }, + { + "epoch": 0.81, + "learning_rate": 2e-05, + "loss": 272.2921, + "step": 243 + }, + { + "epoch": 0.81, + "learning_rate": 2e-05, + "loss": 239.2713, + "step": 244 + }, + { + "epoch": 0.82, + "learning_rate": 2e-05, + "loss": 266.6125, + "step": 245 + }, + { + "epoch": 0.82, + "learning_rate": 2e-05, + "loss": 255.2529, + "step": 246 + }, + { + "epoch": 0.82, + "learning_rate": 2e-05, + "loss": 301.0817, + "step": 247 + }, + { + "epoch": 0.83, + "learning_rate": 2e-05, + "loss": 289.9151, + "step": 248 + }, + { + "epoch": 0.83, + "learning_rate": 2e-05, + "loss": 269.2121, + "step": 249 + }, + { + "epoch": 0.83, + "learning_rate": 2e-05, + "loss": 255.5247, + "step": 250 + }, + { + "epoch": 0.84, + "learning_rate": 2e-05, + "loss": 247.6964, + "step": 251 + }, + { + "epoch": 0.84, + "learning_rate": 2e-05, + "loss": 217.559, + "step": 252 + }, + { + "epoch": 0.84, + "learning_rate": 2e-05, + "loss": 230.6058, + "step": 253 + }, + { + "epoch": 0.85, + "learning_rate": 2e-05, + "loss": 214.143, + "step": 254 + }, + { + "epoch": 0.85, + "learning_rate": 2e-05, + "loss": 284.8932, + "step": 255 + }, + { + "epoch": 0.85, + "learning_rate": 2e-05, + "loss": 235.6913, + "step": 256 + }, + { + "epoch": 0.86, + "learning_rate": 2e-05, + "loss": 284.4674, + "step": 257 + }, + { + "epoch": 0.86, + "learning_rate": 2e-05, + "loss": 288.0122, + "step": 258 + }, + { + "epoch": 0.86, + "learning_rate": 2e-05, + "loss": 248.2686, + "step": 259 + }, + { + "epoch": 0.87, + "learning_rate": 2e-05, + "loss": 279.7363, + "step": 260 + }, + { + "epoch": 0.87, + "learning_rate": 2e-05, + "loss": 308.7455, + "step": 261 + }, + { + "epoch": 0.87, + "learning_rate": 2e-05, + "loss": 277.4221, + "step": 262 + }, + { + "epoch": 0.88, + "learning_rate": 2e-05, + "loss": 254.3973, + "step": 263 + }, + { + "epoch": 0.88, + "learning_rate": 2e-05, + "loss": 254.7358, + "step": 264 + }, + { + "epoch": 0.88, + "learning_rate": 2e-05, + "loss": 233.0448, + "step": 265 + }, + { + "epoch": 0.89, + "learning_rate": 2e-05, + "loss": 235.5645, + "step": 266 + }, + { + "epoch": 0.89, + "learning_rate": 2e-05, + "loss": 295.3665, + "step": 267 + }, + { + "epoch": 0.89, + "learning_rate": 2e-05, + "loss": 270.3642, + "step": 268 + }, + { + "epoch": 0.9, + "learning_rate": 2e-05, + "loss": 279.1059, + "step": 269 + }, + { + "epoch": 0.9, + "learning_rate": 2e-05, + "loss": 222.4796, + "step": 270 + }, + { + "epoch": 0.9, + "learning_rate": 2e-05, + "loss": 227.0107, + "step": 271 + }, + { + "epoch": 0.91, + "learning_rate": 2e-05, + "loss": 244.201, + "step": 272 + }, + { + "epoch": 0.91, + "learning_rate": 2e-05, + "loss": 248.5634, + "step": 273 + }, + { + "epoch": 0.91, + "learning_rate": 2e-05, + "loss": 277.4857, + "step": 274 + }, + { + "epoch": 0.92, + "learning_rate": 2e-05, + "loss": 255.8076, + "step": 275 + }, + { + "epoch": 0.92, + "learning_rate": 2e-05, + "loss": 241.0695, + "step": 276 + }, + { + "epoch": 0.92, + "learning_rate": 2e-05, + "loss": 286.1827, + "step": 277 + }, + { + "epoch": 0.93, + "learning_rate": 2e-05, + "loss": 261.8386, + "step": 278 + }, + { + "epoch": 0.93, + "learning_rate": 2e-05, + "loss": 330.9406, + "step": 279 + }, + { + "epoch": 0.93, + "learning_rate": 2e-05, + "loss": 221.898, + "step": 280 + }, + { + "epoch": 0.94, + "learning_rate": 2e-05, + "loss": 223.2677, + "step": 281 + }, + { + "epoch": 0.94, + "learning_rate": 2e-05, + "loss": 236.0616, + "step": 282 + }, + { + "epoch": 0.94, + "learning_rate": 2e-05, + "loss": 218.6026, + "step": 283 + }, + { + "epoch": 0.95, + "learning_rate": 2e-05, + "loss": 273.3364, + "step": 284 + }, + { + "epoch": 0.95, + "learning_rate": 2e-05, + "loss": 278.8625, + "step": 285 + }, + { + "epoch": 0.95, + "learning_rate": 2e-05, + "loss": 224.1773, + "step": 286 + }, + { + "epoch": 0.96, + "learning_rate": 2e-05, + "loss": 264.6806, + "step": 287 + }, + { + "epoch": 0.96, + "learning_rate": 2e-05, + "loss": 245.4145, + "step": 288 + }, + { + "epoch": 0.96, + "learning_rate": 2e-05, + "loss": 207.8361, + "step": 289 + }, + { + "epoch": 0.97, + "learning_rate": 2e-05, + "loss": 290.9304, + "step": 290 + }, + { + "epoch": 0.97, + "learning_rate": 2e-05, + "loss": 221.5088, + "step": 291 + }, + { + "epoch": 0.97, + "learning_rate": 2e-05, + "loss": 224.0763, + "step": 292 + }, + { + "epoch": 0.98, + "learning_rate": 2e-05, + "loss": 249.4179, + "step": 293 + }, + { + "epoch": 0.98, + "learning_rate": 2e-05, + "loss": 236.3529, + "step": 294 + }, + { + "epoch": 0.98, + "learning_rate": 2e-05, + "loss": 259.7051, + "step": 295 + }, + { + "epoch": 0.99, + "learning_rate": 2e-05, + "loss": 337.4284, + "step": 296 + }, + { + "epoch": 0.99, + "learning_rate": 2e-05, + "loss": 282.2198, + "step": 297 + }, + { + "epoch": 0.99, + "learning_rate": 2e-05, + "loss": 206.7412, + "step": 298 + }, + { + "epoch": 1.0, + "learning_rate": 2e-05, + "loss": 270.6522, + "step": 299 + }, + { + "epoch": 1.0, + "learning_rate": 2e-05, + "loss": 251.5347, + "step": 300 + }, + { + "epoch": 1.0, + "eval_loss": 247.680419921875, + "eval_runtime": 287.8885, + "eval_samples_per_second": 16.673, + "eval_steps_per_second": 1.042, + "step": 300 + }, + { + "epoch": 1.0, + "learning_rate": 2e-05, + "loss": 245.9543, + "step": 301 + }, + { + "epoch": 1.01, + "learning_rate": 2e-05, + "loss": 215.307, + "step": 302 + }, + { + "epoch": 1.01, + "learning_rate": 2e-05, + "loss": 232.8804, + "step": 303 + }, + { + "epoch": 1.01, + "learning_rate": 2e-05, + "loss": 270.2068, + "step": 304 + }, + { + "epoch": 1.02, + "learning_rate": 2e-05, + "loss": 210.0473, + "step": 305 + }, + { + "epoch": 1.02, + "learning_rate": 2e-05, + "loss": 219.887, + "step": 306 + }, + { + "epoch": 1.02, + "learning_rate": 2e-05, + "loss": 234.6451, + "step": 307 + }, + { + "epoch": 1.03, + "learning_rate": 2e-05, + "loss": 241.5837, + "step": 308 + }, + { + "epoch": 1.03, + "learning_rate": 2e-05, + "loss": 236.1227, + "step": 309 + }, + { + "epoch": 1.03, + "learning_rate": 2e-05, + "loss": 211.8941, + "step": 310 + }, + { + "epoch": 1.04, + "learning_rate": 2e-05, + "loss": 204.086, + "step": 311 + }, + { + "epoch": 1.04, + "learning_rate": 2e-05, + "loss": 353.9028, + "step": 312 + }, + { + "epoch": 1.04, + "learning_rate": 2e-05, + "loss": 297.9178, + "step": 313 + }, + { + "epoch": 1.05, + "learning_rate": 2e-05, + "loss": 219.1808, + "step": 314 + }, + { + "epoch": 1.05, + "learning_rate": 2e-05, + "loss": 212.635, + "step": 315 + }, + { + "epoch": 1.05, + "learning_rate": 2e-05, + "loss": 216.6892, + "step": 316 + }, + { + "epoch": 1.06, + "learning_rate": 2e-05, + "loss": 336.1581, + "step": 317 + }, + { + "epoch": 1.06, + "learning_rate": 2e-05, + "loss": 178.8046, + "step": 318 + }, + { + "epoch": 1.06, + "learning_rate": 2e-05, + "loss": 249.9513, + "step": 319 + }, + { + "epoch": 1.07, + "learning_rate": 2e-05, + "loss": 258.9301, + "step": 320 + }, + { + "epoch": 1.07, + "learning_rate": 2e-05, + "loss": 308.1837, + "step": 321 + }, + { + "epoch": 1.07, + "learning_rate": 2e-05, + "loss": 214.1725, + "step": 322 + }, + { + "epoch": 1.08, + "learning_rate": 2e-05, + "loss": 218.7718, + "step": 323 + }, + { + "epoch": 1.08, + "learning_rate": 2e-05, + "loss": 343.1404, + "step": 324 + }, + { + "epoch": 1.08, + "learning_rate": 2e-05, + "loss": 200.2678, + "step": 325 + }, + { + "epoch": 1.09, + "learning_rate": 2e-05, + "loss": 321.7372, + "step": 326 + }, + { + "epoch": 1.09, + "learning_rate": 2e-05, + "loss": 264.7751, + "step": 327 + }, + { + "epoch": 1.09, + "learning_rate": 2e-05, + "loss": 224.7439, + "step": 328 + }, + { + "epoch": 1.1, + "learning_rate": 2e-05, + "loss": 196.1887, + "step": 329 + }, + { + "epoch": 1.1, + "learning_rate": 2e-05, + "loss": 234.8271, + "step": 330 + }, + { + "epoch": 1.1, + "learning_rate": 2e-05, + "loss": 220.26, + "step": 331 + }, + { + "epoch": 1.11, + "learning_rate": 2e-05, + "loss": 213.665, + "step": 332 + }, + { + "epoch": 1.11, + "learning_rate": 2e-05, + "loss": 232.8795, + "step": 333 + }, + { + "epoch": 1.11, + "learning_rate": 2e-05, + "loss": 214.4828, + "step": 334 + }, + { + "epoch": 1.12, + "learning_rate": 2e-05, + "loss": 198.4257, + "step": 335 + }, + { + "epoch": 1.12, + "learning_rate": 2e-05, + "loss": 246.1346, + "step": 336 + }, + { + "epoch": 1.12, + "learning_rate": 2e-05, + "loss": 223.7612, + "step": 337 + }, + { + "epoch": 1.13, + "learning_rate": 2e-05, + "loss": 303.2517, + "step": 338 + }, + { + "epoch": 1.13, + "learning_rate": 2e-05, + "loss": 240.5897, + "step": 339 + }, + { + "epoch": 1.13, + "learning_rate": 2e-05, + "loss": 206.087, + "step": 340 + }, + { + "epoch": 1.14, + "learning_rate": 2e-05, + "loss": 238.7818, + "step": 341 + }, + { + "epoch": 1.14, + "learning_rate": 2e-05, + "loss": 225.7912, + "step": 342 + }, + { + "epoch": 1.14, + "learning_rate": 2e-05, + "loss": 229.2215, + "step": 343 + }, + { + "epoch": 1.15, + "learning_rate": 2e-05, + "loss": 180.5617, + "step": 344 + }, + { + "epoch": 1.15, + "learning_rate": 2e-05, + "loss": 254.5206, + "step": 345 + }, + { + "epoch": 1.15, + "learning_rate": 2e-05, + "loss": 236.183, + "step": 346 + }, + { + "epoch": 1.16, + "learning_rate": 2e-05, + "loss": 235.8767, + "step": 347 + }, + { + "epoch": 1.16, + "learning_rate": 2e-05, + "loss": 223.3795, + "step": 348 + }, + { + "epoch": 1.16, + "learning_rate": 2e-05, + "loss": 236.039, + "step": 349 + }, + { + "epoch": 1.17, + "learning_rate": 2e-05, + "loss": 226.3709, + "step": 350 + }, + { + "epoch": 1.17, + "learning_rate": 2e-05, + "loss": 228.4313, + "step": 351 + }, + { + "epoch": 1.17, + "learning_rate": 2e-05, + "loss": 203.7726, + "step": 352 + }, + { + "epoch": 1.18, + "learning_rate": 2e-05, + "loss": 223.8938, + "step": 353 + }, + { + "epoch": 1.18, + "learning_rate": 2e-05, + "loss": 206.5053, + "step": 354 + }, + { + "epoch": 1.18, + "learning_rate": 2e-05, + "loss": 224.3272, + "step": 355 + }, + { + "epoch": 1.19, + "learning_rate": 2e-05, + "loss": 253.1774, + "step": 356 + }, + { + "epoch": 1.19, + "learning_rate": 2e-05, + "loss": 229.9487, + "step": 357 + }, + { + "epoch": 1.19, + "learning_rate": 2e-05, + "loss": 230.8462, + "step": 358 + }, + { + "epoch": 1.2, + "learning_rate": 2e-05, + "loss": 298.9388, + "step": 359 + }, + { + "epoch": 1.2, + "learning_rate": 2e-05, + "loss": 209.2686, + "step": 360 + }, + { + "epoch": 1.2, + "learning_rate": 2e-05, + "loss": 191.1316, + "step": 361 + }, + { + "epoch": 1.21, + "learning_rate": 2e-05, + "loss": 263.6025, + "step": 362 + }, + { + "epoch": 1.21, + "learning_rate": 2e-05, + "loss": 193.0966, + "step": 363 + }, + { + "epoch": 1.21, + "learning_rate": 2e-05, + "loss": 193.0884, + "step": 364 + }, + { + "epoch": 1.22, + "learning_rate": 2e-05, + "loss": 218.7641, + "step": 365 + }, + { + "epoch": 1.22, + "learning_rate": 2e-05, + "loss": 245.633, + "step": 366 + }, + { + "epoch": 1.22, + "learning_rate": 2e-05, + "loss": 177.7586, + "step": 367 + }, + { + "epoch": 1.23, + "learning_rate": 2e-05, + "loss": 189.6293, + "step": 368 + }, + { + "epoch": 1.23, + "learning_rate": 2e-05, + "loss": 188.3251, + "step": 369 + }, + { + "epoch": 1.23, + "learning_rate": 2e-05, + "loss": 211.644, + "step": 370 + }, + { + "epoch": 1.24, + "learning_rate": 2e-05, + "loss": 165.5959, + "step": 371 + }, + { + "epoch": 1.24, + "learning_rate": 2e-05, + "loss": 203.339, + "step": 372 + }, + { + "epoch": 1.24, + "learning_rate": 2e-05, + "loss": 207.5482, + "step": 373 + }, + { + "epoch": 1.25, + "learning_rate": 2e-05, + "loss": 236.129, + "step": 374 + }, + { + "epoch": 1.25, + "learning_rate": 2e-05, + "loss": 251.883, + "step": 375 + }, + { + "epoch": 1.25, + "learning_rate": 2e-05, + "loss": 214.6921, + "step": 376 + }, + { + "epoch": 1.26, + "learning_rate": 2e-05, + "loss": 249.8723, + "step": 377 + }, + { + "epoch": 1.26, + "learning_rate": 2e-05, + "loss": 234.278, + "step": 378 + }, + { + "epoch": 1.26, + "learning_rate": 2e-05, + "loss": 251.6463, + "step": 379 + }, + { + "epoch": 1.27, + "learning_rate": 2e-05, + "loss": 244.5367, + "step": 380 + }, + { + "epoch": 1.27, + "learning_rate": 2e-05, + "loss": 226.0958, + "step": 381 + }, + { + "epoch": 1.27, + "learning_rate": 2e-05, + "loss": 204.7845, + "step": 382 + }, + { + "epoch": 1.28, + "learning_rate": 2e-05, + "loss": 229.6162, + "step": 383 + }, + { + "epoch": 1.28, + "learning_rate": 2e-05, + "loss": 231.22, + "step": 384 + }, + { + "epoch": 1.28, + "learning_rate": 2e-05, + "loss": 207.0068, + "step": 385 + }, + { + "epoch": 1.29, + "learning_rate": 2e-05, + "loss": 245.944, + "step": 386 + }, + { + "epoch": 1.29, + "learning_rate": 2e-05, + "loss": 181.7492, + "step": 387 + }, + { + "epoch": 1.29, + "learning_rate": 2e-05, + "loss": 195.8122, + "step": 388 + }, + { + "epoch": 1.3, + "learning_rate": 2e-05, + "loss": 205.7455, + "step": 389 + }, + { + "epoch": 1.3, + "learning_rate": 2e-05, + "loss": 210.8664, + "step": 390 + }, + { + "epoch": 1.3, + "learning_rate": 2e-05, + "loss": 205.2473, + "step": 391 + }, + { + "epoch": 1.31, + "learning_rate": 2e-05, + "loss": 200.1387, + "step": 392 + }, + { + "epoch": 1.31, + "learning_rate": 2e-05, + "loss": 189.6015, + "step": 393 + }, + { + "epoch": 1.31, + "learning_rate": 2e-05, + "loss": 271.6579, + "step": 394 + }, + { + "epoch": 1.32, + "learning_rate": 2e-05, + "loss": 256.1616, + "step": 395 + }, + { + "epoch": 1.32, + "learning_rate": 2e-05, + "loss": 227.5033, + "step": 396 + }, + { + "epoch": 1.32, + "learning_rate": 2e-05, + "loss": 230.1321, + "step": 397 + }, + { + "epoch": 1.33, + "learning_rate": 2e-05, + "loss": 211.3426, + "step": 398 + }, + { + "epoch": 1.33, + "learning_rate": 2e-05, + "loss": 187.5162, + "step": 399 + }, + { + "epoch": 1.33, + "learning_rate": 2e-05, + "loss": 211.4105, + "step": 400 + }, + { + "epoch": 1.34, + "learning_rate": 2e-05, + "loss": 250.8305, + "step": 401 + }, + { + "epoch": 1.34, + "learning_rate": 2e-05, + "loss": 233.2176, + "step": 402 + }, + { + "epoch": 1.34, + "learning_rate": 2e-05, + "loss": 186.2552, + "step": 403 + }, + { + "epoch": 1.35, + "learning_rate": 2e-05, + "loss": 182.1218, + "step": 404 + }, + { + "epoch": 1.35, + "learning_rate": 2e-05, + "loss": 245.9604, + "step": 405 + }, + { + "epoch": 1.35, + "learning_rate": 2e-05, + "loss": 186.1751, + "step": 406 + }, + { + "epoch": 1.36, + "learning_rate": 2e-05, + "loss": 238.4419, + "step": 407 + }, + { + "epoch": 1.36, + "learning_rate": 2e-05, + "loss": 222.9325, + "step": 408 + }, + { + "epoch": 1.36, + "learning_rate": 2e-05, + "loss": 183.7545, + "step": 409 + }, + { + "epoch": 1.37, + "learning_rate": 2e-05, + "loss": 258.3376, + "step": 410 + }, + { + "epoch": 1.37, + "learning_rate": 2e-05, + "loss": 207.5193, + "step": 411 + }, + { + "epoch": 1.37, + "learning_rate": 2e-05, + "loss": 209.8826, + "step": 412 + }, + { + "epoch": 1.38, + "learning_rate": 2e-05, + "loss": 206.6928, + "step": 413 + }, + { + "epoch": 1.38, + "learning_rate": 2e-05, + "loss": 170.6046, + "step": 414 + }, + { + "epoch": 1.38, + "learning_rate": 2e-05, + "loss": 210.6746, + "step": 415 + }, + { + "epoch": 1.39, + "learning_rate": 2e-05, + "loss": 208.4622, + "step": 416 + }, + { + "epoch": 1.39, + "learning_rate": 2e-05, + "loss": 214.8073, + "step": 417 + }, + { + "epoch": 1.39, + "learning_rate": 2e-05, + "loss": 252.3761, + "step": 418 + }, + { + "epoch": 1.4, + "learning_rate": 2e-05, + "loss": 194.0715, + "step": 419 + }, + { + "epoch": 1.4, + "learning_rate": 2e-05, + "loss": 195.3139, + "step": 420 + }, + { + "epoch": 1.4, + "learning_rate": 2e-05, + "loss": 200.1256, + "step": 421 + }, + { + "epoch": 1.41, + "learning_rate": 2e-05, + "loss": 203.8806, + "step": 422 + }, + { + "epoch": 1.41, + "learning_rate": 2e-05, + "loss": 363.1582, + "step": 423 + }, + { + "epoch": 1.41, + "learning_rate": 2e-05, + "loss": 214.0313, + "step": 424 + }, + { + "epoch": 1.42, + "learning_rate": 2e-05, + "loss": 248.3994, + "step": 425 + }, + { + "epoch": 1.42, + "learning_rate": 2e-05, + "loss": 203.2281, + "step": 426 + }, + { + "epoch": 1.42, + "learning_rate": 2e-05, + "loss": 234.2976, + "step": 427 + }, + { + "epoch": 1.43, + "learning_rate": 2e-05, + "loss": 237.7386, + "step": 428 + }, + { + "epoch": 1.43, + "learning_rate": 2e-05, + "loss": 254.7703, + "step": 429 + }, + { + "epoch": 1.43, + "learning_rate": 2e-05, + "loss": 226.2335, + "step": 430 + }, + { + "epoch": 1.44, + "learning_rate": 2e-05, + "loss": 195.5936, + "step": 431 + }, + { + "epoch": 1.44, + "learning_rate": 2e-05, + "loss": 202.621, + "step": 432 + }, + { + "epoch": 1.44, + "learning_rate": 2e-05, + "loss": 224.2951, + "step": 433 + }, + { + "epoch": 1.45, + "learning_rate": 2e-05, + "loss": 211.6174, + "step": 434 + }, + { + "epoch": 1.45, + "learning_rate": 2e-05, + "loss": 215.9442, + "step": 435 + }, + { + "epoch": 1.45, + "learning_rate": 2e-05, + "loss": 191.8515, + "step": 436 + }, + { + "epoch": 1.46, + "learning_rate": 2e-05, + "loss": 202.5363, + "step": 437 + }, + { + "epoch": 1.46, + "learning_rate": 2e-05, + "loss": 244.0159, + "step": 438 + }, + { + "epoch": 1.46, + "learning_rate": 2e-05, + "loss": 254.0356, + "step": 439 + }, + { + "epoch": 1.47, + "learning_rate": 2e-05, + "loss": 267.0516, + "step": 440 + }, + { + "epoch": 1.47, + "learning_rate": 2e-05, + "loss": 199.8777, + "step": 441 + }, + { + "epoch": 1.47, + "learning_rate": 2e-05, + "loss": 208.3456, + "step": 442 + }, + { + "epoch": 1.48, + "learning_rate": 2e-05, + "loss": 238.818, + "step": 443 + }, + { + "epoch": 1.48, + "learning_rate": 2e-05, + "loss": 226.0656, + "step": 444 + }, + { + "epoch": 1.48, + "learning_rate": 2e-05, + "loss": 244.9799, + "step": 445 + }, + { + "epoch": 1.49, + "learning_rate": 2e-05, + "loss": 203.9123, + "step": 446 + }, + { + "epoch": 1.49, + "learning_rate": 2e-05, + "loss": 207.2839, + "step": 447 + }, + { + "epoch": 1.49, + "learning_rate": 2e-05, + "loss": 222.766, + "step": 448 + }, + { + "epoch": 1.5, + "learning_rate": 2e-05, + "loss": 252.938, + "step": 449 + }, + { + "epoch": 1.5, + "learning_rate": 2e-05, + "loss": 230.257, + "step": 450 + }, + { + "epoch": 1.5, + "eval_loss": 226.50051879882812, + "eval_runtime": 287.559, + "eval_samples_per_second": 16.692, + "eval_steps_per_second": 1.043, + "step": 450 + }, + { + "epoch": 1.5, + "learning_rate": 2e-05, + "loss": 249.9676, + "step": 451 + }, + { + "epoch": 1.51, + "learning_rate": 2e-05, + "loss": 291.7989, + "step": 452 + }, + { + "epoch": 1.51, + "learning_rate": 2e-05, + "loss": 244.1971, + "step": 453 + }, + { + "epoch": 1.51, + "learning_rate": 2e-05, + "loss": 210.1197, + "step": 454 + }, + { + "epoch": 1.52, + "learning_rate": 2e-05, + "loss": 185.1812, + "step": 455 + }, + { + "epoch": 1.52, + "learning_rate": 2e-05, + "loss": 203.96, + "step": 456 + }, + { + "epoch": 1.52, + "learning_rate": 2e-05, + "loss": 248.2275, + "step": 457 + }, + { + "epoch": 1.53, + "learning_rate": 2e-05, + "loss": 211.4718, + "step": 458 + }, + { + "epoch": 1.53, + "learning_rate": 2e-05, + "loss": 194.2697, + "step": 459 + }, + { + "epoch": 1.53, + "learning_rate": 2e-05, + "loss": 195.9454, + "step": 460 + }, + { + "epoch": 1.54, + "learning_rate": 2e-05, + "loss": 217.1035, + "step": 461 + }, + { + "epoch": 1.54, + "learning_rate": 2e-05, + "loss": 224.5334, + "step": 462 + }, + { + "epoch": 1.54, + "learning_rate": 2e-05, + "loss": 199.3744, + "step": 463 + }, + { + "epoch": 1.55, + "learning_rate": 2e-05, + "loss": 204.6962, + "step": 464 + }, + { + "epoch": 1.55, + "learning_rate": 2e-05, + "loss": 171.7811, + "step": 465 + }, + { + "epoch": 1.55, + "learning_rate": 2e-05, + "loss": 244.8348, + "step": 466 + }, + { + "epoch": 1.56, + "learning_rate": 2e-05, + "loss": 187.4453, + "step": 467 + }, + { + "epoch": 1.56, + "learning_rate": 2e-05, + "loss": 225.2651, + "step": 468 + }, + { + "epoch": 1.56, + "learning_rate": 2e-05, + "loss": 202.3235, + "step": 469 + }, + { + "epoch": 1.57, + "learning_rate": 2e-05, + "loss": 267.9012, + "step": 470 + }, + { + "epoch": 1.57, + "learning_rate": 2e-05, + "loss": 192.601, + "step": 471 + }, + { + "epoch": 1.57, + "learning_rate": 2e-05, + "loss": 215.3246, + "step": 472 + }, + { + "epoch": 1.58, + "learning_rate": 2e-05, + "loss": 206.6324, + "step": 473 + }, + { + "epoch": 1.58, + "learning_rate": 2e-05, + "loss": 224.5978, + "step": 474 + }, + { + "epoch": 1.58, + "learning_rate": 2e-05, + "loss": 194.7379, + "step": 475 + }, + { + "epoch": 1.59, + "learning_rate": 2e-05, + "loss": 181.7932, + "step": 476 + }, + { + "epoch": 1.59, + "learning_rate": 2e-05, + "loss": 207.0792, + "step": 477 + }, + { + "epoch": 1.59, + "learning_rate": 2e-05, + "loss": 209.3582, + "step": 478 + }, + { + "epoch": 1.6, + "learning_rate": 2e-05, + "loss": 155.4181, + "step": 479 + }, + { + "epoch": 1.6, + "learning_rate": 2e-05, + "loss": 208.6905, + "step": 480 + }, + { + "epoch": 1.6, + "learning_rate": 2e-05, + "loss": 179.2064, + "step": 481 + }, + { + "epoch": 1.61, + "learning_rate": 2e-05, + "loss": 244.269, + "step": 482 + }, + { + "epoch": 1.61, + "learning_rate": 2e-05, + "loss": 203.8597, + "step": 483 + }, + { + "epoch": 1.61, + "learning_rate": 2e-05, + "loss": 215.3183, + "step": 484 + }, + { + "epoch": 1.62, + "learning_rate": 2e-05, + "loss": 209.9353, + "step": 485 + }, + { + "epoch": 1.62, + "learning_rate": 2e-05, + "loss": 208.1019, + "step": 486 + }, + { + "epoch": 1.62, + "learning_rate": 2e-05, + "loss": 185.1986, + "step": 487 + }, + { + "epoch": 1.63, + "learning_rate": 2e-05, + "loss": 221.299, + "step": 488 + }, + { + "epoch": 1.63, + "learning_rate": 2e-05, + "loss": 197.2944, + "step": 489 + }, + { + "epoch": 1.63, + "learning_rate": 2e-05, + "loss": 249.1776, + "step": 490 + }, + { + "epoch": 1.64, + "learning_rate": 2e-05, + "loss": 249.6916, + "step": 491 + }, + { + "epoch": 1.64, + "learning_rate": 2e-05, + "loss": 223.0698, + "step": 492 + }, + { + "epoch": 1.64, + "learning_rate": 2e-05, + "loss": 212.4214, + "step": 493 + }, + { + "epoch": 1.65, + "learning_rate": 2e-05, + "loss": 207.3959, + "step": 494 + }, + { + "epoch": 1.65, + "learning_rate": 2e-05, + "loss": 197.5657, + "step": 495 + }, + { + "epoch": 1.65, + "learning_rate": 2e-05, + "loss": 224.0082, + "step": 496 + }, + { + "epoch": 1.66, + "learning_rate": 2e-05, + "loss": 194.0723, + "step": 497 + }, + { + "epoch": 1.66, + "learning_rate": 2e-05, + "loss": 241.0521, + "step": 498 + }, + { + "epoch": 1.66, + "learning_rate": 2e-05, + "loss": 195.3386, + "step": 499 + }, + { + "epoch": 1.67, + "learning_rate": 2e-05, + "loss": 246.2868, + "step": 500 + }, + { + "epoch": 1.67, + "learning_rate": 2e-05, + "loss": 185.8767, + "step": 501 + }, + { + "epoch": 1.67, + "learning_rate": 2e-05, + "loss": 199.1286, + "step": 502 + }, + { + "epoch": 1.68, + "learning_rate": 2e-05, + "loss": 234.8778, + "step": 503 + }, + { + "epoch": 1.68, + "learning_rate": 2e-05, + "loss": 246.7677, + "step": 504 + }, + { + "epoch": 1.68, + "learning_rate": 2e-05, + "loss": 215.8775, + "step": 505 + }, + { + "epoch": 1.69, + "learning_rate": 2e-05, + "loss": 208.5989, + "step": 506 + }, + { + "epoch": 1.69, + "learning_rate": 2e-05, + "loss": 193.0592, + "step": 507 + }, + { + "epoch": 1.69, + "learning_rate": 2e-05, + "loss": 254.8217, + "step": 508 + }, + { + "epoch": 1.7, + "learning_rate": 2e-05, + "loss": 222.7972, + "step": 509 + }, + { + "epoch": 1.7, + "learning_rate": 2e-05, + "loss": 229.1115, + "step": 510 + }, + { + "epoch": 1.7, + "learning_rate": 2e-05, + "loss": 201.3895, + "step": 511 + }, + { + "epoch": 1.71, + "learning_rate": 2e-05, + "loss": 170.4128, + "step": 512 + }, + { + "epoch": 1.71, + "learning_rate": 2e-05, + "loss": 227.652, + "step": 513 + }, + { + "epoch": 1.71, + "learning_rate": 2e-05, + "loss": 203.8975, + "step": 514 + }, + { + "epoch": 1.72, + "learning_rate": 2e-05, + "loss": 209.6851, + "step": 515 + }, + { + "epoch": 1.72, + "learning_rate": 2e-05, + "loss": 188.932, + "step": 516 + }, + { + "epoch": 1.72, + "learning_rate": 2e-05, + "loss": 187.8048, + "step": 517 + }, + { + "epoch": 1.73, + "learning_rate": 2e-05, + "loss": 265.4262, + "step": 518 + }, + { + "epoch": 1.73, + "learning_rate": 2e-05, + "loss": 250.1163, + "step": 519 + }, + { + "epoch": 1.73, + "learning_rate": 2e-05, + "loss": 190.8139, + "step": 520 + }, + { + "epoch": 1.74, + "learning_rate": 2e-05, + "loss": 203.3908, + "step": 521 + }, + { + "epoch": 1.74, + "learning_rate": 2e-05, + "loss": 250.0257, + "step": 522 + }, + { + "epoch": 1.74, + "learning_rate": 2e-05, + "loss": 167.6775, + "step": 523 + }, + { + "epoch": 1.75, + "learning_rate": 2e-05, + "loss": 216.208, + "step": 524 + }, + { + "epoch": 1.75, + "learning_rate": 2e-05, + "loss": 248.1995, + "step": 525 + }, + { + "epoch": 1.75, + "learning_rate": 2e-05, + "loss": 186.5524, + "step": 526 + }, + { + "epoch": 1.76, + "learning_rate": 2e-05, + "loss": 198.6543, + "step": 527 + }, + { + "epoch": 1.76, + "learning_rate": 2e-05, + "loss": 221.3749, + "step": 528 + }, + { + "epoch": 1.76, + "learning_rate": 2e-05, + "loss": 239.5123, + "step": 529 + }, + { + "epoch": 1.77, + "learning_rate": 2e-05, + "loss": 202.19, + "step": 530 + }, + { + "epoch": 1.77, + "learning_rate": 2e-05, + "loss": 187.0039, + "step": 531 + }, + { + "epoch": 1.77, + "learning_rate": 2e-05, + "loss": 281.6668, + "step": 532 + }, + { + "epoch": 1.78, + "learning_rate": 2e-05, + "loss": 225.5882, + "step": 533 + }, + { + "epoch": 1.78, + "learning_rate": 2e-05, + "loss": 154.8565, + "step": 534 + }, + { + "epoch": 1.78, + "learning_rate": 2e-05, + "loss": 206.029, + "step": 535 + }, + { + "epoch": 1.79, + "learning_rate": 2e-05, + "loss": 219.6192, + "step": 536 + }, + { + "epoch": 1.79, + "learning_rate": 2e-05, + "loss": 264.9105, + "step": 537 + }, + { + "epoch": 1.79, + "learning_rate": 2e-05, + "loss": 209.3203, + "step": 538 + }, + { + "epoch": 1.8, + "learning_rate": 2e-05, + "loss": 221.2177, + "step": 539 + }, + { + "epoch": 1.8, + "learning_rate": 2e-05, + "loss": 206.0646, + "step": 540 + }, + { + "epoch": 1.8, + "learning_rate": 2e-05, + "loss": 179.302, + "step": 541 + }, + { + "epoch": 1.81, + "learning_rate": 2e-05, + "loss": 216.4368, + "step": 542 + }, + { + "epoch": 1.81, + "learning_rate": 2e-05, + "loss": 256.2146, + "step": 543 + }, + { + "epoch": 1.81, + "learning_rate": 2e-05, + "loss": 225.4241, + "step": 544 + }, + { + "epoch": 1.82, + "learning_rate": 2e-05, + "loss": 223.213, + "step": 545 + }, + { + "epoch": 1.82, + "learning_rate": 2e-05, + "loss": 224.182, + "step": 546 + }, + { + "epoch": 1.82, + "learning_rate": 2e-05, + "loss": 226.146, + "step": 547 + }, + { + "epoch": 1.83, + "learning_rate": 2e-05, + "loss": 201.2667, + "step": 548 + }, + { + "epoch": 1.83, + "learning_rate": 2e-05, + "loss": 196.2431, + "step": 549 + }, + { + "epoch": 1.83, + "learning_rate": 2e-05, + "loss": 253.1102, + "step": 550 + }, + { + "epoch": 1.84, + "learning_rate": 2e-05, + "loss": 218.3497, + "step": 551 + }, + { + "epoch": 1.84, + "learning_rate": 2e-05, + "loss": 239.5849, + "step": 552 + }, + { + "epoch": 1.84, + "learning_rate": 2e-05, + "loss": 229.3954, + "step": 553 + }, + { + "epoch": 1.85, + "learning_rate": 2e-05, + "loss": 224.5805, + "step": 554 + }, + { + "epoch": 1.85, + "learning_rate": 2e-05, + "loss": 193.1906, + "step": 555 + }, + { + "epoch": 1.85, + "learning_rate": 2e-05, + "loss": 198.8174, + "step": 556 + }, + { + "epoch": 1.86, + "learning_rate": 2e-05, + "loss": 206.2353, + "step": 557 + }, + { + "epoch": 1.86, + "learning_rate": 2e-05, + "loss": 226.555, + "step": 558 + }, + { + "epoch": 1.86, + "learning_rate": 2e-05, + "loss": 259.067, + "step": 559 + }, + { + "epoch": 1.87, + "learning_rate": 2e-05, + "loss": 204.7034, + "step": 560 + }, + { + "epoch": 1.87, + "learning_rate": 2e-05, + "loss": 237.9439, + "step": 561 + }, + { + "epoch": 1.87, + "learning_rate": 2e-05, + "loss": 197.9103, + "step": 562 + }, + { + "epoch": 1.88, + "learning_rate": 2e-05, + "loss": 217.9633, + "step": 563 + }, + { + "epoch": 1.88, + "learning_rate": 2e-05, + "loss": 212.1185, + "step": 564 + }, + { + "epoch": 1.88, + "learning_rate": 2e-05, + "loss": 173.727, + "step": 565 + }, + { + "epoch": 1.89, + "learning_rate": 2e-05, + "loss": 212.4548, + "step": 566 + }, + { + "epoch": 1.89, + "learning_rate": 2e-05, + "loss": 269.4512, + "step": 567 + }, + { + "epoch": 1.89, + "learning_rate": 2e-05, + "loss": 219.3664, + "step": 568 + }, + { + "epoch": 1.9, + "learning_rate": 2e-05, + "loss": 200.041, + "step": 569 + }, + { + "epoch": 1.9, + "learning_rate": 2e-05, + "loss": 196.7029, + "step": 570 + }, + { + "epoch": 1.9, + "learning_rate": 2e-05, + "loss": 225.4566, + "step": 571 + }, + { + "epoch": 1.91, + "learning_rate": 2e-05, + "loss": 212.7281, + "step": 572 + }, + { + "epoch": 1.91, + "learning_rate": 2e-05, + "loss": 187.6301, + "step": 573 + }, + { + "epoch": 1.91, + "learning_rate": 2e-05, + "loss": 335.6991, + "step": 574 + }, + { + "epoch": 1.92, + "learning_rate": 2e-05, + "loss": 153.5664, + "step": 575 + }, + { + "epoch": 1.92, + "learning_rate": 2e-05, + "loss": 217.0094, + "step": 576 + }, + { + "epoch": 1.92, + "learning_rate": 2e-05, + "loss": 255.8123, + "step": 577 + }, + { + "epoch": 1.93, + "learning_rate": 2e-05, + "loss": 206.1462, + "step": 578 + }, + { + "epoch": 1.93, + "learning_rate": 2e-05, + "loss": 242.3889, + "step": 579 + }, + { + "epoch": 1.93, + "learning_rate": 2e-05, + "loss": 218.3518, + "step": 580 + }, + { + "epoch": 1.94, + "learning_rate": 2e-05, + "loss": 200.9317, + "step": 581 + }, + { + "epoch": 1.94, + "learning_rate": 2e-05, + "loss": 212.1096, + "step": 582 + }, + { + "epoch": 1.94, + "learning_rate": 2e-05, + "loss": 213.4619, + "step": 583 + }, + { + "epoch": 1.95, + "learning_rate": 2e-05, + "loss": 232.34, + "step": 584 + }, + { + "epoch": 1.95, + "learning_rate": 2e-05, + "loss": 159.0329, + "step": 585 + }, + { + "epoch": 1.95, + "learning_rate": 2e-05, + "loss": 221.9789, + "step": 586 + }, + { + "epoch": 1.96, + "learning_rate": 2e-05, + "loss": 224.5071, + "step": 587 + }, + { + "epoch": 1.96, + "learning_rate": 2e-05, + "loss": 251.3988, + "step": 588 + }, + { + "epoch": 1.96, + "learning_rate": 2e-05, + "loss": 228.7717, + "step": 589 + }, + { + "epoch": 1.97, + "learning_rate": 2e-05, + "loss": 187.2274, + "step": 590 + }, + { + "epoch": 1.97, + "learning_rate": 2e-05, + "loss": 213.5903, + "step": 591 + }, + { + "epoch": 1.97, + "learning_rate": 2e-05, + "loss": 196.8019, + "step": 592 + }, + { + "epoch": 1.98, + "learning_rate": 2e-05, + "loss": 225.6396, + "step": 593 + }, + { + "epoch": 1.98, + "learning_rate": 2e-05, + "loss": 250.5785, + "step": 594 + }, + { + "epoch": 1.98, + "learning_rate": 2e-05, + "loss": 180.8865, + "step": 595 + }, + { + "epoch": 1.99, + "learning_rate": 2e-05, + "loss": 190.5121, + "step": 596 + }, + { + "epoch": 1.99, + "learning_rate": 2e-05, + "loss": 166.3393, + "step": 597 + }, + { + "epoch": 1.99, + "learning_rate": 2e-05, + "loss": 187.183, + "step": 598 + }, + { + "epoch": 2.0, + "learning_rate": 2e-05, + "loss": 199.8197, + "step": 599 + }, + { + "epoch": 2.0, + "learning_rate": 2e-05, + "loss": 231.2036, + "step": 600 + }, + { + "epoch": 2.0, + "eval_loss": 217.99810791015625, + "eval_runtime": 288.0687, + "eval_samples_per_second": 16.663, + "eval_steps_per_second": 1.041, + "step": 600 + }, + { + "epoch": 2.0, + "learning_rate": 8e-06, + "loss": 175.3337, + "step": 601 + }, + { + "epoch": 2.01, + "learning_rate": 8e-06, + "loss": 182.202, + "step": 602 + }, + { + "epoch": 2.01, + "learning_rate": 8e-06, + "loss": 177.5331, + "step": 603 + }, + { + "epoch": 2.01, + "learning_rate": 8e-06, + "loss": 201.5817, + "step": 604 + }, + { + "epoch": 2.02, + "learning_rate": 8e-06, + "loss": 202.2545, + "step": 605 + }, + { + "epoch": 2.02, + "learning_rate": 8e-06, + "loss": 228.6945, + "step": 606 + }, + { + "epoch": 2.02, + "learning_rate": 8e-06, + "loss": 198.3622, + "step": 607 + }, + { + "epoch": 2.03, + "learning_rate": 8e-06, + "loss": 165.2558, + "step": 608 + }, + { + "epoch": 2.03, + "learning_rate": 8e-06, + "loss": 200.0535, + "step": 609 + }, + { + "epoch": 2.03, + "learning_rate": 8e-06, + "loss": 202.44, + "step": 610 + }, + { + "epoch": 2.04, + "learning_rate": 8e-06, + "loss": 174.4217, + "step": 611 + }, + { + "epoch": 2.04, + "learning_rate": 8e-06, + "loss": 162.0943, + "step": 612 + }, + { + "epoch": 2.04, + "learning_rate": 8e-06, + "loss": 189.5728, + "step": 613 + }, + { + "epoch": 2.05, + "learning_rate": 8e-06, + "loss": 145.0881, + "step": 614 + }, + { + "epoch": 2.05, + "learning_rate": 8e-06, + "loss": 202.7971, + "step": 615 + }, + { + "epoch": 2.05, + "learning_rate": 8e-06, + "loss": 217.9154, + "step": 616 + }, + { + "epoch": 2.06, + "learning_rate": 8e-06, + "loss": 170.6097, + "step": 617 + }, + { + "epoch": 2.06, + "learning_rate": 8e-06, + "loss": 207.4393, + "step": 618 + }, + { + "epoch": 2.06, + "learning_rate": 8e-06, + "loss": 195.7061, + "step": 619 + }, + { + "epoch": 2.07, + "learning_rate": 8e-06, + "loss": 197.8928, + "step": 620 + }, + { + "epoch": 2.07, + "learning_rate": 8e-06, + "loss": 187.5499, + "step": 621 + }, + { + "epoch": 2.07, + "learning_rate": 8e-06, + "loss": 176.1068, + "step": 622 + }, + { + "epoch": 2.08, + "learning_rate": 8e-06, + "loss": 173.0856, + "step": 623 + }, + { + "epoch": 2.08, + "learning_rate": 8e-06, + "loss": 160.888, + "step": 624 + }, + { + "epoch": 2.08, + "learning_rate": 8e-06, + "loss": 148.3637, + "step": 625 + }, + { + "epoch": 2.09, + "learning_rate": 8e-06, + "loss": 167.0729, + "step": 626 + }, + { + "epoch": 2.09, + "learning_rate": 8e-06, + "loss": 189.1506, + "step": 627 + }, + { + "epoch": 2.09, + "learning_rate": 8e-06, + "loss": 155.5463, + "step": 628 + }, + { + "epoch": 2.1, + "learning_rate": 8e-06, + "loss": 145.7486, + "step": 629 + }, + { + "epoch": 2.1, + "learning_rate": 8e-06, + "loss": 174.9353, + "step": 630 + }, + { + "epoch": 2.1, + "learning_rate": 8e-06, + "loss": 194.5591, + "step": 631 + }, + { + "epoch": 2.11, + "learning_rate": 8e-06, + "loss": 147.3489, + "step": 632 + }, + { + "epoch": 2.11, + "learning_rate": 8e-06, + "loss": 163.189, + "step": 633 + }, + { + "epoch": 2.11, + "learning_rate": 8e-06, + "loss": 151.512, + "step": 634 + }, + { + "epoch": 2.12, + "learning_rate": 8e-06, + "loss": 176.8741, + "step": 635 + }, + { + "epoch": 2.12, + "learning_rate": 8e-06, + "loss": 142.8391, + "step": 636 + }, + { + "epoch": 2.12, + "learning_rate": 8e-06, + "loss": 177.6373, + "step": 637 + }, + { + "epoch": 2.13, + "learning_rate": 8e-06, + "loss": 185.2133, + "step": 638 + }, + { + "epoch": 2.13, + "learning_rate": 8e-06, + "loss": 182.5912, + "step": 639 + }, + { + "epoch": 2.13, + "learning_rate": 8e-06, + "loss": 185.6598, + "step": 640 + }, + { + "epoch": 2.14, + "learning_rate": 8e-06, + "loss": 155.6521, + "step": 641 + }, + { + "epoch": 2.14, + "learning_rate": 8e-06, + "loss": 183.4825, + "step": 642 + }, + { + "epoch": 2.14, + "learning_rate": 8e-06, + "loss": 192.4618, + "step": 643 + }, + { + "epoch": 2.15, + "learning_rate": 8e-06, + "loss": 144.5469, + "step": 644 + }, + { + "epoch": 2.15, + "learning_rate": 8e-06, + "loss": 174.8644, + "step": 645 + }, + { + "epoch": 2.15, + "learning_rate": 8e-06, + "loss": 216.1262, + "step": 646 + }, + { + "epoch": 2.16, + "learning_rate": 8e-06, + "loss": 186.6161, + "step": 647 + }, + { + "epoch": 2.16, + "learning_rate": 8e-06, + "loss": 194.0572, + "step": 648 + }, + { + "epoch": 2.16, + "learning_rate": 8e-06, + "loss": 163.9438, + "step": 649 + }, + { + "epoch": 2.17, + "learning_rate": 8e-06, + "loss": 164.6438, + "step": 650 + }, + { + "epoch": 2.17, + "learning_rate": 8e-06, + "loss": 164.0169, + "step": 651 + }, + { + "epoch": 2.17, + "learning_rate": 8e-06, + "loss": 203.8269, + "step": 652 + }, + { + "epoch": 2.18, + "learning_rate": 8e-06, + "loss": 246.5373, + "step": 653 + }, + { + "epoch": 2.18, + "learning_rate": 8e-06, + "loss": 197.9024, + "step": 654 + }, + { + "epoch": 2.18, + "learning_rate": 8e-06, + "loss": 221.8409, + "step": 655 + }, + { + "epoch": 2.19, + "learning_rate": 8e-06, + "loss": 203.5629, + "step": 656 + }, + { + "epoch": 2.19, + "learning_rate": 8e-06, + "loss": 202.3013, + "step": 657 + }, + { + "epoch": 2.19, + "learning_rate": 8e-06, + "loss": 155.4175, + "step": 658 + }, + { + "epoch": 2.2, + "learning_rate": 8e-06, + "loss": 156.6437, + "step": 659 + }, + { + "epoch": 2.2, + "learning_rate": 8e-06, + "loss": 141.6618, + "step": 660 + }, + { + "epoch": 2.2, + "learning_rate": 8e-06, + "loss": 182.4935, + "step": 661 + }, + { + "epoch": 2.21, + "learning_rate": 8e-06, + "loss": 182.7516, + "step": 662 + }, + { + "epoch": 2.21, + "learning_rate": 8e-06, + "loss": 186.1476, + "step": 663 + }, + { + "epoch": 2.21, + "learning_rate": 8e-06, + "loss": 150.8565, + "step": 664 + }, + { + "epoch": 2.22, + "learning_rate": 8e-06, + "loss": 163.2544, + "step": 665 + }, + { + "epoch": 2.22, + "learning_rate": 8e-06, + "loss": 158.1472, + "step": 666 + }, + { + "epoch": 2.22, + "learning_rate": 8e-06, + "loss": 323.656, + "step": 667 + }, + { + "epoch": 2.23, + "learning_rate": 8e-06, + "loss": 164.6997, + "step": 668 + }, + { + "epoch": 2.23, + "learning_rate": 8e-06, + "loss": 164.7367, + "step": 669 + }, + { + "epoch": 2.23, + "learning_rate": 8e-06, + "loss": 148.7803, + "step": 670 + }, + { + "epoch": 2.24, + "learning_rate": 8e-06, + "loss": 150.3279, + "step": 671 + }, + { + "epoch": 2.24, + "learning_rate": 8e-06, + "loss": 195.6078, + "step": 672 + }, + { + "epoch": 2.24, + "learning_rate": 8e-06, + "loss": 200.5983, + "step": 673 + }, + { + "epoch": 2.25, + "learning_rate": 8e-06, + "loss": 242.7517, + "step": 674 + }, + { + "epoch": 2.25, + "learning_rate": 8e-06, + "loss": 151.9127, + "step": 675 + }, + { + "epoch": 2.25, + "learning_rate": 8e-06, + "loss": 186.1397, + "step": 676 + }, + { + "epoch": 2.26, + "learning_rate": 8e-06, + "loss": 213.1484, + "step": 677 + }, + { + "epoch": 2.26, + "learning_rate": 8e-06, + "loss": 198.6462, + "step": 678 + }, + { + "epoch": 2.26, + "learning_rate": 8e-06, + "loss": 216.2724, + "step": 679 + }, + { + "epoch": 2.27, + "learning_rate": 8e-06, + "loss": 185.5919, + "step": 680 + }, + { + "epoch": 2.27, + "learning_rate": 8e-06, + "loss": 175.0293, + "step": 681 + }, + { + "epoch": 2.27, + "learning_rate": 8e-06, + "loss": 174.3937, + "step": 682 + }, + { + "epoch": 2.28, + "learning_rate": 8e-06, + "loss": 169.0712, + "step": 683 + }, + { + "epoch": 2.28, + "learning_rate": 8e-06, + "loss": 188.1937, + "step": 684 + }, + { + "epoch": 2.28, + "learning_rate": 8e-06, + "loss": 153.7814, + "step": 685 + }, + { + "epoch": 2.29, + "learning_rate": 8e-06, + "loss": 140.7339, + "step": 686 + }, + { + "epoch": 2.29, + "learning_rate": 8e-06, + "loss": 153.4352, + "step": 687 + }, + { + "epoch": 2.29, + "learning_rate": 8e-06, + "loss": 183.7314, + "step": 688 + }, + { + "epoch": 2.3, + "learning_rate": 8e-06, + "loss": 175.8271, + "step": 689 + }, + { + "epoch": 2.3, + "learning_rate": 8e-06, + "loss": 183.1224, + "step": 690 + }, + { + "epoch": 2.3, + "learning_rate": 8e-06, + "loss": 210.2152, + "step": 691 + }, + { + "epoch": 2.31, + "learning_rate": 8e-06, + "loss": 154.5453, + "step": 692 + }, + { + "epoch": 2.31, + "learning_rate": 8e-06, + "loss": 155.9603, + "step": 693 + }, + { + "epoch": 2.31, + "learning_rate": 8e-06, + "loss": 182.5823, + "step": 694 + }, + { + "epoch": 2.32, + "learning_rate": 8e-06, + "loss": 187.776, + "step": 695 + }, + { + "epoch": 2.32, + "learning_rate": 8e-06, + "loss": 191.5418, + "step": 696 + }, + { + "epoch": 2.32, + "learning_rate": 8e-06, + "loss": 183.6264, + "step": 697 + }, + { + "epoch": 2.33, + "learning_rate": 8e-06, + "loss": 184.0325, + "step": 698 + }, + { + "epoch": 2.33, + "learning_rate": 8e-06, + "loss": 183.5244, + "step": 699 + }, + { + "epoch": 2.33, + "learning_rate": 8e-06, + "loss": 217.7524, + "step": 700 + }, + { + "epoch": 2.34, + "learning_rate": 8e-06, + "loss": 175.4666, + "step": 701 + }, + { + "epoch": 2.34, + "learning_rate": 8e-06, + "loss": 191.1435, + "step": 702 + }, + { + "epoch": 2.34, + "learning_rate": 8e-06, + "loss": 183.9761, + "step": 703 + }, + { + "epoch": 2.35, + "learning_rate": 8e-06, + "loss": 203.3655, + "step": 704 + }, + { + "epoch": 2.35, + "learning_rate": 8e-06, + "loss": 209.4119, + "step": 705 + }, + { + "epoch": 2.35, + "learning_rate": 8e-06, + "loss": 184.6994, + "step": 706 + }, + { + "epoch": 2.36, + "learning_rate": 8e-06, + "loss": 178.2667, + "step": 707 + }, + { + "epoch": 2.36, + "learning_rate": 8e-06, + "loss": 140.5428, + "step": 708 + }, + { + "epoch": 2.36, + "learning_rate": 8e-06, + "loss": 202.2665, + "step": 709 + }, + { + "epoch": 2.37, + "learning_rate": 8e-06, + "loss": 207.6225, + "step": 710 + }, + { + "epoch": 2.37, + "learning_rate": 8e-06, + "loss": 206.2856, + "step": 711 + }, + { + "epoch": 2.37, + "learning_rate": 8e-06, + "loss": 169.7143, + "step": 712 + }, + { + "epoch": 2.38, + "learning_rate": 8e-06, + "loss": 172.6555, + "step": 713 + }, + { + "epoch": 2.38, + "learning_rate": 8e-06, + "loss": 169.8565, + "step": 714 + }, + { + "epoch": 2.38, + "learning_rate": 8e-06, + "loss": 156.0905, + "step": 715 + }, + { + "epoch": 2.39, + "learning_rate": 8e-06, + "loss": 170.2141, + "step": 716 + }, + { + "epoch": 2.39, + "learning_rate": 8e-06, + "loss": 130.7449, + "step": 717 + }, + { + "epoch": 2.39, + "learning_rate": 8e-06, + "loss": 169.6455, + "step": 718 + }, + { + "epoch": 2.4, + "learning_rate": 8e-06, + "loss": 159.3731, + "step": 719 + }, + { + "epoch": 2.4, + "learning_rate": 8e-06, + "loss": 165.2614, + "step": 720 + }, + { + "epoch": 2.4, + "learning_rate": 8e-06, + "loss": 177.168, + "step": 721 + }, + { + "epoch": 2.41, + "learning_rate": 8e-06, + "loss": 167.8181, + "step": 722 + }, + { + "epoch": 2.41, + "learning_rate": 8e-06, + "loss": 180.6999, + "step": 723 + }, + { + "epoch": 2.41, + "learning_rate": 8e-06, + "loss": 215.0352, + "step": 724 + }, + { + "epoch": 2.42, + "learning_rate": 8e-06, + "loss": 156.2193, + "step": 725 + }, + { + "epoch": 2.42, + "learning_rate": 8e-06, + "loss": 149.2104, + "step": 726 + }, + { + "epoch": 2.42, + "learning_rate": 8e-06, + "loss": 150.7269, + "step": 727 + }, + { + "epoch": 2.43, + "learning_rate": 8e-06, + "loss": 173.6506, + "step": 728 + }, + { + "epoch": 2.43, + "learning_rate": 8e-06, + "loss": 162.7161, + "step": 729 + }, + { + "epoch": 2.43, + "learning_rate": 8e-06, + "loss": 169.6832, + "step": 730 + }, + { + "epoch": 2.44, + "learning_rate": 8e-06, + "loss": 150.0918, + "step": 731 + }, + { + "epoch": 2.44, + "learning_rate": 8e-06, + "loss": 160.509, + "step": 732 + }, + { + "epoch": 2.44, + "learning_rate": 8e-06, + "loss": 194.7803, + "step": 733 + }, + { + "epoch": 2.45, + "learning_rate": 8e-06, + "loss": 197.1913, + "step": 734 + }, + { + "epoch": 2.45, + "learning_rate": 8e-06, + "loss": 141.4682, + "step": 735 + }, + { + "epoch": 2.45, + "learning_rate": 8e-06, + "loss": 221.3066, + "step": 736 + }, + { + "epoch": 2.46, + "learning_rate": 8e-06, + "loss": 153.1761, + "step": 737 + }, + { + "epoch": 2.46, + "learning_rate": 8e-06, + "loss": 186.1345, + "step": 738 + }, + { + "epoch": 2.46, + "learning_rate": 8e-06, + "loss": 176.7299, + "step": 739 + }, + { + "epoch": 2.47, + "learning_rate": 8e-06, + "loss": 178.7766, + "step": 740 + }, + { + "epoch": 2.47, + "learning_rate": 8e-06, + "loss": 157.2248, + "step": 741 + }, + { + "epoch": 2.47, + "learning_rate": 8e-06, + "loss": 166.3764, + "step": 742 + }, + { + "epoch": 2.48, + "learning_rate": 8e-06, + "loss": 157.3144, + "step": 743 + }, + { + "epoch": 2.48, + "learning_rate": 8e-06, + "loss": 210.9758, + "step": 744 + }, + { + "epoch": 2.48, + "learning_rate": 8e-06, + "loss": 175.2808, + "step": 745 + }, + { + "epoch": 2.49, + "learning_rate": 8e-06, + "loss": 154.6651, + "step": 746 + }, + { + "epoch": 2.49, + "learning_rate": 8e-06, + "loss": 129.8414, + "step": 747 + }, + { + "epoch": 2.49, + "learning_rate": 8e-06, + "loss": 183.9736, + "step": 748 + }, + { + "epoch": 2.5, + "learning_rate": 8e-06, + "loss": 172.8187, + "step": 749 + }, + { + "epoch": 2.5, + "learning_rate": 8e-06, + "loss": 149.0044, + "step": 750 + }, + { + "epoch": 2.5, + "eval_loss": 187.71217346191406, + "eval_runtime": 140.4705, + "eval_samples_per_second": 34.171, + "eval_steps_per_second": 2.136, + "step": 750 + }, + { + "epoch": 2.5, + "learning_rate": 8e-06, + "loss": 203.7368, + "step": 751 + }, + { + "epoch": 2.51, + "learning_rate": 8e-06, + "loss": 151.7148, + "step": 752 + }, + { + "epoch": 2.51, + "learning_rate": 8e-06, + "loss": 143.8918, + "step": 753 + }, + { + "epoch": 2.51, + "learning_rate": 8e-06, + "loss": 144.7953, + "step": 754 + }, + { + "epoch": 2.52, + "learning_rate": 8e-06, + "loss": 191.2098, + "step": 755 + }, + { + "epoch": 2.52, + "learning_rate": 8e-06, + "loss": 157.7352, + "step": 756 + }, + { + "epoch": 2.52, + "learning_rate": 8e-06, + "loss": 195.8432, + "step": 757 + }, + { + "epoch": 2.53, + "learning_rate": 8e-06, + "loss": 149.4046, + "step": 758 + }, + { + "epoch": 2.53, + "learning_rate": 8e-06, + "loss": 165.5607, + "step": 759 + }, + { + "epoch": 2.53, + "learning_rate": 8e-06, + "loss": 145.3999, + "step": 760 + }, + { + "epoch": 2.54, + "learning_rate": 8e-06, + "loss": 139.3573, + "step": 761 + }, + { + "epoch": 2.54, + "learning_rate": 8e-06, + "loss": 169.0986, + "step": 762 + }, + { + "epoch": 2.54, + "learning_rate": 8e-06, + "loss": 198.4961, + "step": 763 + }, + { + "epoch": 2.55, + "learning_rate": 8e-06, + "loss": 163.7185, + "step": 764 + }, + { + "epoch": 2.55, + "learning_rate": 8e-06, + "loss": 157.8521, + "step": 765 + }, + { + "epoch": 2.55, + "learning_rate": 8e-06, + "loss": 162.5461, + "step": 766 + }, + { + "epoch": 2.56, + "learning_rate": 8e-06, + "loss": 155.7953, + "step": 767 + }, + { + "epoch": 2.56, + "learning_rate": 8e-06, + "loss": 178.8524, + "step": 768 + }, + { + "epoch": 2.56, + "learning_rate": 8e-06, + "loss": 163.8322, + "step": 769 + }, + { + "epoch": 2.57, + "learning_rate": 8e-06, + "loss": 146.2827, + "step": 770 + }, + { + "epoch": 2.57, + "learning_rate": 8e-06, + "loss": 138.1372, + "step": 771 + }, + { + "epoch": 2.57, + "learning_rate": 8e-06, + "loss": 147.3132, + "step": 772 + }, + { + "epoch": 2.58, + "learning_rate": 8e-06, + "loss": 158.0063, + "step": 773 + }, + { + "epoch": 2.58, + "learning_rate": 8e-06, + "loss": 183.7001, + "step": 774 + }, + { + "epoch": 2.58, + "learning_rate": 8e-06, + "loss": 131.2975, + "step": 775 + }, + { + "epoch": 2.59, + "learning_rate": 8e-06, + "loss": 169.3791, + "step": 776 + }, + { + "epoch": 2.59, + "learning_rate": 8e-06, + "loss": 172.7045, + "step": 777 + }, + { + "epoch": 2.59, + "learning_rate": 8e-06, + "loss": 183.9179, + "step": 778 + }, + { + "epoch": 2.6, + "learning_rate": 8e-06, + "loss": 153.5893, + "step": 779 + }, + { + "epoch": 2.6, + "learning_rate": 8e-06, + "loss": 144.8223, + "step": 780 + }, + { + "epoch": 2.6, + "learning_rate": 8e-06, + "loss": 170.5944, + "step": 781 + }, + { + "epoch": 2.61, + "learning_rate": 8e-06, + "loss": 136.3611, + "step": 782 + }, + { + "epoch": 2.61, + "learning_rate": 8e-06, + "loss": 168.1483, + "step": 783 + }, + { + "epoch": 2.61, + "learning_rate": 8e-06, + "loss": 132.7393, + "step": 784 + }, + { + "epoch": 2.62, + "learning_rate": 8e-06, + "loss": 169.2686, + "step": 785 + }, + { + "epoch": 2.62, + "learning_rate": 8e-06, + "loss": 139.7592, + "step": 786 + }, + { + "epoch": 2.62, + "learning_rate": 8e-06, + "loss": 156.953, + "step": 787 + }, + { + "epoch": 2.63, + "learning_rate": 8e-06, + "loss": 166.6686, + "step": 788 + }, + { + "epoch": 2.63, + "learning_rate": 8e-06, + "loss": 207.3799, + "step": 789 + }, + { + "epoch": 2.63, + "learning_rate": 8e-06, + "loss": 169.0426, + "step": 790 + }, + { + "epoch": 2.64, + "learning_rate": 8e-06, + "loss": 148.4689, + "step": 791 + }, + { + "epoch": 2.64, + "learning_rate": 8e-06, + "loss": 170.8208, + "step": 792 + }, + { + "epoch": 2.64, + "learning_rate": 8e-06, + "loss": 148.8021, + "step": 793 + }, + { + "epoch": 2.65, + "learning_rate": 8e-06, + "loss": 170.2678, + "step": 794 + }, + { + "epoch": 2.65, + "learning_rate": 8e-06, + "loss": 176.5266, + "step": 795 + }, + { + "epoch": 2.65, + "learning_rate": 8e-06, + "loss": 155.2612, + "step": 796 + }, + { + "epoch": 2.66, + "learning_rate": 8e-06, + "loss": 170.6114, + "step": 797 + }, + { + "epoch": 2.66, + "learning_rate": 8e-06, + "loss": 181.4791, + "step": 798 + }, + { + "epoch": 2.66, + "learning_rate": 8e-06, + "loss": 158.4708, + "step": 799 + }, + { + "epoch": 2.67, + "learning_rate": 8e-06, + "loss": 178.9341, + "step": 800 + }, + { + "epoch": 2.67, + "learning_rate": 8e-06, + "loss": 171.6622, + "step": 801 + }, + { + "epoch": 2.67, + "learning_rate": 8e-06, + "loss": 170.4995, + "step": 802 + }, + { + "epoch": 2.68, + "learning_rate": 8e-06, + "loss": 165.4084, + "step": 803 + }, + { + "epoch": 2.68, + "learning_rate": 8e-06, + "loss": 216.8477, + "step": 804 + }, + { + "epoch": 2.68, + "learning_rate": 8e-06, + "loss": 208.0631, + "step": 805 + }, + { + "epoch": 2.69, + "learning_rate": 8e-06, + "loss": 157.7889, + "step": 806 + }, + { + "epoch": 2.69, + "learning_rate": 8e-06, + "loss": 130.6747, + "step": 807 + }, + { + "epoch": 2.69, + "learning_rate": 8e-06, + "loss": 155.1428, + "step": 808 + }, + { + "epoch": 2.7, + "learning_rate": 8e-06, + "loss": 125.2736, + "step": 809 + }, + { + "epoch": 2.7, + "learning_rate": 8e-06, + "loss": 153.7274, + "step": 810 + }, + { + "epoch": 2.7, + "learning_rate": 8e-06, + "loss": 171.2344, + "step": 811 + }, + { + "epoch": 2.71, + "learning_rate": 8e-06, + "loss": 141.068, + "step": 812 + }, + { + "epoch": 2.71, + "learning_rate": 8e-06, + "loss": 171.9363, + "step": 813 + }, + { + "epoch": 2.71, + "learning_rate": 8e-06, + "loss": 137.3767, + "step": 814 + }, + { + "epoch": 2.72, + "learning_rate": 8e-06, + "loss": 154.4406, + "step": 815 + }, + { + "epoch": 2.72, + "learning_rate": 8e-06, + "loss": 152.9395, + "step": 816 + }, + { + "epoch": 2.72, + "learning_rate": 8e-06, + "loss": 174.0528, + "step": 817 + }, + { + "epoch": 2.73, + "learning_rate": 8e-06, + "loss": 161.1511, + "step": 818 + }, + { + "epoch": 2.73, + "learning_rate": 8e-06, + "loss": 183.789, + "step": 819 + }, + { + "epoch": 2.73, + "learning_rate": 8e-06, + "loss": 163.4563, + "step": 820 + }, + { + "epoch": 2.74, + "learning_rate": 8e-06, + "loss": 142.3525, + "step": 821 + }, + { + "epoch": 2.74, + "learning_rate": 8e-06, + "loss": 169.8715, + "step": 822 + }, + { + "epoch": 2.74, + "learning_rate": 8e-06, + "loss": 159.4444, + "step": 823 + }, + { + "epoch": 2.75, + "learning_rate": 8e-06, + "loss": 152.8666, + "step": 824 + }, + { + "epoch": 2.75, + "learning_rate": 8e-06, + "loss": 137.0646, + "step": 825 + }, + { + "epoch": 2.75, + "learning_rate": 8e-06, + "loss": 146.0876, + "step": 826 + }, + { + "epoch": 2.76, + "learning_rate": 8e-06, + "loss": 202.4309, + "step": 827 + }, + { + "epoch": 2.76, + "learning_rate": 8e-06, + "loss": 136.3753, + "step": 828 + }, + { + "epoch": 2.76, + "learning_rate": 8e-06, + "loss": 139.9753, + "step": 829 + }, + { + "epoch": 2.77, + "learning_rate": 8e-06, + "loss": 181.1168, + "step": 830 + }, + { + "epoch": 2.77, + "learning_rate": 8e-06, + "loss": 164.8888, + "step": 831 + }, + { + "epoch": 2.77, + "learning_rate": 8e-06, + "loss": 128.8818, + "step": 832 + }, + { + "epoch": 2.78, + "learning_rate": 8e-06, + "loss": 136.7745, + "step": 833 + }, + { + "epoch": 2.78, + "learning_rate": 8e-06, + "loss": 145.64, + "step": 834 + }, + { + "epoch": 2.78, + "learning_rate": 8e-06, + "loss": 162.9099, + "step": 835 + }, + { + "epoch": 2.79, + "learning_rate": 8e-06, + "loss": 178.4833, + "step": 836 + }, + { + "epoch": 2.79, + "learning_rate": 8e-06, + "loss": 125.9664, + "step": 837 + }, + { + "epoch": 2.79, + "learning_rate": 8e-06, + "loss": 185.7289, + "step": 838 + }, + { + "epoch": 2.8, + "learning_rate": 8e-06, + "loss": 133.5958, + "step": 839 + }, + { + "epoch": 2.8, + "learning_rate": 8e-06, + "loss": 177.2631, + "step": 840 + }, + { + "epoch": 2.8, + "learning_rate": 8e-06, + "loss": 150.5079, + "step": 841 + }, + { + "epoch": 2.81, + "learning_rate": 8e-06, + "loss": 169.4113, + "step": 842 + }, + { + "epoch": 2.81, + "learning_rate": 8e-06, + "loss": 147.1604, + "step": 843 + }, + { + "epoch": 2.81, + "learning_rate": 8e-06, + "loss": 198.2325, + "step": 844 + }, + { + "epoch": 2.82, + "learning_rate": 8e-06, + "loss": 135.1947, + "step": 845 + }, + { + "epoch": 2.82, + "learning_rate": 8e-06, + "loss": 155.64, + "step": 846 + }, + { + "epoch": 2.82, + "learning_rate": 8e-06, + "loss": 165.5759, + "step": 847 + }, + { + "epoch": 2.83, + "learning_rate": 8e-06, + "loss": 192.6142, + "step": 848 + }, + { + "epoch": 2.83, + "learning_rate": 8e-06, + "loss": 181.6491, + "step": 849 + }, + { + "epoch": 2.83, + "learning_rate": 8e-06, + "loss": 144.8011, + "step": 850 + }, + { + "epoch": 2.84, + "learning_rate": 8e-06, + "loss": 171.0445, + "step": 851 + }, + { + "epoch": 2.84, + "learning_rate": 8e-06, + "loss": 151.7704, + "step": 852 + }, + { + "epoch": 2.84, + "learning_rate": 8e-06, + "loss": 187.4667, + "step": 853 + }, + { + "epoch": 2.85, + "learning_rate": 8e-06, + "loss": 165.5133, + "step": 854 + }, + { + "epoch": 2.85, + "learning_rate": 8e-06, + "loss": 152.9709, + "step": 855 + }, + { + "epoch": 2.85, + "learning_rate": 8e-06, + "loss": 148.6145, + "step": 856 + }, + { + "epoch": 2.86, + "learning_rate": 8e-06, + "loss": 169.0189, + "step": 857 + }, + { + "epoch": 2.86, + "learning_rate": 8e-06, + "loss": 152.7328, + "step": 858 + }, + { + "epoch": 2.86, + "learning_rate": 8e-06, + "loss": 150.3401, + "step": 859 + }, + { + "epoch": 2.87, + "learning_rate": 8e-06, + "loss": 177.4704, + "step": 860 + }, + { + "epoch": 2.87, + "learning_rate": 8e-06, + "loss": 144.8234, + "step": 861 + }, + { + "epoch": 2.87, + "learning_rate": 8e-06, + "loss": 145.4043, + "step": 862 + }, + { + "epoch": 2.88, + "learning_rate": 8e-06, + "loss": 124.1652, + "step": 863 + }, + { + "epoch": 2.88, + "learning_rate": 8e-06, + "loss": 133.8708, + "step": 864 + }, + { + "epoch": 2.88, + "learning_rate": 8e-06, + "loss": 126.4331, + "step": 865 + }, + { + "epoch": 2.89, + "learning_rate": 8e-06, + "loss": 165.6183, + "step": 866 + }, + { + "epoch": 2.89, + "learning_rate": 8e-06, + "loss": 154.3025, + "step": 867 + }, + { + "epoch": 2.89, + "learning_rate": 8e-06, + "loss": 180.9241, + "step": 868 + }, + { + "epoch": 2.9, + "learning_rate": 8e-06, + "loss": 140.5349, + "step": 869 + }, + { + "epoch": 2.9, + "learning_rate": 8e-06, + "loss": 166.5607, + "step": 870 + }, + { + "epoch": 2.9, + "learning_rate": 8e-06, + "loss": 205.8565, + "step": 871 + }, + { + "epoch": 2.91, + "learning_rate": 8e-06, + "loss": 178.8804, + "step": 872 + }, + { + "epoch": 2.91, + "learning_rate": 8e-06, + "loss": 177.8319, + "step": 873 + }, + { + "epoch": 2.91, + "learning_rate": 8e-06, + "loss": 182.9185, + "step": 874 + }, + { + "epoch": 2.92, + "learning_rate": 8e-06, + "loss": 209.176, + "step": 875 + }, + { + "epoch": 2.92, + "learning_rate": 8e-06, + "loss": 165.3932, + "step": 876 + }, + { + "epoch": 2.92, + "learning_rate": 8e-06, + "loss": 133.6421, + "step": 877 + }, + { + "epoch": 2.93, + "learning_rate": 8e-06, + "loss": 172.0088, + "step": 878 + }, + { + "epoch": 2.93, + "learning_rate": 8e-06, + "loss": 134.22, + "step": 879 + }, + { + "epoch": 2.93, + "learning_rate": 8e-06, + "loss": 176.1048, + "step": 880 + }, + { + "epoch": 2.94, + "learning_rate": 8e-06, + "loss": 133.1551, + "step": 881 + }, + { + "epoch": 2.94, + "learning_rate": 8e-06, + "loss": 200.7317, + "step": 882 + }, + { + "epoch": 2.94, + "learning_rate": 8e-06, + "loss": 164.1982, + "step": 883 + }, + { + "epoch": 2.95, + "learning_rate": 8e-06, + "loss": 134.334, + "step": 884 + }, + { + "epoch": 2.95, + "learning_rate": 8e-06, + "loss": 151.3014, + "step": 885 + }, + { + "epoch": 2.95, + "learning_rate": 8e-06, + "loss": 167.496, + "step": 886 + }, + { + "epoch": 2.96, + "learning_rate": 8e-06, + "loss": 163.0481, + "step": 887 + }, + { + "epoch": 2.96, + "learning_rate": 8e-06, + "loss": 151.8579, + "step": 888 + }, + { + "epoch": 2.96, + "learning_rate": 8e-06, + "loss": 150.1856, + "step": 889 + }, + { + "epoch": 2.97, + "learning_rate": 8e-06, + "loss": 155.185, + "step": 890 + }, + { + "epoch": 2.97, + "learning_rate": 8e-06, + "loss": 162.8825, + "step": 891 + }, + { + "epoch": 2.97, + "learning_rate": 8e-06, + "loss": 131.0793, + "step": 892 + }, + { + "epoch": 2.98, + "learning_rate": 8e-06, + "loss": 166.3559, + "step": 893 + }, + { + "epoch": 2.98, + "learning_rate": 8e-06, + "loss": 159.2781, + "step": 894 + }, + { + "epoch": 2.98, + "learning_rate": 8e-06, + "loss": 183.3366, + "step": 895 + }, + { + "epoch": 2.99, + "learning_rate": 8e-06, + "loss": 136.9031, + "step": 896 + }, + { + "epoch": 2.99, + "learning_rate": 8e-06, + "loss": 159.9371, + "step": 897 + }, + { + "epoch": 2.99, + "learning_rate": 8e-06, + "loss": 153.7727, + "step": 898 + }, + { + "epoch": 3.0, + "learning_rate": 8e-06, + "loss": 151.6584, + "step": 899 + }, + { + "epoch": 3.0, + "learning_rate": 8e-06, + "loss": 151.9758, + "step": 900 + }, + { + "epoch": 3.0, + "eval_loss": 180.1415557861328, + "eval_runtime": 140.6334, + "eval_samples_per_second": 34.131, + "eval_steps_per_second": 2.133, + "step": 900 + } + ], + "logging_steps": 1.0, + "max_steps": 1200, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 20, + "total_flos": 3.661088642039808e+17, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}