diff --git "a/mixtral-8x7b/trainer_state.json" "b/mixtral-8x7b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/mixtral-8x7b/trainer_state.json" @@ -0,0 +1,18999 @@ +{ + "best_metric": 1.9152195453643799, + "best_model_checkpoint": "./lora-out/checkpoint-2492", + "epoch": 2.2076541459957477, + "eval_steps": 89, + "global_step": 3115, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 2.0441, + "step": 1 + }, + { + "epoch": 0.0, + "eval_loss": 1.9822026491165161, + "eval_runtime": 63.2922, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 1.9015, + "step": 2 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 2.1358, + "step": 3 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 2.1365, + "step": 4 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 2.2622, + "step": 5 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 1.844, + "step": 6 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001, + "loss": 1.914, + "step": 7 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.631, + "step": 8 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9554, + "step": 9 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9661, + "step": 10 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9605, + "step": 11 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.8921, + "step": 12 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.7569, + "step": 13 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9932, + "step": 14 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.7259, + "step": 15 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.8129, + "step": 16 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9666, + "step": 17 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.9193, + "step": 18 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 1.722, + "step": 19 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 2.0487, + "step": 20 + }, + { + "epoch": 0.01, + "learning_rate": 0.0001, + "loss": 2.175, + "step": 21 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.5683, + "step": 22 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.9788, + "step": 23 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 2.0265, + "step": 24 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 2.278, + "step": 25 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.758, + "step": 26 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 2.0227, + "step": 27 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 2.1572, + "step": 28 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.7157, + "step": 29 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 4.0886, + "step": 30 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.791, + "step": 31 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.9486, + "step": 32 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 2.1066, + "step": 33 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.9275, + "step": 34 + }, + { + "epoch": 0.02, + "learning_rate": 0.0001, + "loss": 1.936, + "step": 35 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.791, + "step": 36 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.5042, + "step": 37 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 2.0767, + "step": 38 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.8809, + "step": 39 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.8557, + "step": 40 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.9781, + "step": 41 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.7982, + "step": 42 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.4546, + "step": 43 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.8912, + "step": 44 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.6241, + "step": 45 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.6991, + "step": 46 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.7723, + "step": 47 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.8136, + "step": 48 + }, + { + "epoch": 0.03, + "learning_rate": 0.0001, + "loss": 1.8724, + "step": 49 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.9829, + "step": 50 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.6399, + "step": 51 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.7688, + "step": 52 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 2.135, + "step": 53 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.7454, + "step": 54 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.9948, + "step": 55 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 2.1128, + "step": 56 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.8485, + "step": 57 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 2.1956, + "step": 58 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.7755, + "step": 59 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 2.2515, + "step": 60 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 2.5705, + "step": 61 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.8593, + "step": 62 + }, + { + "epoch": 0.04, + "learning_rate": 0.0001, + "loss": 1.7371, + "step": 63 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.8086, + "step": 64 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.8876, + "step": 65 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 2.1627, + "step": 66 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.9502, + "step": 67 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.8382, + "step": 68 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.7379, + "step": 69 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.5922, + "step": 70 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.6613, + "step": 71 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.7454, + "step": 72 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 2.0684, + "step": 73 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.7692, + "step": 74 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.6665, + "step": 75 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.8411, + "step": 76 + }, + { + "epoch": 0.05, + "learning_rate": 0.0001, + "loss": 1.9995, + "step": 77 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.6812, + "step": 78 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.6067, + "step": 79 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.8403, + "step": 80 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.7304, + "step": 81 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.7673, + "step": 82 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.528, + "step": 83 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 2.0542, + "step": 84 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.7475, + "step": 85 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 2.097, + "step": 86 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.8509, + "step": 87 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.9533, + "step": 88 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 2.1251, + "step": 89 + }, + { + "epoch": 0.06, + "eval_loss": 1.9632471799850464, + "eval_runtime": 63.4652, + "eval_samples_per_second": 0.236, + "eval_steps_per_second": 0.236, + "step": 89 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 1.6769, + "step": 90 + }, + { + "epoch": 0.06, + "learning_rate": 0.0001, + "loss": 2.1536, + "step": 91 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.4783, + "step": 92 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.9199, + "step": 93 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.8886, + "step": 94 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.6926, + "step": 95 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.9292, + "step": 96 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 2.2107, + "step": 97 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.9204, + "step": 98 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.7179, + "step": 99 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.7827, + "step": 100 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.7291, + "step": 101 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.8231, + "step": 102 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 2.0241, + "step": 103 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.7931, + "step": 104 + }, + { + "epoch": 0.07, + "learning_rate": 0.0001, + "loss": 1.8712, + "step": 105 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.9743, + "step": 106 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.7699, + "step": 107 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.8796, + "step": 108 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.7526, + "step": 109 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 2.1716, + "step": 110 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.633, + "step": 111 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.6981, + "step": 112 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 2.0707, + "step": 113 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 2.0217, + "step": 114 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.6625, + "step": 115 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.9057, + "step": 116 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.8445, + "step": 117 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 2.1162, + "step": 118 + }, + { + "epoch": 0.08, + "learning_rate": 0.0001, + "loss": 1.5243, + "step": 119 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.7193, + "step": 120 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.7774, + "step": 121 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.568, + "step": 122 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.9103, + "step": 123 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.3418, + "step": 124 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.801, + "step": 125 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.9232, + "step": 126 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.7224, + "step": 127 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.7781, + "step": 128 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.0237, + "step": 129 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.0258, + "step": 130 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.8393, + "step": 131 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.0934, + "step": 132 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 2.0792, + "step": 133 + }, + { + "epoch": 0.09, + "learning_rate": 0.0001, + "loss": 1.9444, + "step": 134 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.7126, + "step": 135 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.9845, + "step": 136 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 2.1419, + "step": 137 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 2.154, + "step": 138 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.8715, + "step": 139 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.6492, + "step": 140 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.8449, + "step": 141 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.7595, + "step": 142 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.7164, + "step": 143 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 2.1169, + "step": 144 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.8675, + "step": 145 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.8906, + "step": 146 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 1.5967, + "step": 147 + }, + { + "epoch": 0.1, + "learning_rate": 0.0001, + "loss": 2.303, + "step": 148 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.7977, + "step": 149 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.7305, + "step": 150 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.7159, + "step": 151 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.5643, + "step": 152 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.8055, + "step": 153 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.7207, + "step": 154 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.9894, + "step": 155 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.6363, + "step": 156 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 2.0697, + "step": 157 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 2.1718, + "step": 158 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 2.0707, + "step": 159 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.6, + "step": 160 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 2.1345, + "step": 161 + }, + { + "epoch": 0.11, + "learning_rate": 0.0001, + "loss": 1.931, + "step": 162 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.9392, + "step": 163 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.8305, + "step": 164 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.1091, + "step": 165 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.5311, + "step": 166 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.0479, + "step": 167 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.0156, + "step": 168 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.6879, + "step": 169 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.0537, + "step": 170 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.8243, + "step": 171 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.1142, + "step": 172 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.9021, + "step": 173 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 2.0993, + "step": 174 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.8727, + "step": 175 + }, + { + "epoch": 0.12, + "learning_rate": 0.0001, + "loss": 1.6527, + "step": 176 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.69, + "step": 177 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 3.2868, + "step": 178 + }, + { + "epoch": 0.13, + "eval_loss": 1.9570825099945068, + "eval_runtime": 63.7243, + "eval_samples_per_second": 0.235, + "eval_steps_per_second": 0.235, + "step": 178 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.9875, + "step": 179 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.5371, + "step": 180 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.8503, + "step": 181 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.8186, + "step": 182 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.9381, + "step": 183 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.9236, + "step": 184 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.5678, + "step": 185 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 2.0947, + "step": 186 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.7998, + "step": 187 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 2.0272, + "step": 188 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.9726, + "step": 189 + }, + { + "epoch": 0.13, + "learning_rate": 0.0001, + "loss": 1.7585, + "step": 190 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.9109, + "step": 191 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.9456, + "step": 192 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 0.9121, + "step": 193 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.8506, + "step": 194 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.8635, + "step": 195 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 2.1572, + "step": 196 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.7493, + "step": 197 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.898, + "step": 198 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.8544, + "step": 199 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.7581, + "step": 200 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.7702, + "step": 201 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.8468, + "step": 202 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.8313, + "step": 203 + }, + { + "epoch": 0.14, + "learning_rate": 0.0001, + "loss": 1.7005, + "step": 204 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.8035, + "step": 205 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.6932, + "step": 206 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.6722, + "step": 207 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 2.0045, + "step": 208 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.83, + "step": 209 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.9204, + "step": 210 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.8177, + "step": 211 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.7142, + "step": 212 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.6527, + "step": 213 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.9286, + "step": 214 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 2.2043, + "step": 215 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.6803, + "step": 216 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 1.7659, + "step": 217 + }, + { + "epoch": 0.15, + "learning_rate": 0.0001, + "loss": 2.0085, + "step": 218 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.6455, + "step": 219 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.9728, + "step": 220 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.9912, + "step": 221 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.7641, + "step": 222 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.8288, + "step": 223 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 2.1472, + "step": 224 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.601, + "step": 225 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.9818, + "step": 226 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.8828, + "step": 227 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.6993, + "step": 228 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.6852, + "step": 229 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.6201, + "step": 230 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 2.0215, + "step": 231 + }, + { + "epoch": 0.16, + "learning_rate": 0.0001, + "loss": 1.7473, + "step": 232 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 2.0189, + "step": 233 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 2.0028, + "step": 234 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.7689, + "step": 235 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 2.1462, + "step": 236 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.78, + "step": 237 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.789, + "step": 238 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.809, + "step": 239 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.9535, + "step": 240 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 2.0378, + "step": 241 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.9864, + "step": 242 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.6717, + "step": 243 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.7778, + "step": 244 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.6944, + "step": 245 + }, + { + "epoch": 0.17, + "learning_rate": 0.0001, + "loss": 1.6622, + "step": 246 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.9127, + "step": 247 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.9625, + "step": 248 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.8896, + "step": 249 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 2.2713, + "step": 250 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.8324, + "step": 251 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 2.1619, + "step": 252 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 2.0863, + "step": 253 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 2.0065, + "step": 254 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.9237, + "step": 255 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.7379, + "step": 256 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.6704, + "step": 257 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.7054, + "step": 258 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.9157, + "step": 259 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.6166, + "step": 260 + }, + { + "epoch": 0.18, + "learning_rate": 0.0001, + "loss": 1.6808, + "step": 261 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.9651, + "step": 262 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.8011, + "step": 263 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 2.068, + "step": 264 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 2.0059, + "step": 265 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.7092, + "step": 266 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.9871, + "step": 267 + }, + { + "epoch": 0.19, + "eval_loss": 1.9494396448135376, + "eval_runtime": 63.3247, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 267 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.709, + "step": 268 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.9579, + "step": 269 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.9672, + "step": 270 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.7042, + "step": 271 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 2.1476, + "step": 272 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.8924, + "step": 273 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.9211, + "step": 274 + }, + { + "epoch": 0.19, + "learning_rate": 0.0001, + "loss": 1.7151, + "step": 275 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.8345, + "step": 276 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.2576, + "step": 277 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.9847, + "step": 278 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.8366, + "step": 279 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.0653, + "step": 280 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.0599, + "step": 281 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.1345, + "step": 282 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.6369, + "step": 283 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.2584, + "step": 284 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.9308, + "step": 285 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 2.2002, + "step": 286 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.5348, + "step": 287 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.6495, + "step": 288 + }, + { + "epoch": 0.2, + "learning_rate": 0.0001, + "loss": 1.8517, + "step": 289 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8762, + "step": 290 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 2.0101, + "step": 291 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8954, + "step": 292 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8616, + "step": 293 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.6444, + "step": 294 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.9117, + "step": 295 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8955, + "step": 296 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 2.1447, + "step": 297 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8076, + "step": 298 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 2.7244, + "step": 299 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 2.0754, + "step": 300 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8664, + "step": 301 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8048, + "step": 302 + }, + { + "epoch": 0.21, + "learning_rate": 0.0001, + "loss": 1.8097, + "step": 303 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.8538, + "step": 304 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.7444, + "step": 305 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.6428, + "step": 306 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.7982, + "step": 307 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 2.2178, + "step": 308 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.8692, + "step": 309 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.9983, + "step": 310 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 2.1014, + "step": 311 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 2.0145, + "step": 312 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.824, + "step": 313 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.5223, + "step": 314 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.5907, + "step": 315 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 2.147, + "step": 316 + }, + { + "epoch": 0.22, + "learning_rate": 0.0001, + "loss": 1.9247, + "step": 317 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.9899, + "step": 318 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 2.0483, + "step": 319 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.6109, + "step": 320 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 2.1799, + "step": 321 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 2.0768, + "step": 322 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.9323, + "step": 323 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.776, + "step": 324 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.7566, + "step": 325 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.9118, + "step": 326 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.783, + "step": 327 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.757, + "step": 328 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.9424, + "step": 329 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.8284, + "step": 330 + }, + { + "epoch": 0.23, + "learning_rate": 0.0001, + "loss": 1.971, + "step": 331 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 2.0255, + "step": 332 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 2.0988, + "step": 333 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.8124, + "step": 334 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 2.2742, + "step": 335 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 2.1846, + "step": 336 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.7225, + "step": 337 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 2.3234, + "step": 338 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.8142, + "step": 339 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.8602, + "step": 340 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.8396, + "step": 341 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.9045, + "step": 342 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.7244, + "step": 343 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.8749, + "step": 344 + }, + { + "epoch": 0.24, + "learning_rate": 0.0001, + "loss": 1.7685, + "step": 345 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.7743, + "step": 346 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 2.1782, + "step": 347 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.7325, + "step": 348 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.7464, + "step": 349 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.6492, + "step": 350 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.7778, + "step": 351 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.8004, + "step": 352 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.9375, + "step": 353 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.6744, + "step": 354 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.6975, + "step": 355 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.6466, + "step": 356 + }, + { + "epoch": 0.25, + "eval_loss": 1.944254755973816, + "eval_runtime": 63.4073, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 356 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.9863, + "step": 357 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 2.005, + "step": 358 + }, + { + "epoch": 0.25, + "learning_rate": 0.0001, + "loss": 1.8214, + "step": 359 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.9551, + "step": 360 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 2.1185, + "step": 361 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.981, + "step": 362 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.8581, + "step": 363 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 2.1336, + "step": 364 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.6766, + "step": 365 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.8778, + "step": 366 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.937, + "step": 367 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 2.0644, + "step": 368 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.8518, + "step": 369 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.9557, + "step": 370 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 2.2913, + "step": 371 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 2.1082, + "step": 372 + }, + { + "epoch": 0.26, + "learning_rate": 0.0001, + "loss": 1.7143, + "step": 373 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.7147, + "step": 374 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.8943, + "step": 375 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.8132, + "step": 376 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.7547, + "step": 377 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.0025, + "step": 378 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.0956, + "step": 379 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.0107, + "step": 380 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.8986, + "step": 381 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.0028, + "step": 382 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.7969, + "step": 383 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.6261, + "step": 384 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.8805, + "step": 385 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 1.979, + "step": 386 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.2747, + "step": 387 + }, + { + "epoch": 0.27, + "learning_rate": 0.0001, + "loss": 2.0176, + "step": 388 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 2.7858, + "step": 389 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.6598, + "step": 390 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.9503, + "step": 391 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 2.1978, + "step": 392 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 2.162, + "step": 393 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.9228, + "step": 394 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.7917, + "step": 395 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.6078, + "step": 396 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.7725, + "step": 397 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.6455, + "step": 398 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 0.9346, + "step": 399 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.7049, + "step": 400 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 2.1055, + "step": 401 + }, + { + "epoch": 0.28, + "learning_rate": 0.0001, + "loss": 1.7491, + "step": 402 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.9236, + "step": 403 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.8366, + "step": 404 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 2.22, + "step": 405 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 2.4014, + "step": 406 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.6996, + "step": 407 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.9783, + "step": 408 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.8839, + "step": 409 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.9483, + "step": 410 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.9146, + "step": 411 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.9852, + "step": 412 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.6182, + "step": 413 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.5836, + "step": 414 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 1.8784, + "step": 415 + }, + { + "epoch": 0.29, + "learning_rate": 0.0001, + "loss": 2.3023, + "step": 416 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 2.2135, + "step": 417 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.8413, + "step": 418 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.8896, + "step": 419 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.737, + "step": 420 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.8703, + "step": 421 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 2.3035, + "step": 422 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.4616, + "step": 423 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 2.0092, + "step": 424 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.7678, + "step": 425 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.8462, + "step": 426 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.9557, + "step": 427 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.8582, + "step": 428 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.5842, + "step": 429 + }, + { + "epoch": 0.3, + "learning_rate": 0.0001, + "loss": 1.889, + "step": 430 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.8304, + "step": 431 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.498, + "step": 432 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.9945, + "step": 433 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.865, + "step": 434 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.7832, + "step": 435 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.7251, + "step": 436 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.9067, + "step": 437 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.9828, + "step": 438 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.8872, + "step": 439 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.7999, + "step": 440 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.7111, + "step": 441 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.6856, + "step": 442 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.8373, + "step": 443 + }, + { + "epoch": 0.31, + "learning_rate": 0.0001, + "loss": 1.8404, + "step": 444 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 2.0483, + "step": 445 + }, + { + "epoch": 0.32, + "eval_loss": 1.9440066814422607, + "eval_runtime": 63.3115, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 445 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.545, + "step": 446 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 2.0416, + "step": 447 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.8243, + "step": 448 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.7714, + "step": 449 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.7783, + "step": 450 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.8671, + "step": 451 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 2.2084, + "step": 452 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.9869, + "step": 453 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 2.0856, + "step": 454 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.8568, + "step": 455 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.8395, + "step": 456 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 1.8131, + "step": 457 + }, + { + "epoch": 0.32, + "learning_rate": 0.0001, + "loss": 2.1822, + "step": 458 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 2.0111, + "step": 459 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.7242, + "step": 460 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.9868, + "step": 461 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 2.2219, + "step": 462 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.67, + "step": 463 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 2.3471, + "step": 464 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.9498, + "step": 465 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 2.1829, + "step": 466 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.9964, + "step": 467 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.7926, + "step": 468 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.747, + "step": 469 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.9916, + "step": 470 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 1.8207, + "step": 471 + }, + { + "epoch": 0.33, + "learning_rate": 0.0001, + "loss": 2.2, + "step": 472 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.9425, + "step": 473 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.745, + "step": 474 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.7128, + "step": 475 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 2.1567, + "step": 476 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.8316, + "step": 477 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.9554, + "step": 478 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 2.3731, + "step": 479 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.9406, + "step": 480 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.7005, + "step": 481 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.945, + "step": 482 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.8803, + "step": 483 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.5901, + "step": 484 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.4123, + "step": 485 + }, + { + "epoch": 0.34, + "learning_rate": 0.0001, + "loss": 1.8637, + "step": 486 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 2.067, + "step": 487 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.6928, + "step": 488 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.7906, + "step": 489 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.8901, + "step": 490 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 2.3061, + "step": 491 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.7507, + "step": 492 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.835, + "step": 493 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 2.423, + "step": 494 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.9678, + "step": 495 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.9251, + "step": 496 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 2.0379, + "step": 497 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 2.1843, + "step": 498 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.9862, + "step": 499 + }, + { + "epoch": 0.35, + "learning_rate": 0.0001, + "loss": 1.8093, + "step": 500 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 2.0504, + "step": 501 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.6295, + "step": 502 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.9604, + "step": 503 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 2.1754, + "step": 504 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 2.3377, + "step": 505 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.9869, + "step": 506 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 2.1284, + "step": 507 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.9906, + "step": 508 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.9586, + "step": 509 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.9825, + "step": 510 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.77, + "step": 511 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.993, + "step": 512 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.3908, + "step": 513 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 2.136, + "step": 514 + }, + { + "epoch": 0.36, + "learning_rate": 0.0001, + "loss": 1.8478, + "step": 515 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.8101, + "step": 516 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.905, + "step": 517 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.5357, + "step": 518 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.7017, + "step": 519 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 2.3692, + "step": 520 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 2.338, + "step": 521 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.7604, + "step": 522 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.7367, + "step": 523 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.9482, + "step": 524 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.7398, + "step": 525 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.7947, + "step": 526 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.8654, + "step": 527 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 2.2075, + "step": 528 + }, + { + "epoch": 0.37, + "learning_rate": 0.0001, + "loss": 1.8858, + "step": 529 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.8016, + "step": 530 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.8049, + "step": 531 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.8417, + "step": 532 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.9884, + "step": 533 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 2.1068, + "step": 534 + }, + { + "epoch": 0.38, + "eval_loss": 1.9413771629333496, + "eval_runtime": 63.0102, + "eval_samples_per_second": 0.238, + "eval_steps_per_second": 0.238, + "step": 534 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.9799, + "step": 535 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.5547, + "step": 536 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.8111, + "step": 537 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 2.0949, + "step": 538 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 2.0072, + "step": 539 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.8145, + "step": 540 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 2.2703, + "step": 541 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.886, + "step": 542 + }, + { + "epoch": 0.38, + "learning_rate": 0.0001, + "loss": 1.7234, + "step": 543 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.8692, + "step": 544 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.6844, + "step": 545 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.8156, + "step": 546 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.003, + "step": 547 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.2078, + "step": 548 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.3006, + "step": 549 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.8744, + "step": 550 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.8663, + "step": 551 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.1766, + "step": 552 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.6755, + "step": 553 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.054, + "step": 554 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 2.1876, + "step": 555 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.7568, + "step": 556 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001, + "loss": 1.7363, + "step": 557 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.5848, + "step": 558 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.8316, + "step": 559 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.8921, + "step": 560 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 2.362, + "step": 561 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.8229, + "step": 562 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 2.0896, + "step": 563 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.663, + "step": 564 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 2.0634, + "step": 565 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.7088, + "step": 566 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.7795, + "step": 567 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 2.0422, + "step": 568 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.7394, + "step": 569 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 2.1291, + "step": 570 + }, + { + "epoch": 0.4, + "learning_rate": 0.0001, + "loss": 1.7925, + "step": 571 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.7514, + "step": 572 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.8402, + "step": 573 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.0839, + "step": 574 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.7288, + "step": 575 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.079, + "step": 576 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.0067, + "step": 577 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.9023, + "step": 578 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.8314, + "step": 579 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.2753, + "step": 580 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.6298, + "step": 581 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.1362, + "step": 582 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.9468, + "step": 583 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 1.7839, + "step": 584 + }, + { + "epoch": 0.41, + "learning_rate": 0.0001, + "loss": 2.2562, + "step": 585 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.9038, + "step": 586 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 2.4895, + "step": 587 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.7705, + "step": 588 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.896, + "step": 589 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.7152, + "step": 590 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.8502, + "step": 591 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.8092, + "step": 592 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.6355, + "step": 593 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 2.0179, + "step": 594 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.8476, + "step": 595 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.1576, + "step": 596 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.6568, + "step": 597 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 1.8347, + "step": 598 + }, + { + "epoch": 0.42, + "learning_rate": 0.0001, + "loss": 2.0324, + "step": 599 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.8477, + "step": 600 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.9895, + "step": 601 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.7988, + "step": 602 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 2.0336, + "step": 603 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 2.0764, + "step": 604 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 2.075, + "step": 605 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.817, + "step": 606 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.6837, + "step": 607 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.9109, + "step": 608 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.8303, + "step": 609 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.6014, + "step": 610 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.7344, + "step": 611 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.6878, + "step": 612 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001, + "loss": 1.6137, + "step": 613 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.8045, + "step": 614 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.8799, + "step": 615 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.9057, + "step": 616 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.8276, + "step": 617 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.9961, + "step": 618 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.7941, + "step": 619 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 2.0407, + "step": 620 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.9108, + "step": 621 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.723, + "step": 622 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.8459, + "step": 623 + }, + { + "epoch": 0.44, + "eval_loss": 1.9396083354949951, + "eval_runtime": 63.0773, + "eval_samples_per_second": 0.238, + "eval_steps_per_second": 0.238, + "step": 623 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.6835, + "step": 624 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 2.1874, + "step": 625 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.8581, + "step": 626 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001, + "loss": 1.759, + "step": 627 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.7377, + "step": 628 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.1439, + "step": 629 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.0657, + "step": 630 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.818, + "step": 631 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.1494, + "step": 632 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.0133, + "step": 633 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.2054, + "step": 634 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.3004, + "step": 635 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.7399, + "step": 636 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.459, + "step": 637 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 2.162, + "step": 638 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.7658, + "step": 639 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.7536, + "step": 640 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.928, + "step": 641 + }, + { + "epoch": 0.45, + "learning_rate": 0.0001, + "loss": 1.7327, + "step": 642 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.6013, + "step": 643 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.8346, + "step": 644 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.7372, + "step": 645 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 2.0391, + "step": 646 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.6627, + "step": 647 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.8777, + "step": 648 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.6497, + "step": 649 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.8247, + "step": 650 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 2.1537, + "step": 651 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.3305, + "step": 652 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 2.1525, + "step": 653 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.4565, + "step": 654 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 1.7444, + "step": 655 + }, + { + "epoch": 0.46, + "learning_rate": 0.0001, + "loss": 2.1156, + "step": 656 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.9933, + "step": 657 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.7606, + "step": 658 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.8464, + "step": 659 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.5007, + "step": 660 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 2.0624, + "step": 661 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.8997, + "step": 662 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.886, + "step": 663 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 2.1107, + "step": 664 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 2.148, + "step": 665 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 2.0676, + "step": 666 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 2.1844, + "step": 667 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.9355, + "step": 668 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.8314, + "step": 669 + }, + { + "epoch": 0.47, + "learning_rate": 0.0001, + "loss": 1.7697, + "step": 670 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.8071, + "step": 671 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.5623, + "step": 672 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.4279, + "step": 673 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.6706, + "step": 674 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.9821, + "step": 675 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.8106, + "step": 676 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.7764, + "step": 677 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.7891, + "step": 678 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.7374, + "step": 679 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.6009, + "step": 680 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.7285, + "step": 681 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 2.4078, + "step": 682 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.9803, + "step": 683 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001, + "loss": 1.8895, + "step": 684 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.7304, + "step": 685 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.0424, + "step": 686 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.8297, + "step": 687 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.6459, + "step": 688 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.1642, + "step": 689 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.3089, + "step": 690 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.9653, + "step": 691 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.7582, + "step": 692 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.6467, + "step": 693 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.03, + "step": 694 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.6653, + "step": 695 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 1.6032, + "step": 696 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.1151, + "step": 697 + }, + { + "epoch": 0.49, + "learning_rate": 0.0001, + "loss": 2.1455, + "step": 698 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.0998, + "step": 699 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.1302, + "step": 700 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.0294, + "step": 701 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 1.8739, + "step": 702 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.2857, + "step": 703 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 1.9235, + "step": 704 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.2205, + "step": 705 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 1.8419, + "step": 706 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.0892, + "step": 707 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.0241, + "step": 708 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.0023, + "step": 709 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 1.7279, + "step": 710 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 2.2514, + "step": 711 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001, + "loss": 1.9354, + "step": 712 + }, + { + "epoch": 0.5, + "eval_loss": 1.9365569353103638, + "eval_runtime": 63.9569, + "eval_samples_per_second": 0.235, + "eval_steps_per_second": 0.235, + "step": 712 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.052, + "step": 713 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.944, + "step": 714 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.2362, + "step": 715 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.9593, + "step": 716 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.9306, + "step": 717 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.0885, + "step": 718 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.0129, + "step": 719 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.1342, + "step": 720 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.071, + "step": 721 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 2.2079, + "step": 722 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.8284, + "step": 723 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.7264, + "step": 724 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.7658, + "step": 725 + }, + { + "epoch": 0.51, + "learning_rate": 0.0001, + "loss": 1.7782, + "step": 726 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.8509, + "step": 727 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 2.1615, + "step": 728 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.7406, + "step": 729 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.4262, + "step": 730 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.5428, + "step": 731 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.8637, + "step": 732 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.9984, + "step": 733 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.5804, + "step": 734 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.6744, + "step": 735 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.9612, + "step": 736 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 2.1637, + "step": 737 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 2.1121, + "step": 738 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.8631, + "step": 739 + }, + { + "epoch": 0.52, + "learning_rate": 0.0001, + "loss": 1.9868, + "step": 740 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 2.0133, + "step": 741 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.9917, + "step": 742 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.9804, + "step": 743 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.6401, + "step": 744 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.8266, + "step": 745 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.8297, + "step": 746 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.7772, + "step": 747 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.839, + "step": 748 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.6934, + "step": 749 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.9337, + "step": 750 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 2.2892, + "step": 751 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.6004, + "step": 752 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.8708, + "step": 753 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001, + "loss": 1.9127, + "step": 754 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.5868, + "step": 755 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.4811, + "step": 756 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 2.4089, + "step": 757 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.9274, + "step": 758 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.7435, + "step": 759 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.7207, + "step": 760 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 2.277, + "step": 761 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.7653, + "step": 762 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 2.0132, + "step": 763 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 2.0948, + "step": 764 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.6895, + "step": 765 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 2.2286, + "step": 766 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.7136, + "step": 767 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001, + "loss": 1.5417, + "step": 768 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 2.1137, + "step": 769 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.691, + "step": 770 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.7187, + "step": 771 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.8303, + "step": 772 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.4065, + "step": 773 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.7329, + "step": 774 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.8106, + "step": 775 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.7972, + "step": 776 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 2.0359, + "step": 777 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.859, + "step": 778 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.8877, + "step": 779 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.567, + "step": 780 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 2.1961, + "step": 781 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.4229, + "step": 782 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001, + "loss": 1.8659, + "step": 783 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.9721, + "step": 784 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.9758, + "step": 785 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.7506, + "step": 786 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.9322, + "step": 787 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.85, + "step": 788 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 2.1555, + "step": 789 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.828, + "step": 790 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.9786, + "step": 791 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.9605, + "step": 792 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 2.0576, + "step": 793 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.7564, + "step": 794 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.907, + "step": 795 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.8087, + "step": 796 + }, + { + "epoch": 0.56, + "learning_rate": 0.0001, + "loss": 1.4803, + "step": 797 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.8621, + "step": 798 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 2.1555, + "step": 799 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 2.0289, + "step": 800 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 2.1406, + "step": 801 + }, + { + "epoch": 0.57, + "eval_loss": 1.9425642490386963, + "eval_runtime": 63.6511, + "eval_samples_per_second": 0.236, + "eval_steps_per_second": 0.236, + "step": 801 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.9211, + "step": 802 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.6879, + "step": 803 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.969, + "step": 804 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.9358, + "step": 805 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.8678, + "step": 806 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.7304, + "step": 807 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 2.0451, + "step": 808 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 2.1152, + "step": 809 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.8649, + "step": 810 + }, + { + "epoch": 0.57, + "learning_rate": 0.0001, + "loss": 1.7242, + "step": 811 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 2.1989, + "step": 812 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 2.1717, + "step": 813 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.7763, + "step": 814 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.6333, + "step": 815 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.7224, + "step": 816 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 2.064, + "step": 817 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.7847, + "step": 818 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.8256, + "step": 819 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.899, + "step": 820 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.7559, + "step": 821 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.5752, + "step": 822 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.8866, + "step": 823 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 1.9823, + "step": 824 + }, + { + "epoch": 0.58, + "learning_rate": 0.0001, + "loss": 2.0882, + "step": 825 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 2.0811, + "step": 826 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.507, + "step": 827 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.5937, + "step": 828 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.758, + "step": 829 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.6184, + "step": 830 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 2.0566, + "step": 831 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.7442, + "step": 832 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 2.0686, + "step": 833 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.9081, + "step": 834 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.7023, + "step": 835 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.9597, + "step": 836 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 2.0299, + "step": 837 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 1.9144, + "step": 838 + }, + { + "epoch": 0.59, + "learning_rate": 0.0001, + "loss": 2.0248, + "step": 839 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.8697, + "step": 840 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.91, + "step": 841 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.6724, + "step": 842 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 2.1238, + "step": 843 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.9037, + "step": 844 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 2.0531, + "step": 845 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.7025, + "step": 846 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.7491, + "step": 847 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.7714, + "step": 848 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.6563, + "step": 849 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.6405, + "step": 850 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.703, + "step": 851 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 2.035, + "step": 852 + }, + { + "epoch": 0.6, + "learning_rate": 0.0001, + "loss": 1.7017, + "step": 853 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.7439, + "step": 854 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.6272, + "step": 855 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 2.2708, + "step": 856 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 2.0115, + "step": 857 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 2.1788, + "step": 858 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.9604, + "step": 859 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 2.1387, + "step": 860 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.6086, + "step": 861 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.7241, + "step": 862 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.6834, + "step": 863 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.8049, + "step": 864 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.9939, + "step": 865 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 2.1317, + "step": 866 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001, + "loss": 1.9823, + "step": 867 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.367, + "step": 868 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.7732, + "step": 869 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.9179, + "step": 870 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.4351, + "step": 871 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.7716, + "step": 872 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 2.0547, + "step": 873 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.7476, + "step": 874 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.8424, + "step": 875 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.8973, + "step": 876 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 2.0303, + "step": 877 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 2.1306, + "step": 878 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.6427, + "step": 879 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.7279, + "step": 880 + }, + { + "epoch": 0.62, + "learning_rate": 0.0001, + "loss": 1.8452, + "step": 881 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 2.0339, + "step": 882 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.6724, + "step": 883 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 2.0652, + "step": 884 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.9059, + "step": 885 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.7697, + "step": 886 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.915, + "step": 887 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.6871, + "step": 888 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 2.2855, + "step": 889 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.7402, + "step": 890 + }, + { + "epoch": 0.63, + "eval_loss": 1.9399702548980713, + "eval_runtime": 63.6234, + "eval_samples_per_second": 0.236, + "eval_steps_per_second": 0.236, + "step": 890 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.8773, + "step": 891 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.7505, + "step": 892 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.8977, + "step": 893 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.7634, + "step": 894 + }, + { + "epoch": 0.63, + "learning_rate": 0.0001, + "loss": 1.877, + "step": 895 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.7833, + "step": 896 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.8658, + "step": 897 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.6749, + "step": 898 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.9534, + "step": 899 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 2.1272, + "step": 900 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.8252, + "step": 901 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 2.1331, + "step": 902 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.6467, + "step": 903 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.9082, + "step": 904 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.8215, + "step": 905 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.6768, + "step": 906 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 2.0549, + "step": 907 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.8063, + "step": 908 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 1.7412, + "step": 909 + }, + { + "epoch": 0.64, + "learning_rate": 0.0001, + "loss": 2.023, + "step": 910 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.4766, + "step": 911 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.6504, + "step": 912 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.9332, + "step": 913 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 2.2145, + "step": 914 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.9344, + "step": 915 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 2.0618, + "step": 916 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 2.0762, + "step": 917 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.8004, + "step": 918 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.8428, + "step": 919 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.9834, + "step": 920 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.6267, + "step": 921 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 2.0119, + "step": 922 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.787, + "step": 923 + }, + { + "epoch": 0.65, + "learning_rate": 0.0001, + "loss": 1.7364, + "step": 924 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 2.3403, + "step": 925 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.9449, + "step": 926 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.5749, + "step": 927 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.6718, + "step": 928 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.7097, + "step": 929 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.7844, + "step": 930 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.7391, + "step": 931 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.673, + "step": 932 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.7397, + "step": 933 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 2.1254, + "step": 934 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.9851, + "step": 935 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.9661, + "step": 936 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.6277, + "step": 937 + }, + { + "epoch": 0.66, + "learning_rate": 0.0001, + "loss": 1.9356, + "step": 938 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.9407, + "step": 939 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 2.1318, + "step": 940 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.8852, + "step": 941 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.7339, + "step": 942 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.8122, + "step": 943 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.9678, + "step": 944 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.8383, + "step": 945 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.5181, + "step": 946 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 2.1875, + "step": 947 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.9489, + "step": 948 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.7456, + "step": 949 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.9907, + "step": 950 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 1.6801, + "step": 951 + }, + { + "epoch": 0.67, + "learning_rate": 0.0001, + "loss": 2.0002, + "step": 952 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.897, + "step": 953 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.7125, + "step": 954 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.9057, + "step": 955 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.8706, + "step": 956 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.9746, + "step": 957 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 2.0179, + "step": 958 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.7476, + "step": 959 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.6901, + "step": 960 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 2.1064, + "step": 961 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.5333, + "step": 962 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.8571, + "step": 963 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.5436, + "step": 964 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.8606, + "step": 965 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001, + "loss": 1.6715, + "step": 966 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.9352, + "step": 967 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.8625, + "step": 968 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.1652, + "step": 969 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.0572, + "step": 970 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.878, + "step": 971 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.6297, + "step": 972 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.1613, + "step": 973 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.6987, + "step": 974 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.2919, + "step": 975 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.0742, + "step": 976 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 2.0878, + "step": 977 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.9415, + "step": 978 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.6681, + "step": 979 + }, + { + "epoch": 0.69, + "eval_loss": 1.9275270700454712, + "eval_runtime": 63.3457, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 979 + }, + { + "epoch": 0.69, + "learning_rate": 0.0001, + "loss": 1.9574, + "step": 980 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.9479, + "step": 981 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 2.0122, + "step": 982 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.8012, + "step": 983 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 2.2897, + "step": 984 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 2.2231, + "step": 985 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.8032, + "step": 986 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 2.0259, + "step": 987 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.9555, + "step": 988 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.7052, + "step": 989 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.8336, + "step": 990 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 2.3175, + "step": 991 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.9695, + "step": 992 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.928, + "step": 993 + }, + { + "epoch": 0.7, + "learning_rate": 0.0001, + "loss": 1.9005, + "step": 994 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.0191, + "step": 995 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.9545, + "step": 996 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.9822, + "step": 997 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.1344, + "step": 998 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.7322, + "step": 999 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.0879, + "step": 1000 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.5315, + "step": 1001 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.2744, + "step": 1002 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.9764, + "step": 1003 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.1761, + "step": 1004 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.8844, + "step": 1005 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.7518, + "step": 1006 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 2.014, + "step": 1007 + }, + { + "epoch": 0.71, + "learning_rate": 0.0001, + "loss": 1.9195, + "step": 1008 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 2.4106, + "step": 1009 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.8069, + "step": 1010 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.8921, + "step": 1011 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.7956, + "step": 1012 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.9134, + "step": 1013 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.7994, + "step": 1014 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.8817, + "step": 1015 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.8973, + "step": 1016 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.1899, + "step": 1017 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.7837, + "step": 1018 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 2.1721, + "step": 1019 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.9353, + "step": 1020 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 2.2914, + "step": 1021 + }, + { + "epoch": 0.72, + "learning_rate": 0.0001, + "loss": 1.5998, + "step": 1022 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 2.018, + "step": 1023 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.9586, + "step": 1024 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.7786, + "step": 1025 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.9252, + "step": 1026 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.9068, + "step": 1027 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.7848, + "step": 1028 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.6465, + "step": 1029 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.6251, + "step": 1030 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.8101, + "step": 1031 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.8243, + "step": 1032 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.9419, + "step": 1033 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 2.2066, + "step": 1034 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.7756, + "step": 1035 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.6383, + "step": 1036 + }, + { + "epoch": 0.73, + "learning_rate": 0.0001, + "loss": 1.977, + "step": 1037 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.7765, + "step": 1038 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.6438, + "step": 1039 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.7157, + "step": 1040 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.5114, + "step": 1041 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.3201, + "step": 1042 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.048, + "step": 1043 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.1625, + "step": 1044 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.719, + "step": 1045 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.9991, + "step": 1046 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.2528, + "step": 1047 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.8541, + "step": 1048 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 2.041, + "step": 1049 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.6358, + "step": 1050 + }, + { + "epoch": 0.74, + "learning_rate": 0.0001, + "loss": 1.887, + "step": 1051 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8368, + "step": 1052 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 2.0044, + "step": 1053 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8862, + "step": 1054 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8118, + "step": 1055 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8667, + "step": 1056 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 2.1339, + "step": 1057 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 2.0375, + "step": 1058 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8577, + "step": 1059 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8087, + "step": 1060 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.8469, + "step": 1061 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.9569, + "step": 1062 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 2.018, + "step": 1063 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.6045, + "step": 1064 + }, + { + "epoch": 0.75, + "learning_rate": 0.0001, + "loss": 1.7991, + "step": 1065 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.757, + "step": 1066 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 2.4132, + "step": 1067 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 2.1097, + "step": 1068 + }, + { + "epoch": 0.76, + "eval_loss": 1.9269508123397827, + "eval_runtime": 63.2347, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1068 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 2.0247, + "step": 1069 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.7639, + "step": 1070 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.7065, + "step": 1071 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.6924, + "step": 1072 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.6135, + "step": 1073 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.8963, + "step": 1074 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 2.2926, + "step": 1075 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.7036, + "step": 1076 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 2.1936, + "step": 1077 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.6418, + "step": 1078 + }, + { + "epoch": 0.76, + "learning_rate": 0.0001, + "loss": 1.7638, + "step": 1079 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.8062, + "step": 1080 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.9147, + "step": 1081 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.7236, + "step": 1082 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 2.153, + "step": 1083 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 2.1311, + "step": 1084 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 2.1591, + "step": 1085 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.8319, + "step": 1086 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 2.0848, + "step": 1087 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.8143, + "step": 1088 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.775, + "step": 1089 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.9926, + "step": 1090 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.685, + "step": 1091 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 2.0678, + "step": 1092 + }, + { + "epoch": 0.77, + "learning_rate": 0.0001, + "loss": 1.8919, + "step": 1093 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.6619, + "step": 1094 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.9407, + "step": 1095 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 2.1989, + "step": 1096 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.7825, + "step": 1097 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.7172, + "step": 1098 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 2.2864, + "step": 1099 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.5073, + "step": 1100 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.8081, + "step": 1101 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.9721, + "step": 1102 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.8955, + "step": 1103 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.9065, + "step": 1104 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 2.0126, + "step": 1105 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 2.1909, + "step": 1106 + }, + { + "epoch": 0.78, + "learning_rate": 0.0001, + "loss": 1.871, + "step": 1107 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.8791, + "step": 1108 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.8209, + "step": 1109 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.5004, + "step": 1110 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.7917, + "step": 1111 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.6135, + "step": 1112 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.9156, + "step": 1113 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 2.0848, + "step": 1114 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.8023, + "step": 1115 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 2.0328, + "step": 1116 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.8428, + "step": 1117 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.9741, + "step": 1118 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.7626, + "step": 1119 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.5649, + "step": 1120 + }, + { + "epoch": 0.79, + "learning_rate": 0.0001, + "loss": 1.7526, + "step": 1121 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 2.0545, + "step": 1122 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 2.2046, + "step": 1123 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.636, + "step": 1124 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 2.1327, + "step": 1125 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.3032, + "step": 1126 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.6396, + "step": 1127 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.7123, + "step": 1128 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.9677, + "step": 1129 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.7477, + "step": 1130 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.7764, + "step": 1131 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.9469, + "step": 1132 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.9805, + "step": 1133 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 1.8433, + "step": 1134 + }, + { + "epoch": 0.8, + "learning_rate": 0.0001, + "loss": 2.1573, + "step": 1135 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.7865, + "step": 1136 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.6911, + "step": 1137 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.9271, + "step": 1138 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.8844, + "step": 1139 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 2.1905, + "step": 1140 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.8506, + "step": 1141 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 2.2661, + "step": 1142 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.7232, + "step": 1143 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.9991, + "step": 1144 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.5619, + "step": 1145 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.974, + "step": 1146 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 2.0721, + "step": 1147 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.7806, + "step": 1148 + }, + { + "epoch": 0.81, + "learning_rate": 0.0001, + "loss": 1.9031, + "step": 1149 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.9436, + "step": 1150 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.722, + "step": 1151 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 2.0907, + "step": 1152 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.9344, + "step": 1153 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.8904, + "step": 1154 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 2.1853, + "step": 1155 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.477, + "step": 1156 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 2.0179, + "step": 1157 + }, + { + "epoch": 0.82, + "eval_loss": 1.9268152713775635, + "eval_runtime": 64.1513, + "eval_samples_per_second": 0.234, + "eval_steps_per_second": 0.234, + "step": 1157 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.6802, + "step": 1158 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.948, + "step": 1159 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.8921, + "step": 1160 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.709, + "step": 1161 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.8763, + "step": 1162 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.601, + "step": 1163 + }, + { + "epoch": 0.82, + "learning_rate": 0.0001, + "loss": 1.7018, + "step": 1164 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 2.0424, + "step": 1165 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.8201, + "step": 1166 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.7211, + "step": 1167 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.8988, + "step": 1168 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.6791, + "step": 1169 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.8997, + "step": 1170 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.4643, + "step": 1171 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.6351, + "step": 1172 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.711, + "step": 1173 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.7918, + "step": 1174 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.6947, + "step": 1175 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 1.4431, + "step": 1176 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 2.1208, + "step": 1177 + }, + { + "epoch": 0.83, + "learning_rate": 0.0001, + "loss": 2.2914, + "step": 1178 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 2.1222, + "step": 1179 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.9073, + "step": 1180 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.7412, + "step": 1181 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.6939, + "step": 1182 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 2.9741, + "step": 1183 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.5645, + "step": 1184 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.8372, + "step": 1185 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 2.1551, + "step": 1186 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 2.2469, + "step": 1187 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.9648, + "step": 1188 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.7695, + "step": 1189 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.6717, + "step": 1190 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.7777, + "step": 1191 + }, + { + "epoch": 0.84, + "learning_rate": 0.0001, + "loss": 1.904, + "step": 1192 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.9375, + "step": 1193 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.7475, + "step": 1194 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 2.176, + "step": 1195 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.7629, + "step": 1196 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.6129, + "step": 1197 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 2.2184, + "step": 1198 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.8761, + "step": 1199 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.7564, + "step": 1200 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.7802, + "step": 1201 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 2.2242, + "step": 1202 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.9903, + "step": 1203 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 2.2617, + "step": 1204 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.8889, + "step": 1205 + }, + { + "epoch": 0.85, + "learning_rate": 0.0001, + "loss": 1.7064, + "step": 1206 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.5691, + "step": 1207 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 2.046, + "step": 1208 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.7292, + "step": 1209 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.8063, + "step": 1210 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.8968, + "step": 1211 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.89, + "step": 1212 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 2.0841, + "step": 1213 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.8561, + "step": 1214 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.7414, + "step": 1215 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.8544, + "step": 1216 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 1.9211, + "step": 1217 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 2.2022, + "step": 1218 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 2.1967, + "step": 1219 + }, + { + "epoch": 0.86, + "learning_rate": 0.0001, + "loss": 2.1005, + "step": 1220 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.6813, + "step": 1221 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.8846, + "step": 1222 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.7852, + "step": 1223 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 2.1345, + "step": 1224 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.9738, + "step": 1225 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 2.2913, + "step": 1226 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.916, + "step": 1227 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 2.0093, + "step": 1228 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.8367, + "step": 1229 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.7715, + "step": 1230 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.509, + "step": 1231 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.8631, + "step": 1232 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.476, + "step": 1233 + }, + { + "epoch": 0.87, + "learning_rate": 0.0001, + "loss": 1.9756, + "step": 1234 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.246, + "step": 1235 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.7165, + "step": 1236 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.8429, + "step": 1237 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.8487, + "step": 1238 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.0161, + "step": 1239 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.0203, + "step": 1240 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.1596, + "step": 1241 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.8874, + "step": 1242 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.934, + "step": 1243 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.6817, + "step": 1244 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.7907, + "step": 1245 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.0975, + "step": 1246 + }, + { + "epoch": 0.88, + "eval_loss": 1.9242541790008545, + "eval_runtime": 63.3922, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1246 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 1.8325, + "step": 1247 + }, + { + "epoch": 0.88, + "learning_rate": 0.0001, + "loss": 2.1478, + "step": 1248 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.9718, + "step": 1249 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 2.0378, + "step": 1250 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 2.0409, + "step": 1251 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.9091, + "step": 1252 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.9239, + "step": 1253 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 2.1454, + "step": 1254 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.9088, + "step": 1255 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.6272, + "step": 1256 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.8373, + "step": 1257 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.9723, + "step": 1258 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 2.0182, + "step": 1259 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.8097, + "step": 1260 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.7913, + "step": 1261 + }, + { + "epoch": 0.89, + "learning_rate": 0.0001, + "loss": 1.3804, + "step": 1262 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.5729, + "step": 1263 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.8805, + "step": 1264 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.873, + "step": 1265 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 2.1149, + "step": 1266 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.7163, + "step": 1267 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.92, + "step": 1268 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 2.0008, + "step": 1269 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.9228, + "step": 1270 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.8849, + "step": 1271 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.9436, + "step": 1272 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.6194, + "step": 1273 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.8257, + "step": 1274 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.6582, + "step": 1275 + }, + { + "epoch": 0.9, + "learning_rate": 0.0001, + "loss": 1.7683, + "step": 1276 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.7602, + "step": 1277 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.4264, + "step": 1278 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.9449, + "step": 1279 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.6017, + "step": 1280 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.9083, + "step": 1281 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 2.0122, + "step": 1282 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 2.1099, + "step": 1283 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.8277, + "step": 1284 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.5713, + "step": 1285 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 2.352, + "step": 1286 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.698, + "step": 1287 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 2.0338, + "step": 1288 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.9852, + "step": 1289 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.7762, + "step": 1290 + }, + { + "epoch": 0.91, + "learning_rate": 0.0001, + "loss": 1.6687, + "step": 1291 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.0932, + "step": 1292 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.9123, + "step": 1293 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.3361, + "step": 1294 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.2811, + "step": 1295 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.887, + "step": 1296 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.7575, + "step": 1297 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.7125, + "step": 1298 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.168, + "step": 1299 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.697, + "step": 1300 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.5156, + "step": 1301 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.104, + "step": 1302 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.8325, + "step": 1303 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 2.1019, + "step": 1304 + }, + { + "epoch": 0.92, + "learning_rate": 0.0001, + "loss": 1.8462, + "step": 1305 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.9057, + "step": 1306 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.9698, + "step": 1307 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.9946, + "step": 1308 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.6985, + "step": 1309 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.9255, + "step": 1310 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.5848, + "step": 1311 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.6719, + "step": 1312 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.6498, + "step": 1313 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 2.1591, + "step": 1314 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.5006, + "step": 1315 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 2.2985, + "step": 1316 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.8674, + "step": 1317 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 2.2023, + "step": 1318 + }, + { + "epoch": 0.93, + "learning_rate": 0.0001, + "loss": 1.7105, + "step": 1319 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 2.1111, + "step": 1320 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.5925, + "step": 1321 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.7701, + "step": 1322 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.8487, + "step": 1323 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.8425, + "step": 1324 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 2.1951, + "step": 1325 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.2588, + "step": 1326 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.9891, + "step": 1327 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 2.0407, + "step": 1328 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.6058, + "step": 1329 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.8978, + "step": 1330 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 2.2373, + "step": 1331 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.985, + "step": 1332 + }, + { + "epoch": 0.94, + "learning_rate": 0.0001, + "loss": 1.707, + "step": 1333 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.6868, + "step": 1334 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.7191, + "step": 1335 + }, + { + "epoch": 0.95, + "eval_loss": 1.9246184825897217, + "eval_runtime": 63.3696, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1335 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 2.1398, + "step": 1336 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.6078, + "step": 1337 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 2.3569, + "step": 1338 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.9199, + "step": 1339 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 2.0135, + "step": 1340 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.9322, + "step": 1341 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.6794, + "step": 1342 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.7662, + "step": 1343 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 2.181, + "step": 1344 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.535, + "step": 1345 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 1.8235, + "step": 1346 + }, + { + "epoch": 0.95, + "learning_rate": 0.0001, + "loss": 2.2163, + "step": 1347 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.9156, + "step": 1348 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.917, + "step": 1349 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.5048, + "step": 1350 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.766, + "step": 1351 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8382, + "step": 1352 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8445, + "step": 1353 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 2.0187, + "step": 1354 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8424, + "step": 1355 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.899, + "step": 1356 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8172, + "step": 1357 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 2.0487, + "step": 1358 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8386, + "step": 1359 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.8118, + "step": 1360 + }, + { + "epoch": 0.96, + "learning_rate": 0.0001, + "loss": 1.9987, + "step": 1361 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 2.0707, + "step": 1362 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.732, + "step": 1363 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.9519, + "step": 1364 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.8025, + "step": 1365 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.7155, + "step": 1366 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.858, + "step": 1367 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.6507, + "step": 1368 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.7709, + "step": 1369 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.7205, + "step": 1370 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.661, + "step": 1371 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.8501, + "step": 1372 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.6316, + "step": 1373 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.6305, + "step": 1374 + }, + { + "epoch": 0.97, + "learning_rate": 0.0001, + "loss": 1.9647, + "step": 1375 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.9386, + "step": 1376 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.9541, + "step": 1377 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.6228, + "step": 1378 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.9268, + "step": 1379 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.9813, + "step": 1380 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.6918, + "step": 1381 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 2.0742, + "step": 1382 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.6277, + "step": 1383 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.465, + "step": 1384 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.9921, + "step": 1385 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.7742, + "step": 1386 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.8034, + "step": 1387 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 1.6988, + "step": 1388 + }, + { + "epoch": 0.98, + "learning_rate": 0.0001, + "loss": 2.1144, + "step": 1389 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.7633, + "step": 1390 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.9004, + "step": 1391 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.5653, + "step": 1392 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.5337, + "step": 1393 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 2.2658, + "step": 1394 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.6347, + "step": 1395 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.8069, + "step": 1396 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.557, + "step": 1397 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.6775, + "step": 1398 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.8611, + "step": 1399 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 2.1561, + "step": 1400 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.7615, + "step": 1401 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.7642, + "step": 1402 + }, + { + "epoch": 0.99, + "learning_rate": 0.0001, + "loss": 1.6408, + "step": 1403 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.6142, + "step": 1404 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.9238, + "step": 1405 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.432, + "step": 1406 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.6958, + "step": 1407 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.7486, + "step": 1408 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 2.1143, + "step": 1409 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.688, + "step": 1410 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.6025, + "step": 1411 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.6945, + "step": 1412 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.7512, + "step": 1413 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.6596, + "step": 1414 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.5227, + "step": 1415 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.7139, + "step": 1416 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.322, + "step": 1417 + }, + { + "epoch": 1.0, + "learning_rate": 0.0001, + "loss": 1.5952, + "step": 1418 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.69, + "step": 1419 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.6942, + "step": 1420 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.7167, + "step": 1421 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.8557, + "step": 1422 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.6262, + "step": 1423 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.6579, + "step": 1424 + }, + { + "epoch": 1.01, + "eval_loss": 1.9249701499938965, + "eval_runtime": 63.6027, + "eval_samples_per_second": 0.236, + "eval_steps_per_second": 0.236, + "step": 1424 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.6662, + "step": 1425 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.7813, + "step": 1426 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.7495, + "step": 1427 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.4632, + "step": 1428 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.4123, + "step": 1429 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.6351, + "step": 1430 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.7105, + "step": 1431 + }, + { + "epoch": 1.01, + "learning_rate": 0.0001, + "loss": 1.4796, + "step": 1432 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.8745, + "step": 1433 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 2.0024, + "step": 1434 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.872, + "step": 1435 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.8067, + "step": 1436 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.6025, + "step": 1437 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 2.0765, + "step": 1438 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 2.1456, + "step": 1439 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.8758, + "step": 1440 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.7925, + "step": 1441 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.7886, + "step": 1442 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.5564, + "step": 1443 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 2.0534, + "step": 1444 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 1.5871, + "step": 1445 + }, + { + "epoch": 1.02, + "learning_rate": 0.0001, + "loss": 2.1777, + "step": 1446 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.9262, + "step": 1447 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 2.2582, + "step": 1448 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.3886, + "step": 1449 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.4893, + "step": 1450 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 2.0009, + "step": 1451 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.8651, + "step": 1452 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.7718, + "step": 1453 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.8694, + "step": 1454 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.6745, + "step": 1455 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.6261, + "step": 1456 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.6007, + "step": 1457 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.5633, + "step": 1458 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.7258, + "step": 1459 + }, + { + "epoch": 1.03, + "learning_rate": 0.0001, + "loss": 1.9641, + "step": 1460 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.4323, + "step": 1461 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.736, + "step": 1462 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.8733, + "step": 1463 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.4102, + "step": 1464 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.7476, + "step": 1465 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.7776, + "step": 1466 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.1355, + "step": 1467 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.9294, + "step": 1468 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.419, + "step": 1469 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.5387, + "step": 1470 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.6258, + "step": 1471 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.6755, + "step": 1472 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.6637, + "step": 1473 + }, + { + "epoch": 1.04, + "learning_rate": 0.0001, + "loss": 1.9036, + "step": 1474 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.7545, + "step": 1475 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.661, + "step": 1476 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.8836, + "step": 1477 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.6304, + "step": 1478 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.5102, + "step": 1479 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.7512, + "step": 1480 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 2.0653, + "step": 1481 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.7533, + "step": 1482 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.9943, + "step": 1483 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.7757, + "step": 1484 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.5386, + "step": 1485 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.6653, + "step": 1486 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.701, + "step": 1487 + }, + { + "epoch": 1.05, + "learning_rate": 0.0001, + "loss": 1.8606, + "step": 1488 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.9038, + "step": 1489 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.6232, + "step": 1490 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.9431, + "step": 1491 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 2.0489, + "step": 1492 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.3047, + "step": 1493 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.2695, + "step": 1494 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.5749, + "step": 1495 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.6473, + "step": 1496 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.4843, + "step": 1497 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.6731, + "step": 1498 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.814, + "step": 1499 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.4837, + "step": 1500 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.6458, + "step": 1501 + }, + { + "epoch": 1.06, + "learning_rate": 0.0001, + "loss": 1.7321, + "step": 1502 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.8786, + "step": 1503 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.6265, + "step": 1504 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.786, + "step": 1505 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 2.0774, + "step": 1506 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 2.0586, + "step": 1507 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.8174, + "step": 1508 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.8266, + "step": 1509 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.7024, + "step": 1510 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.7298, + "step": 1511 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.5261, + "step": 1512 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.3073, + "step": 1513 + }, + { + "epoch": 1.07, + "eval_loss": 1.945878267288208, + "eval_runtime": 63.3819, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1513 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.5695, + "step": 1514 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.3913, + "step": 1515 + }, + { + "epoch": 1.07, + "learning_rate": 0.0001, + "loss": 1.7174, + "step": 1516 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.3493, + "step": 1517 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 2.0249, + "step": 1518 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.758, + "step": 1519 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.6841, + "step": 1520 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.4539, + "step": 1521 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.5051, + "step": 1522 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.664, + "step": 1523 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.738, + "step": 1524 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.8014, + "step": 1525 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.3785, + "step": 1526 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.5086, + "step": 1527 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 0.9883, + "step": 1528 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.2551, + "step": 1529 + }, + { + "epoch": 1.08, + "learning_rate": 0.0001, + "loss": 1.5289, + "step": 1530 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.9074, + "step": 1531 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.6652, + "step": 1532 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.8049, + "step": 1533 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.8685, + "step": 1534 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.4461, + "step": 1535 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.4461, + "step": 1536 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.7549, + "step": 1537 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.7061, + "step": 1538 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 2.2212, + "step": 1539 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.9488, + "step": 1540 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.0722, + "step": 1541 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.6502, + "step": 1542 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.0228, + "step": 1543 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.7266, + "step": 1544 + }, + { + "epoch": 1.09, + "learning_rate": 0.0001, + "loss": 1.9718, + "step": 1545 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.9295, + "step": 1546 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.9487, + "step": 1547 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.7107, + "step": 1548 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.5999, + "step": 1549 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.634, + "step": 1550 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.9449, + "step": 1551 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.6744, + "step": 1552 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.5443, + "step": 1553 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.9725, + "step": 1554 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.6028, + "step": 1555 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.0176, + "step": 1556 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 2.079, + "step": 1557 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.9176, + "step": 1558 + }, + { + "epoch": 1.1, + "learning_rate": 0.0001, + "loss": 1.6218, + "step": 1559 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.7058, + "step": 1560 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.0993, + "step": 1561 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 2.0321, + "step": 1562 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.5849, + "step": 1563 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.1113, + "step": 1564 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.658, + "step": 1565 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 2.0374, + "step": 1566 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.8152, + "step": 1567 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.8465, + "step": 1568 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.6291, + "step": 1569 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.9207, + "step": 1570 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.6625, + "step": 1571 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.8167, + "step": 1572 + }, + { + "epoch": 1.11, + "learning_rate": 0.0001, + "loss": 1.9215, + "step": 1573 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.6872, + "step": 1574 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.6717, + "step": 1575 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.9117, + "step": 1576 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.8216, + "step": 1577 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.9405, + "step": 1578 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.8041, + "step": 1579 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.7379, + "step": 1580 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.5512, + "step": 1581 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.7232, + "step": 1582 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.4192, + "step": 1583 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.7359, + "step": 1584 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 2.0339, + "step": 1585 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.65, + "step": 1586 + }, + { + "epoch": 1.12, + "learning_rate": 0.0001, + "loss": 1.3965, + "step": 1587 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.77, + "step": 1588 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7265, + "step": 1589 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7378, + "step": 1590 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 2.0955, + "step": 1591 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7967, + "step": 1592 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7285, + "step": 1593 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 2.072, + "step": 1594 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.5848, + "step": 1595 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.9955, + "step": 1596 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7664, + "step": 1597 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.8978, + "step": 1598 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.7799, + "step": 1599 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 2.1265, + "step": 1600 + }, + { + "epoch": 1.13, + "learning_rate": 0.0001, + "loss": 1.8273, + "step": 1601 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 2.101, + "step": 1602 + }, + { + "epoch": 1.14, + "eval_loss": 1.9494655132293701, + "eval_runtime": 63.7161, + "eval_samples_per_second": 0.235, + "eval_steps_per_second": 0.235, + "step": 1602 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.3001, + "step": 1603 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.7177, + "step": 1604 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.4687, + "step": 1605 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.6364, + "step": 1606 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 2.0634, + "step": 1607 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.9044, + "step": 1608 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.9386, + "step": 1609 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.9726, + "step": 1610 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.9748, + "step": 1611 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.706, + "step": 1612 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 2.0128, + "step": 1613 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 1.6177, + "step": 1614 + }, + { + "epoch": 1.14, + "learning_rate": 0.0001, + "loss": 2.0811, + "step": 1615 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.1331, + "step": 1616 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.7143, + "step": 1617 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.7179, + "step": 1618 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 2.0058, + "step": 1619 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.9216, + "step": 1620 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 2.0152, + "step": 1621 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.5641, + "step": 1622 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.4389, + "step": 1623 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 2.0769, + "step": 1624 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.8027, + "step": 1625 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.5527, + "step": 1626 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.8551, + "step": 1627 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.369, + "step": 1628 + }, + { + "epoch": 1.15, + "learning_rate": 0.0001, + "loss": 1.6694, + "step": 1629 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.7313, + "step": 1630 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.7985, + "step": 1631 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.9329, + "step": 1632 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.4417, + "step": 1633 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.7297, + "step": 1634 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 2.2117, + "step": 1635 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.6652, + "step": 1636 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.6456, + "step": 1637 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.7847, + "step": 1638 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.715, + "step": 1639 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.7279, + "step": 1640 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.6032, + "step": 1641 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.638, + "step": 1642 + }, + { + "epoch": 1.16, + "learning_rate": 0.0001, + "loss": 1.5824, + "step": 1643 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.1817, + "step": 1644 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.808, + "step": 1645 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.661, + "step": 1646 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.9131, + "step": 1647 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.8737, + "step": 1648 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.7667, + "step": 1649 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.4645, + "step": 1650 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.6157, + "step": 1651 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.5624, + "step": 1652 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.7286, + "step": 1653 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.7261, + "step": 1654 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 2.1463, + "step": 1655 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.6054, + "step": 1656 + }, + { + "epoch": 1.17, + "learning_rate": 0.0001, + "loss": 1.6082, + "step": 1657 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.976, + "step": 1658 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.4098, + "step": 1659 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.7593, + "step": 1660 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.6741, + "step": 1661 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.6047, + "step": 1662 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 2.0147, + "step": 1663 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.4743, + "step": 1664 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 2.0332, + "step": 1665 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.7499, + "step": 1666 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.6595, + "step": 1667 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.5235, + "step": 1668 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.4865, + "step": 1669 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.0421, + "step": 1670 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.6683, + "step": 1671 + }, + { + "epoch": 1.18, + "learning_rate": 0.0001, + "loss": 1.8902, + "step": 1672 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.9754, + "step": 1673 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.7393, + "step": 1674 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.6147, + "step": 1675 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.7635, + "step": 1676 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.6425, + "step": 1677 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.8237, + "step": 1678 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.1755, + "step": 1679 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.8289, + "step": 1680 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.2977, + "step": 1681 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.7313, + "step": 1682 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.621, + "step": 1683 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.7129, + "step": 1684 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.7145, + "step": 1685 + }, + { + "epoch": 1.19, + "learning_rate": 0.0001, + "loss": 1.6115, + "step": 1686 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.6744, + "step": 1687 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.3673, + "step": 1688 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.2074, + "step": 1689 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.8567, + "step": 1690 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 2.1437, + "step": 1691 + }, + { + "epoch": 1.2, + "eval_loss": 1.9496909379959106, + "eval_runtime": 63.4305, + "eval_samples_per_second": 0.236, + "eval_steps_per_second": 0.236, + "step": 1691 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.2095, + "step": 1692 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 2.185, + "step": 1693 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.5063, + "step": 1694 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.7863, + "step": 1695 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.2752, + "step": 1696 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.8377, + "step": 1697 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.6214, + "step": 1698 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.6109, + "step": 1699 + }, + { + "epoch": 1.2, + "learning_rate": 0.0001, + "loss": 1.7731, + "step": 1700 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.6409, + "step": 1701 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 2.0841, + "step": 1702 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.2003, + "step": 1703 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 0.5053, + "step": 1704 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.704, + "step": 1705 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.235, + "step": 1706 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.8261, + "step": 1707 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.5029, + "step": 1708 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.8124, + "step": 1709 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.0043, + "step": 1710 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.5111, + "step": 1711 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.7786, + "step": 1712 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.8279, + "step": 1713 + }, + { + "epoch": 1.21, + "learning_rate": 0.0001, + "loss": 1.6733, + "step": 1714 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.7766, + "step": 1715 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.6897, + "step": 1716 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.6191, + "step": 1717 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.9135, + "step": 1718 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.4385, + "step": 1719 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.5635, + "step": 1720 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.5073, + "step": 1721 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.7283, + "step": 1722 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 2.0348, + "step": 1723 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.3186, + "step": 1724 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.6705, + "step": 1725 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.6546, + "step": 1726 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.7882, + "step": 1727 + }, + { + "epoch": 1.22, + "learning_rate": 0.0001, + "loss": 1.56, + "step": 1728 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.8134, + "step": 1729 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.8556, + "step": 1730 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.5644, + "step": 1731 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.8346, + "step": 1732 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.6463, + "step": 1733 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.7104, + "step": 1734 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.9973, + "step": 1735 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.3831, + "step": 1736 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.6848, + "step": 1737 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.6916, + "step": 1738 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.6643, + "step": 1739 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.9115, + "step": 1740 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.5245, + "step": 1741 + }, + { + "epoch": 1.23, + "learning_rate": 0.0001, + "loss": 1.6909, + "step": 1742 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 2.0958, + "step": 1743 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.726, + "step": 1744 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 2.0767, + "step": 1745 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 2.005, + "step": 1746 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.63, + "step": 1747 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.6437, + "step": 1748 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.1731, + "step": 1749 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.6128, + "step": 1750 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.7805, + "step": 1751 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.7622, + "step": 1752 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 2.159, + "step": 1753 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.2994, + "step": 1754 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.5899, + "step": 1755 + }, + { + "epoch": 1.24, + "learning_rate": 0.0001, + "loss": 1.8124, + "step": 1756 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.5785, + "step": 1757 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.8023, + "step": 1758 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.6181, + "step": 1759 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.8971, + "step": 1760 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.7568, + "step": 1761 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.1524, + "step": 1762 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.5943, + "step": 1763 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.6098, + "step": 1764 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 1765 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.8629, + "step": 1766 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.8022, + "step": 1767 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.3385, + "step": 1768 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.5391, + "step": 1769 + }, + { + "epoch": 1.25, + "learning_rate": 0.0001, + "loss": 1.9298, + "step": 1770 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.7262, + "step": 1771 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.9157, + "step": 1772 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.6035, + "step": 1773 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.4934, + "step": 1774 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.6645, + "step": 1775 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.9337, + "step": 1776 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.3919, + "step": 1777 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.8542, + "step": 1778 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.7419, + "step": 1779 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.7049, + "step": 1780 + }, + { + "epoch": 1.26, + "eval_loss": 1.9455102682113647, + "eval_runtime": 63.2982, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1780 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.4795, + "step": 1781 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.7615, + "step": 1782 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.2493, + "step": 1783 + }, + { + "epoch": 1.26, + "learning_rate": 0.0001, + "loss": 1.6768, + "step": 1784 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.3559, + "step": 1785 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.5953, + "step": 1786 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.5762, + "step": 1787 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.6695, + "step": 1788 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.5976, + "step": 1789 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.3982, + "step": 1790 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.4088, + "step": 1791 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.6316, + "step": 1792 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.4846, + "step": 1793 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.8331, + "step": 1794 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.5196, + "step": 1795 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.7144, + "step": 1796 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.874, + "step": 1797 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.2909, + "step": 1798 + }, + { + "epoch": 1.27, + "learning_rate": 0.0001, + "loss": 1.9295, + "step": 1799 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.6554, + "step": 1800 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.756, + "step": 1801 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.5286, + "step": 1802 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.2475, + "step": 1803 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.3504, + "step": 1804 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.6761, + "step": 1805 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.7771, + "step": 1806 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.5513, + "step": 1807 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.8992, + "step": 1808 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.6001, + "step": 1809 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.0107, + "step": 1810 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.6516, + "step": 1811 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.4, + "step": 1812 + }, + { + "epoch": 1.28, + "learning_rate": 0.0001, + "loss": 1.9661, + "step": 1813 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.8378, + "step": 1814 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.4342, + "step": 1815 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.9191, + "step": 1816 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.4751, + "step": 1817 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.2309, + "step": 1818 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.8478, + "step": 1819 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 2.0416, + "step": 1820 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.7383, + "step": 1821 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.7763, + "step": 1822 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.7933, + "step": 1823 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 2.2124, + "step": 1824 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.7276, + "step": 1825 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.7122, + "step": 1826 + }, + { + "epoch": 1.29, + "learning_rate": 0.0001, + "loss": 1.8052, + "step": 1827 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.5244, + "step": 1828 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.4488, + "step": 1829 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.4439, + "step": 1830 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.3783, + "step": 1831 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.4073, + "step": 1832 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.7249, + "step": 1833 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.7947, + "step": 1834 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.9997, + "step": 1835 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.391, + "step": 1836 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.9791, + "step": 1837 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.9542, + "step": 1838 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.6866, + "step": 1839 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.4514, + "step": 1840 + }, + { + "epoch": 1.3, + "learning_rate": 0.0001, + "loss": 1.3061, + "step": 1841 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 2.2513, + "step": 1842 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.656, + "step": 1843 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 2.0115, + "step": 1844 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 2.0642, + "step": 1845 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.8812, + "step": 1846 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.9698, + "step": 1847 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.7101, + "step": 1848 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.6115, + "step": 1849 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.4584, + "step": 1850 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.6468, + "step": 1851 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.91, + "step": 1852 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.9338, + "step": 1853 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 1.7772, + "step": 1854 + }, + { + "epoch": 1.31, + "learning_rate": 0.0001, + "loss": 2.0151, + "step": 1855 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.6151, + "step": 1856 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.8803, + "step": 1857 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.8203, + "step": 1858 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.5396, + "step": 1859 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.1366, + "step": 1860 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.724, + "step": 1861 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 2.2467, + "step": 1862 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.5939, + "step": 1863 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.6961, + "step": 1864 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.3679, + "step": 1865 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.6815, + "step": 1866 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.7757, + "step": 1867 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.7754, + "step": 1868 + }, + { + "epoch": 1.32, + "learning_rate": 0.0001, + "loss": 1.7663, + "step": 1869 + }, + { + "epoch": 1.32, + "eval_loss": 1.9497803449630737, + "eval_runtime": 63.1891, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1869 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.8653, + "step": 1870 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.9247, + "step": 1871 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.5392, + "step": 1872 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.9172, + "step": 1873 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.4651, + "step": 1874 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.8057, + "step": 1875 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.5014, + "step": 1876 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.5483, + "step": 1877 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.2298, + "step": 1878 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.7679, + "step": 1879 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.6813, + "step": 1880 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 2.0494, + "step": 1881 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.1676, + "step": 1882 + }, + { + "epoch": 1.33, + "learning_rate": 0.0001, + "loss": 1.4746, + "step": 1883 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.6711, + "step": 1884 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 0.7828, + "step": 1885 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.7618, + "step": 1886 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.7174, + "step": 1887 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.6545, + "step": 1888 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.0469, + "step": 1889 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.6121, + "step": 1890 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.3722, + "step": 1891 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.6516, + "step": 1892 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.7147, + "step": 1893 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.5476, + "step": 1894 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.7388, + "step": 1895 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 1.4691, + "step": 1896 + }, + { + "epoch": 1.34, + "learning_rate": 0.0001, + "loss": 0.8619, + "step": 1897 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.8191, + "step": 1898 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.4955, + "step": 1899 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.6596, + "step": 1900 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.4618, + "step": 1901 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 2.01, + "step": 1902 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 2.2292, + "step": 1903 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.7525, + "step": 1904 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.9812, + "step": 1905 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.5675, + "step": 1906 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 2.0256, + "step": 1907 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.8676, + "step": 1908 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.5826, + "step": 1909 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.6933, + "step": 1910 + }, + { + "epoch": 1.35, + "learning_rate": 0.0001, + "loss": 1.7385, + "step": 1911 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.7074, + "step": 1912 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 2.1587, + "step": 1913 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.5677, + "step": 1914 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.4544, + "step": 1915 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.7986, + "step": 1916 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.6512, + "step": 1917 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.5898, + "step": 1918 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.8204, + "step": 1919 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.7865, + "step": 1920 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.8506, + "step": 1921 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.6889, + "step": 1922 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.9491, + "step": 1923 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.1764, + "step": 1924 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.7009, + "step": 1925 + }, + { + "epoch": 1.36, + "learning_rate": 0.0001, + "loss": 1.6492, + "step": 1926 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.6188, + "step": 1927 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.7481, + "step": 1928 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.2431, + "step": 1929 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.5776, + "step": 1930 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 2.0481, + "step": 1931 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.6605, + "step": 1932 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.6945, + "step": 1933 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.5777, + "step": 1934 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.7031, + "step": 1935 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.208, + "step": 1936 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.516, + "step": 1937 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 2.1038, + "step": 1938 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.6318, + "step": 1939 + }, + { + "epoch": 1.37, + "learning_rate": 0.0001, + "loss": 1.6635, + "step": 1940 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.8281, + "step": 1941 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.5875, + "step": 1942 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.9745, + "step": 1943 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.8608, + "step": 1944 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.6523, + "step": 1945 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.0997, + "step": 1946 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.6535, + "step": 1947 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.794, + "step": 1948 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.6849, + "step": 1949 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.6897, + "step": 1950 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.7836, + "step": 1951 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.6999, + "step": 1952 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.3648, + "step": 1953 + }, + { + "epoch": 1.38, + "learning_rate": 0.0001, + "loss": 1.902, + "step": 1954 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.6792, + "step": 1955 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.4847, + "step": 1956 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.2519, + "step": 1957 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 2.2991, + "step": 1958 + }, + { + "epoch": 1.39, + "eval_loss": 1.94557785987854, + "eval_runtime": 63.316, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 1958 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 0.4304, + "step": 1959 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.5284, + "step": 1960 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 2.0169, + "step": 1961 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.6981, + "step": 1962 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.9624, + "step": 1963 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.6356, + "step": 1964 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.8033, + "step": 1965 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.6553, + "step": 1966 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 0.8921, + "step": 1967 + }, + { + "epoch": 1.39, + "learning_rate": 0.0001, + "loss": 1.8046, + "step": 1968 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.6188, + "step": 1969 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.7137, + "step": 1970 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.4256, + "step": 1971 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.6248, + "step": 1972 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.7923, + "step": 1973 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.746, + "step": 1974 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 2.097, + "step": 1975 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.4685, + "step": 1976 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.5148, + "step": 1977 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.6608, + "step": 1978 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.7547, + "step": 1979 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.4941, + "step": 1980 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.4028, + "step": 1981 + }, + { + "epoch": 1.4, + "learning_rate": 0.0001, + "loss": 1.8809, + "step": 1982 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6349, + "step": 1983 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.5897, + "step": 1984 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 2.0326, + "step": 1985 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6788, + "step": 1986 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.7301, + "step": 1987 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6708, + "step": 1988 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.5554, + "step": 1989 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.8654, + "step": 1990 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.5192, + "step": 1991 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.8438, + "step": 1992 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.5421, + "step": 1993 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6985, + "step": 1994 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6697, + "step": 1995 + }, + { + "epoch": 1.41, + "learning_rate": 0.0001, + "loss": 1.6654, + "step": 1996 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.7798, + "step": 1997 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.2505, + "step": 1998 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.0637, + "step": 1999 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.9193, + "step": 2000 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.1348, + "step": 2001 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.1788, + "step": 2002 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.4732, + "step": 2003 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.7411, + "step": 2004 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.553, + "step": 2005 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.5364, + "step": 2006 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.5912, + "step": 2007 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.7875, + "step": 2008 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.3467, + "step": 2009 + }, + { + "epoch": 1.42, + "learning_rate": 0.0001, + "loss": 1.4521, + "step": 2010 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.8209, + "step": 2011 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.6564, + "step": 2012 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.3963, + "step": 2013 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.3426, + "step": 2014 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.6588, + "step": 2015 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.5932, + "step": 2016 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.5325, + "step": 2017 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 0.7287, + "step": 2018 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.2835, + "step": 2019 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.2958, + "step": 2020 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.6532, + "step": 2021 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.4859, + "step": 2022 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.6937, + "step": 2023 + }, + { + "epoch": 1.43, + "learning_rate": 0.0001, + "loss": 1.5219, + "step": 2024 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.98, + "step": 2025 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.3487, + "step": 2026 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 0.5937, + "step": 2027 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.6386, + "step": 2028 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.6734, + "step": 2029 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.5446, + "step": 2030 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.1022, + "step": 2031 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.5405, + "step": 2032 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.7499, + "step": 2033 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.8317, + "step": 2034 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 2.0198, + "step": 2035 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.4415, + "step": 2036 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 2.0274, + "step": 2037 + }, + { + "epoch": 1.44, + "learning_rate": 0.0001, + "loss": 1.6697, + "step": 2038 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.4828, + "step": 2039 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.2746, + "step": 2040 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.5675, + "step": 2041 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.5846, + "step": 2042 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.3824, + "step": 2043 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 0.9352, + "step": 2044 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.7914, + "step": 2045 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.8138, + "step": 2046 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 2.0033, + "step": 2047 + }, + { + "epoch": 1.45, + "eval_loss": 1.9548592567443848, + "eval_runtime": 63.3069, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2047 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 2.1568, + "step": 2048 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 2.233, + "step": 2049 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.4521, + "step": 2050 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.6895, + "step": 2051 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.9095, + "step": 2052 + }, + { + "epoch": 1.45, + "learning_rate": 0.0001, + "loss": 1.9172, + "step": 2053 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.6656, + "step": 2054 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.6938, + "step": 2055 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.957, + "step": 2056 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 2.0622, + "step": 2057 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.6446, + "step": 2058 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.5751, + "step": 2059 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 0.6129, + "step": 2060 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.9518, + "step": 2061 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.4289, + "step": 2062 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.9601, + "step": 2063 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.422, + "step": 2064 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.51, + "step": 2065 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.4685, + "step": 2066 + }, + { + "epoch": 1.46, + "learning_rate": 0.0001, + "loss": 1.7567, + "step": 2067 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.7331, + "step": 2068 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.4658, + "step": 2069 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.7173, + "step": 2070 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 2.0777, + "step": 2071 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.3091, + "step": 2072 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.8495, + "step": 2073 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.873, + "step": 2074 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.6774, + "step": 2075 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.6781, + "step": 2076 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.414, + "step": 2077 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.8324, + "step": 2078 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.9027, + "step": 2079 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.768, + "step": 2080 + }, + { + "epoch": 1.47, + "learning_rate": 0.0001, + "loss": 1.6498, + "step": 2081 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.5036, + "step": 2082 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.5631, + "step": 2083 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.6814, + "step": 2084 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.7344, + "step": 2085 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.3276, + "step": 2086 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.7072, + "step": 2087 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.3724, + "step": 2088 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 2.0051, + "step": 2089 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 2.0886, + "step": 2090 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.3221, + "step": 2091 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.6156, + "step": 2092 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.9769, + "step": 2093 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 2.1, + "step": 2094 + }, + { + "epoch": 1.48, + "learning_rate": 0.0001, + "loss": 1.6958, + "step": 2095 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.4951, + "step": 2096 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.8243, + "step": 2097 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.853, + "step": 2098 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.7474, + "step": 2099 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 2.2166, + "step": 2100 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.7209, + "step": 2101 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.9388, + "step": 2102 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 2.3084, + "step": 2103 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.9936, + "step": 2104 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.834, + "step": 2105 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.4571, + "step": 2106 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.7694, + "step": 2107 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.8069, + "step": 2108 + }, + { + "epoch": 1.49, + "learning_rate": 0.0001, + "loss": 1.6298, + "step": 2109 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.9137, + "step": 2110 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.7929, + "step": 2111 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.5167, + "step": 2112 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.7159, + "step": 2113 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.9249, + "step": 2114 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.8169, + "step": 2115 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.532, + "step": 2116 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.4262, + "step": 2117 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.7636, + "step": 2118 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.2482, + "step": 2119 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.843, + "step": 2120 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 2.0457, + "step": 2121 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 0.7014, + "step": 2122 + }, + { + "epoch": 1.5, + "learning_rate": 0.0001, + "loss": 1.7491, + "step": 2123 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.8619, + "step": 2124 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.6708, + "step": 2125 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.6336, + "step": 2126 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.8629, + "step": 2127 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.729, + "step": 2128 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.8063, + "step": 2129 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.5898, + "step": 2130 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.5987, + "step": 2131 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.6923, + "step": 2132 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.5842, + "step": 2133 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.6873, + "step": 2134 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.8391, + "step": 2135 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.4333, + "step": 2136 + }, + { + "epoch": 1.51, + "eval_loss": 1.9572677612304688, + "eval_runtime": 63.3227, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2136 + }, + { + "epoch": 1.51, + "learning_rate": 0.0001, + "loss": 1.5515, + "step": 2137 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.5277, + "step": 2138 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.6514, + "step": 2139 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.3987, + "step": 2140 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.3374, + "step": 2141 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.8335, + "step": 2142 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.6756, + "step": 2143 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 2.2042, + "step": 2144 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.7945, + "step": 2145 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.7943, + "step": 2146 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.5509, + "step": 2147 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.4139, + "step": 2148 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.4761, + "step": 2149 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.7191, + "step": 2150 + }, + { + "epoch": 1.52, + "learning_rate": 0.0001, + "loss": 1.4714, + "step": 2151 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.7201, + "step": 2152 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.7022, + "step": 2153 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.28, + "step": 2154 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.9461, + "step": 2155 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.5434, + "step": 2156 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.3597, + "step": 2157 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.3406, + "step": 2158 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.5633, + "step": 2159 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.7484, + "step": 2160 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.6305, + "step": 2161 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.9536, + "step": 2162 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.5475, + "step": 2163 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 1.6209, + "step": 2164 + }, + { + "epoch": 1.53, + "learning_rate": 0.0001, + "loss": 2.0189, + "step": 2165 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.6423, + "step": 2166 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.8053, + "step": 2167 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.7525, + "step": 2168 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.7533, + "step": 2169 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.9834, + "step": 2170 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.685, + "step": 2171 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.7198, + "step": 2172 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.9222, + "step": 2173 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.8482, + "step": 2174 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.715, + "step": 2175 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.648, + "step": 2176 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.5392, + "step": 2177 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.7144, + "step": 2178 + }, + { + "epoch": 1.54, + "learning_rate": 0.0001, + "loss": 1.7394, + "step": 2179 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.8028, + "step": 2180 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.8343, + "step": 2181 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.5511, + "step": 2182 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.7091, + "step": 2183 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.4151, + "step": 2184 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.8709, + "step": 2185 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 2.1461, + "step": 2186 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.6045, + "step": 2187 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.6907, + "step": 2188 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.7896, + "step": 2189 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.6306, + "step": 2190 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.8552, + "step": 2191 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.9654, + "step": 2192 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.7559, + "step": 2193 + }, + { + "epoch": 1.55, + "learning_rate": 0.0001, + "loss": 1.7947, + "step": 2194 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.6279, + "step": 2195 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.3537, + "step": 2196 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.5188, + "step": 2197 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.669, + "step": 2198 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.6248, + "step": 2199 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.7497, + "step": 2200 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 2.1234, + "step": 2201 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.7143, + "step": 2202 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.8488, + "step": 2203 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.5546, + "step": 2204 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.7054, + "step": 2205 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.9943, + "step": 2206 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 2.2176, + "step": 2207 + }, + { + "epoch": 1.56, + "learning_rate": 0.0001, + "loss": 1.9456, + "step": 2208 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.1183, + "step": 2209 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.7406, + "step": 2210 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.7848, + "step": 2211 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.523, + "step": 2212 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.3125, + "step": 2213 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.3959, + "step": 2214 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.5737, + "step": 2215 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.5742, + "step": 2216 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.5875, + "step": 2217 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.3907, + "step": 2218 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.4028, + "step": 2219 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 0.7931, + "step": 2220 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.8324, + "step": 2221 + }, + { + "epoch": 1.57, + "learning_rate": 0.0001, + "loss": 1.2867, + "step": 2222 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.7109, + "step": 2223 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.4247, + "step": 2224 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.7488, + "step": 2225 + }, + { + "epoch": 1.58, + "eval_loss": 1.9493993520736694, + "eval_runtime": 63.25, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2225 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.6847, + "step": 2226 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.7369, + "step": 2227 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.6423, + "step": 2228 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.8399, + "step": 2229 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.9058, + "step": 2230 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.3917, + "step": 2231 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.6074, + "step": 2232 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 2.0381, + "step": 2233 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.8444, + "step": 2234 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.3975, + "step": 2235 + }, + { + "epoch": 1.58, + "learning_rate": 0.0001, + "loss": 1.9342, + "step": 2236 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.7459, + "step": 2237 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.7877, + "step": 2238 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6361, + "step": 2239 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.8995, + "step": 2240 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6166, + "step": 2241 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6959, + "step": 2242 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.4736, + "step": 2243 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6249, + "step": 2244 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.7316, + "step": 2245 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6915, + "step": 2246 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.4745, + "step": 2247 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.6979, + "step": 2248 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 2.0348, + "step": 2249 + }, + { + "epoch": 1.59, + "learning_rate": 0.0001, + "loss": 1.4898, + "step": 2250 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.565, + "step": 2251 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.8307, + "step": 2252 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.7863, + "step": 2253 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.329, + "step": 2254 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 2.2293, + "step": 2255 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.884, + "step": 2256 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 2.3033, + "step": 2257 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.8332, + "step": 2258 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.1095, + "step": 2259 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.7517, + "step": 2260 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.3958, + "step": 2261 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 1.3808, + "step": 2262 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 0.4472, + "step": 2263 + }, + { + "epoch": 1.6, + "learning_rate": 0.0001, + "loss": 2.1132, + "step": 2264 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.5803, + "step": 2265 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.8338, + "step": 2266 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.7135, + "step": 2267 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.7059, + "step": 2268 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.5172, + "step": 2269 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 2.2147, + "step": 2270 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.829, + "step": 2271 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.559, + "step": 2272 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 2.006, + "step": 2273 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 2.0057, + "step": 2274 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.9951, + "step": 2275 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.5782, + "step": 2276 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.7614, + "step": 2277 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001, + "loss": 1.8505, + "step": 2278 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.9637, + "step": 2279 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.6992, + "step": 2280 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.3782, + "step": 2281 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.9716, + "step": 2282 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.8205, + "step": 2283 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.3438, + "step": 2284 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.6779, + "step": 2285 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.7075, + "step": 2286 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.4277, + "step": 2287 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.4122, + "step": 2288 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.9063, + "step": 2289 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.6385, + "step": 2290 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.9247, + "step": 2291 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001, + "loss": 1.8782, + "step": 2292 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.6627, + "step": 2293 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.4947, + "step": 2294 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 2.1013, + "step": 2295 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.4027, + "step": 2296 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.4889, + "step": 2297 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.7282, + "step": 2298 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.5374, + "step": 2299 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.8052, + "step": 2300 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.7062, + "step": 2301 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.8201, + "step": 2302 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.7471, + "step": 2303 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.797, + "step": 2304 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.6596, + "step": 2305 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001, + "loss": 1.5401, + "step": 2306 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.99, + "step": 2307 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.6496, + "step": 2308 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.6506, + "step": 2309 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 2.4952, + "step": 2310 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.9681, + "step": 2311 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.7594, + "step": 2312 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.3277, + "step": 2313 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.5412, + "step": 2314 + }, + { + "epoch": 1.64, + "eval_loss": 1.941941738128662, + "eval_runtime": 63.2762, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2314 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 2.0652, + "step": 2315 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.6917, + "step": 2316 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.5771, + "step": 2317 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.7891, + "step": 2318 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 2.133, + "step": 2319 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.8564, + "step": 2320 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001, + "loss": 1.5066, + "step": 2321 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.3182, + "step": 2322 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.9367, + "step": 2323 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.4495, + "step": 2324 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.2675, + "step": 2325 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.6023, + "step": 2326 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.8345, + "step": 2327 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.931, + "step": 2328 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.4096, + "step": 2329 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.8191, + "step": 2330 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 2.0991, + "step": 2331 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.6903, + "step": 2332 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.9351, + "step": 2333 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.4963, + "step": 2334 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001, + "loss": 1.7835, + "step": 2335 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.6446, + "step": 2336 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.7654, + "step": 2337 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.5187, + "step": 2338 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.8398, + "step": 2339 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.9034, + "step": 2340 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.7141, + "step": 2341 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.2103, + "step": 2342 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.3621, + "step": 2343 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.7412, + "step": 2344 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.5401, + "step": 2345 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.8617, + "step": 2346 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.8876, + "step": 2347 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.6503, + "step": 2348 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001, + "loss": 1.3997, + "step": 2349 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.8268, + "step": 2350 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.3809, + "step": 2351 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 2.0533, + "step": 2352 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.6516, + "step": 2353 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.6651, + "step": 2354 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.3901, + "step": 2355 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.5695, + "step": 2356 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.4726, + "step": 2357 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.51, + "step": 2358 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.8257, + "step": 2359 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.4308, + "step": 2360 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.65, + "step": 2361 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 0.4819, + "step": 2362 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001, + "loss": 1.8962, + "step": 2363 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.3952, + "step": 2364 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.7141, + "step": 2365 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.8731, + "step": 2366 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.8623, + "step": 2367 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 2.0278, + "step": 2368 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.5434, + "step": 2369 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.6841, + "step": 2370 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 2.2931, + "step": 2371 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.6228, + "step": 2372 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.9986, + "step": 2373 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.7351, + "step": 2374 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.6073, + "step": 2375 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.789, + "step": 2376 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001, + "loss": 1.7887, + "step": 2377 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.9432, + "step": 2378 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.3352, + "step": 2379 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.2658, + "step": 2380 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.7648, + "step": 2381 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.5526, + "step": 2382 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.1565, + "step": 2383 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.8835, + "step": 2384 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.52, + "step": 2385 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.8278, + "step": 2386 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.5035, + "step": 2387 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 2.0755, + "step": 2388 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.6354, + "step": 2389 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 1.3066, + "step": 2390 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001, + "loss": 0.7428, + "step": 2391 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.716, + "step": 2392 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.6708, + "step": 2393 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.9492, + "step": 2394 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.6931, + "step": 2395 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.7181, + "step": 2396 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.7919, + "step": 2397 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.2295, + "step": 2398 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.0446, + "step": 2399 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.5291, + "step": 2400 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.7485, + "step": 2401 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.7949, + "step": 2402 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.5192, + "step": 2403 + }, + { + "epoch": 1.7, + "eval_loss": 1.941814661026001, + "eval_runtime": 63.312, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2403 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.4747, + "step": 2404 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001, + "loss": 1.5267, + "step": 2405 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.7725, + "step": 2406 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 0.8154, + "step": 2407 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.8533, + "step": 2408 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.8081, + "step": 2409 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.9258, + "step": 2410 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.4583, + "step": 2411 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.8748, + "step": 2412 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.8387, + "step": 2413 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.5373, + "step": 2414 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.6066, + "step": 2415 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.6424, + "step": 2416 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.6698, + "step": 2417 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.6096, + "step": 2418 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001, + "loss": 1.8888, + "step": 2419 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.4126, + "step": 2420 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.5126, + "step": 2421 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.5429, + "step": 2422 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 2.1063, + "step": 2423 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.5776, + "step": 2424 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.883, + "step": 2425 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.7832, + "step": 2426 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 2.0501, + "step": 2427 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.6044, + "step": 2428 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.8412, + "step": 2429 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.5932, + "step": 2430 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 2.0127, + "step": 2431 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.6922, + "step": 2432 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001, + "loss": 1.8851, + "step": 2433 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.834, + "step": 2434 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 2.0284, + "step": 2435 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.4536, + "step": 2436 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.8322, + "step": 2437 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.785, + "step": 2438 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.8675, + "step": 2439 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 2.0374, + "step": 2440 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.5691, + "step": 2441 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 2.0808, + "step": 2442 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 2.0225, + "step": 2443 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 2.0647, + "step": 2444 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.8164, + "step": 2445 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.3948, + "step": 2446 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.5665, + "step": 2447 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001, + "loss": 1.6356, + "step": 2448 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.6156, + "step": 2449 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.6938, + "step": 2450 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.4307, + "step": 2451 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.8946, + "step": 2452 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.7617, + "step": 2453 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.4449, + "step": 2454 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.5896, + "step": 2455 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 2.0445, + "step": 2456 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.6394, + "step": 2457 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.4042, + "step": 2458 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.7783, + "step": 2459 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.5665, + "step": 2460 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.7404, + "step": 2461 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001, + "loss": 1.6734, + "step": 2462 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.449, + "step": 2463 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.7111, + "step": 2464 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.5936, + "step": 2465 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.6375, + "step": 2466 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.619, + "step": 2467 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.8224, + "step": 2468 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.5868, + "step": 2469 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.9109, + "step": 2470 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 0.9905, + "step": 2471 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.883, + "step": 2472 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.6497, + "step": 2473 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.7252, + "step": 2474 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.6498, + "step": 2475 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001, + "loss": 1.6029, + "step": 2476 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.8377, + "step": 2477 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.6171, + "step": 2478 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.5802, + "step": 2479 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.8093, + "step": 2480 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.5764, + "step": 2481 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.6686, + "step": 2482 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 2.239, + "step": 2483 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.7683, + "step": 2484 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.7532, + "step": 2485 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 2.2636, + "step": 2486 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.9181, + "step": 2487 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.7028, + "step": 2488 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 1.8716, + "step": 2489 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001, + "loss": 2.1601, + "step": 2490 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.6618, + "step": 2491 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.6578, + "step": 2492 + }, + { + "epoch": 1.77, + "eval_loss": 1.9152195453643799, + "eval_runtime": 63.2534, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2492 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.4873, + "step": 2493 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.4569, + "step": 2494 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.4759, + "step": 2495 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.9898, + "step": 2496 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.7562, + "step": 2497 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.2761, + "step": 2498 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.945, + "step": 2499 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.3639, + "step": 2500 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 2.1342, + "step": 2501 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.6777, + "step": 2502 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.8349, + "step": 2503 + }, + { + "epoch": 1.77, + "learning_rate": 0.0001, + "loss": 1.454, + "step": 2504 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.7057, + "step": 2505 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 0.5438, + "step": 2506 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.6234, + "step": 2507 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.8509, + "step": 2508 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.523, + "step": 2509 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.7016, + "step": 2510 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.5004, + "step": 2511 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.3284, + "step": 2512 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.6981, + "step": 2513 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.6419, + "step": 2514 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.5043, + "step": 2515 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.6367, + "step": 2516 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.9902, + "step": 2517 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001, + "loss": 1.5845, + "step": 2518 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6787, + "step": 2519 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.8606, + "step": 2520 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.9643, + "step": 2521 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 2.1181, + "step": 2522 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.4689, + "step": 2523 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6674, + "step": 2524 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.7858, + "step": 2525 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.8322, + "step": 2526 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6417, + "step": 2527 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6293, + "step": 2528 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6914, + "step": 2529 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.5508, + "step": 2530 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.6809, + "step": 2531 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001, + "loss": 1.1123, + "step": 2532 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.5997, + "step": 2533 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.9995, + "step": 2534 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.9685, + "step": 2535 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.7552, + "step": 2536 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.7751, + "step": 2537 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.9814, + "step": 2538 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.5761, + "step": 2539 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.5378, + "step": 2540 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.6212, + "step": 2541 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.591, + "step": 2542 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.9129, + "step": 2543 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 1.7009, + "step": 2544 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 2.2537, + "step": 2545 + }, + { + "epoch": 1.8, + "learning_rate": 0.0001, + "loss": 2.0506, + "step": 2546 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.1722, + "step": 2547 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.8024, + "step": 2548 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.7322, + "step": 2549 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 2.0994, + "step": 2550 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.4175, + "step": 2551 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.9683, + "step": 2552 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.6565, + "step": 2553 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.5251, + "step": 2554 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.5623, + "step": 2555 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.7258, + "step": 2556 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.7024, + "step": 2557 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.8889, + "step": 2558 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.6398, + "step": 2559 + }, + { + "epoch": 1.81, + "learning_rate": 0.0001, + "loss": 1.6301, + "step": 2560 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.3574, + "step": 2561 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 2.0855, + "step": 2562 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.5887, + "step": 2563 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.858, + "step": 2564 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.5578, + "step": 2565 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.5976, + "step": 2566 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.5237, + "step": 2567 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.7434, + "step": 2568 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.2861, + "step": 2569 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.1723, + "step": 2570 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.6484, + "step": 2571 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.8064, + "step": 2572 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.4881, + "step": 2573 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.47, + "step": 2574 + }, + { + "epoch": 1.82, + "learning_rate": 0.0001, + "loss": 1.8893, + "step": 2575 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.7198, + "step": 2576 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.693, + "step": 2577 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.7258, + "step": 2578 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.8139, + "step": 2579 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.6889, + "step": 2580 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.8038, + "step": 2581 + }, + { + "epoch": 1.83, + "eval_loss": 1.9265897274017334, + "eval_runtime": 63.3236, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2581 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.9457, + "step": 2582 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.7206, + "step": 2583 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 2.1033, + "step": 2584 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.7663, + "step": 2585 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.7475, + "step": 2586 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.8686, + "step": 2587 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.8261, + "step": 2588 + }, + { + "epoch": 1.83, + "learning_rate": 0.0001, + "loss": 1.6224, + "step": 2589 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 2.08, + "step": 2590 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.7612, + "step": 2591 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.6357, + "step": 2592 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.7953, + "step": 2593 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.2593, + "step": 2594 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.3511, + "step": 2595 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.8183, + "step": 2596 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.7128, + "step": 2597 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 2.1215, + "step": 2598 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.8563, + "step": 2599 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.9561, + "step": 2600 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.358, + "step": 2601 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.7819, + "step": 2602 + }, + { + "epoch": 1.84, + "learning_rate": 0.0001, + "loss": 1.0457, + "step": 2603 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.7559, + "step": 2604 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.9412, + "step": 2605 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.9919, + "step": 2606 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.4496, + "step": 2607 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.6087, + "step": 2608 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 2.254, + "step": 2609 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.8273, + "step": 2610 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.0315, + "step": 2611 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.3562, + "step": 2612 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.7184, + "step": 2613 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.7009, + "step": 2614 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.723, + "step": 2615 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.7788, + "step": 2616 + }, + { + "epoch": 1.85, + "learning_rate": 0.0001, + "loss": 1.8376, + "step": 2617 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.238, + "step": 2618 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.3556, + "step": 2619 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 2.0462, + "step": 2620 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.4285, + "step": 2621 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 2.0715, + "step": 2622 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.4119, + "step": 2623 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.5916, + "step": 2624 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.653, + "step": 2625 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.3955, + "step": 2626 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.8657, + "step": 2627 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.9601, + "step": 2628 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.8456, + "step": 2629 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.7702, + "step": 2630 + }, + { + "epoch": 1.86, + "learning_rate": 0.0001, + "loss": 1.1798, + "step": 2631 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.4066, + "step": 2632 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.8081, + "step": 2633 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.8842, + "step": 2634 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.4988, + "step": 2635 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.6347, + "step": 2636 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.446, + "step": 2637 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.828, + "step": 2638 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 2.2426, + "step": 2639 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.4354, + "step": 2640 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.3373, + "step": 2641 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 2.2794, + "step": 2642 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.501, + "step": 2643 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.7322, + "step": 2644 + }, + { + "epoch": 1.87, + "learning_rate": 0.0001, + "loss": 1.6483, + "step": 2645 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.2015, + "step": 2646 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.7921, + "step": 2647 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.208, + "step": 2648 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.9795, + "step": 2649 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.3015, + "step": 2650 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.6489, + "step": 2651 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.6695, + "step": 2652 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.68, + "step": 2653 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 2.0194, + "step": 2654 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.898, + "step": 2655 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.6211, + "step": 2656 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.7236, + "step": 2657 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.2475, + "step": 2658 + }, + { + "epoch": 1.88, + "learning_rate": 0.0001, + "loss": 1.5767, + "step": 2659 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.3151, + "step": 2660 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.1052, + "step": 2661 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.7223, + "step": 2662 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.4352, + "step": 2663 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.953, + "step": 2664 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 0.5043, + "step": 2665 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.4209, + "step": 2666 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.8378, + "step": 2667 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.8208, + "step": 2668 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.7256, + "step": 2669 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.4924, + "step": 2670 + }, + { + "epoch": 1.89, + "eval_loss": 1.9226765632629395, + "eval_runtime": 63.3101, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2670 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.6056, + "step": 2671 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.7528, + "step": 2672 + }, + { + "epoch": 1.89, + "learning_rate": 0.0001, + "loss": 1.584, + "step": 2673 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.8719, + "step": 2674 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.9973, + "step": 2675 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.2507, + "step": 2676 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.5755, + "step": 2677 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 2.0067, + "step": 2678 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.778, + "step": 2679 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.9394, + "step": 2680 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.7947, + "step": 2681 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.8206, + "step": 2682 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.3763, + "step": 2683 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.6195, + "step": 2684 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.9644, + "step": 2685 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.6878, + "step": 2686 + }, + { + "epoch": 1.9, + "learning_rate": 0.0001, + "loss": 1.5424, + "step": 2687 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.7385, + "step": 2688 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 0.7432, + "step": 2689 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.837, + "step": 2690 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.3839, + "step": 2691 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.984, + "step": 2692 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 0.3608, + "step": 2693 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.8002, + "step": 2694 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.897, + "step": 2695 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.6177, + "step": 2696 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.549, + "step": 2697 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.7092, + "step": 2698 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 1.3091, + "step": 2699 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 2.0898, + "step": 2700 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 2.1801, + "step": 2701 + }, + { + "epoch": 1.91, + "learning_rate": 0.0001, + "loss": 2.1392, + "step": 2702 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.7921, + "step": 2703 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 0.9768, + "step": 2704 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.7327, + "step": 2705 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.795, + "step": 2706 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 2.0648, + "step": 2707 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.6076, + "step": 2708 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.3042, + "step": 2709 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.4946, + "step": 2710 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.979, + "step": 2711 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.6103, + "step": 2712 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.3993, + "step": 2713 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.6409, + "step": 2714 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 1.5887, + "step": 2715 + }, + { + "epoch": 1.92, + "learning_rate": 0.0001, + "loss": 0.6077, + "step": 2716 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.8271, + "step": 2717 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.7189, + "step": 2718 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.8223, + "step": 2719 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.1416, + "step": 2720 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.5699, + "step": 2721 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.9349, + "step": 2722 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.5515, + "step": 2723 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.7212, + "step": 2724 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.854, + "step": 2725 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 2.0959, + "step": 2726 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.4586, + "step": 2727 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.7782, + "step": 2728 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.6242, + "step": 2729 + }, + { + "epoch": 1.93, + "learning_rate": 0.0001, + "loss": 1.9806, + "step": 2730 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.9451, + "step": 2731 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.8613, + "step": 2732 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 0.7664, + "step": 2733 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 2.0921, + "step": 2734 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 2.0128, + "step": 2735 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 2.0426, + "step": 2736 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.8469, + "step": 2737 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.5478, + "step": 2738 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.6441, + "step": 2739 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.7818, + "step": 2740 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.7672, + "step": 2741 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.6841, + "step": 2742 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.7536, + "step": 2743 + }, + { + "epoch": 1.94, + "learning_rate": 0.0001, + "loss": 1.3982, + "step": 2744 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.8058, + "step": 2745 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.9629, + "step": 2746 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.5468, + "step": 2747 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.6874, + "step": 2748 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.4725, + "step": 2749 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.6246, + "step": 2750 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.4377, + "step": 2751 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.9542, + "step": 2752 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.8339, + "step": 2753 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.7892, + "step": 2754 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.6087, + "step": 2755 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.8224, + "step": 2756 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.5155, + "step": 2757 + }, + { + "epoch": 1.95, + "learning_rate": 0.0001, + "loss": 1.6059, + "step": 2758 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.9394, + "step": 2759 + }, + { + "epoch": 1.96, + "eval_loss": 1.9242284297943115, + "eval_runtime": 63.2955, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2759 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.2392, + "step": 2760 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.4219, + "step": 2761 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.8267, + "step": 2762 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.6931, + "step": 2763 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.5292, + "step": 2764 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.5965, + "step": 2765 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.6908, + "step": 2766 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.3551, + "step": 2767 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 2.0456, + "step": 2768 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.6867, + "step": 2769 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 2.1226, + "step": 2770 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.8618, + "step": 2771 + }, + { + "epoch": 1.96, + "learning_rate": 0.0001, + "loss": 1.5721, + "step": 2772 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.943, + "step": 2773 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.5674, + "step": 2774 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.6498, + "step": 2775 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.618, + "step": 2776 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.7271, + "step": 2777 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 2.0565, + "step": 2778 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.7193, + "step": 2779 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.7028, + "step": 2780 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.7317, + "step": 2781 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.3021, + "step": 2782 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.9647, + "step": 2783 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.3934, + "step": 2784 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.5857, + "step": 2785 + }, + { + "epoch": 1.97, + "learning_rate": 0.0001, + "loss": 1.4456, + "step": 2786 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.732, + "step": 2787 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.7059, + "step": 2788 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.5927, + "step": 2789 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.474, + "step": 2790 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.9162, + "step": 2791 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.3629, + "step": 2792 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 2.0695, + "step": 2793 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.6624, + "step": 2794 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.6845, + "step": 2795 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.2693, + "step": 2796 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.4815, + "step": 2797 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 0.9657, + "step": 2798 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.4711, + "step": 2799 + }, + { + "epoch": 1.98, + "learning_rate": 0.0001, + "loss": 1.6022, + "step": 2800 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.3576, + "step": 2801 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 2.0143, + "step": 2802 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.6853, + "step": 2803 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.5396, + "step": 2804 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.1702, + "step": 2805 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.8778, + "step": 2806 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 2.0549, + "step": 2807 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.4965, + "step": 2808 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.9139, + "step": 2809 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.5924, + "step": 2810 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.767, + "step": 2811 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.707, + "step": 2812 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 1.8036, + "step": 2813 + }, + { + "epoch": 1.99, + "learning_rate": 0.0001, + "loss": 0.8704, + "step": 2814 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.981, + "step": 2815 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.9706, + "step": 2816 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.7082, + "step": 2817 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.8859, + "step": 2818 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.5537, + "step": 2819 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.7741, + "step": 2820 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.3702, + "step": 2821 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.6073, + "step": 2822 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.2448, + "step": 2823 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.635, + "step": 2824 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 0.5791, + "step": 2825 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.7933, + "step": 2826 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.219, + "step": 2827 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 1.5369, + "step": 2828 + }, + { + "epoch": 2.0, + "learning_rate": 0.0001, + "loss": 0.9028, + "step": 2829 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 0.822, + "step": 2830 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.3018, + "step": 2831 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 0.463, + "step": 2832 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.7014, + "step": 2833 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.3853, + "step": 2834 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.0673, + "step": 2835 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.3896, + "step": 2836 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.4485, + "step": 2837 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.6396, + "step": 2838 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.267, + "step": 2839 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 0.5682, + "step": 2840 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.1262, + "step": 2841 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.1669, + "step": 2842 + }, + { + "epoch": 2.01, + "learning_rate": 0.0001, + "loss": 1.1529, + "step": 2843 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.5141, + "step": 2844 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 0.6332, + "step": 2845 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.0783, + "step": 2846 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 0.7377, + "step": 2847 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.6548, + "step": 2848 + }, + { + "epoch": 2.02, + "eval_loss": 1.9688668251037598, + "eval_runtime": 63.0897, + "eval_samples_per_second": 0.238, + "eval_steps_per_second": 0.238, + "step": 2848 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.7965, + "step": 2849 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.3564, + "step": 2850 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.8707, + "step": 2851 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.3645, + "step": 2852 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 2.075, + "step": 2853 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.687, + "step": 2854 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.8485, + "step": 2855 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 0.9599, + "step": 2856 + }, + { + "epoch": 2.02, + "learning_rate": 0.0001, + "loss": 1.8548, + "step": 2857 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.5952, + "step": 2858 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.3366, + "step": 2859 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.6, + "step": 2860 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.3705, + "step": 2861 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.5681, + "step": 2862 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.7038, + "step": 2863 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.5863, + "step": 2864 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.8632, + "step": 2865 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 0.5035, + "step": 2866 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 0.6718, + "step": 2867 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.3039, + "step": 2868 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 2.0235, + "step": 2869 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.1147, + "step": 2870 + }, + { + "epoch": 2.03, + "learning_rate": 0.0001, + "loss": 1.0334, + "step": 2871 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.3147, + "step": 2872 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.5608, + "step": 2873 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.5192, + "step": 2874 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 0.9243, + "step": 2875 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.4827, + "step": 2876 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.3059, + "step": 2877 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.0104, + "step": 2878 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 0.3634, + "step": 2879 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.4002, + "step": 2880 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.3309, + "step": 2881 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.2439, + "step": 2882 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.0298, + "step": 2883 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 1.1379, + "step": 2884 + }, + { + "epoch": 2.04, + "learning_rate": 0.0001, + "loss": 0.5683, + "step": 2885 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.8931, + "step": 2886 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 0.8919, + "step": 2887 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.2447, + "step": 2888 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.8962, + "step": 2889 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.1161, + "step": 2890 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.449, + "step": 2891 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 0.6601, + "step": 2892 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.5903, + "step": 2893 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.5398, + "step": 2894 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 0.4101, + "step": 2895 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.1869, + "step": 2896 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.881, + "step": 2897 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.522, + "step": 2898 + }, + { + "epoch": 2.05, + "learning_rate": 0.0001, + "loss": 1.5101, + "step": 2899 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.3477, + "step": 2900 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.6562, + "step": 2901 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.5594, + "step": 2902 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.6611, + "step": 2903 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.2379, + "step": 2904 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.5454, + "step": 2905 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.963, + "step": 2906 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.3608, + "step": 2907 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.3523, + "step": 2908 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.8226, + "step": 2909 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.5616, + "step": 2910 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.3086, + "step": 2911 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 0.4654, + "step": 2912 + }, + { + "epoch": 2.06, + "learning_rate": 0.0001, + "loss": 1.3791, + "step": 2913 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 0.7251, + "step": 2914 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.5218, + "step": 2915 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.0877, + "step": 2916 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 0.9064, + "step": 2917 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.5342, + "step": 2918 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.5085, + "step": 2919 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.5776, + "step": 2920 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.2814, + "step": 2921 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 0.5607, + "step": 2922 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.1059, + "step": 2923 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.6847, + "step": 2924 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.1263, + "step": 2925 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.2434, + "step": 2926 + }, + { + "epoch": 2.07, + "learning_rate": 0.0001, + "loss": 1.5862, + "step": 2927 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.3876, + "step": 2928 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.1373, + "step": 2929 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.1165, + "step": 2930 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.5908, + "step": 2931 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.4916, + "step": 2932 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.6608, + "step": 2933 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.2057, + "step": 2934 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.5263, + "step": 2935 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 0.8576, + "step": 2936 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 0.8807, + "step": 2937 + }, + { + "epoch": 2.08, + "eval_loss": 1.975825548171997, + "eval_runtime": 63.1925, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 2937 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 0.4102, + "step": 2938 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.0603, + "step": 2939 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 0.7982, + "step": 2940 + }, + { + "epoch": 2.08, + "learning_rate": 0.0001, + "loss": 1.7413, + "step": 2941 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 0.8135, + "step": 2942 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.0232, + "step": 2943 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.6834, + "step": 2944 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.9067, + "step": 2945 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.3022, + "step": 2946 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 0.5756, + "step": 2947 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 0.2923, + "step": 2948 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.2609, + "step": 2949 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.5069, + "step": 2950 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.5851, + "step": 2951 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.5711, + "step": 2952 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.3841, + "step": 2953 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.1074, + "step": 2954 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 1.5035, + "step": 2955 + }, + { + "epoch": 2.09, + "learning_rate": 0.0001, + "loss": 0.1691, + "step": 2956 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.096, + "step": 2957 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.3059, + "step": 2958 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.5081, + "step": 2959 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.085, + "step": 2960 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.8378, + "step": 2961 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 0.2955, + "step": 2962 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.332, + "step": 2963 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 0.6217, + "step": 2964 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.0905, + "step": 2965 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 0.4529, + "step": 2966 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.5902, + "step": 2967 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.8744, + "step": 2968 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 0.3619, + "step": 2969 + }, + { + "epoch": 2.1, + "learning_rate": 0.0001, + "loss": 1.3977, + "step": 2970 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 0.4417, + "step": 2971 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.9822, + "step": 2972 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.0055, + "step": 2973 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.9314, + "step": 2974 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.6707, + "step": 2975 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 0.6705, + "step": 2976 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.5405, + "step": 2977 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.345, + "step": 2978 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.8347, + "step": 2979 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.5503, + "step": 2980 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.9026, + "step": 2981 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.8455, + "step": 2982 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 0.664, + "step": 2983 + }, + { + "epoch": 2.11, + "learning_rate": 0.0001, + "loss": 1.5278, + "step": 2984 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 0.408, + "step": 2985 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 0.9972, + "step": 2986 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.2713, + "step": 2987 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.6054, + "step": 2988 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 0.3705, + "step": 2989 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.2854, + "step": 2990 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.2701, + "step": 2991 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.9614, + "step": 2992 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.5666, + "step": 2993 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.6366, + "step": 2994 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 0.6076, + "step": 2995 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.1878, + "step": 2996 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.2469, + "step": 2997 + }, + { + "epoch": 2.12, + "learning_rate": 0.0001, + "loss": 1.5916, + "step": 2998 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.5048, + "step": 2999 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.4771, + "step": 3000 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 0.6642, + "step": 3001 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.2121, + "step": 3002 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.5255, + "step": 3003 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 0.6095, + "step": 3004 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.1559, + "step": 3005 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.5125, + "step": 3006 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.4957, + "step": 3007 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.0496, + "step": 3008 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.1834, + "step": 3009 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 0.8042, + "step": 3010 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.5617, + "step": 3011 + }, + { + "epoch": 2.13, + "learning_rate": 0.0001, + "loss": 1.4704, + "step": 3012 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.9066, + "step": 3013 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.6455, + "step": 3014 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.7287, + "step": 3015 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.8853, + "step": 3016 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.6917, + "step": 3017 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.9259, + "step": 3018 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.2029, + "step": 3019 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.5024, + "step": 3020 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.4719, + "step": 3021 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.2142, + "step": 3022 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.6582, + "step": 3023 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 0.2984, + "step": 3024 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.6371, + "step": 3025 + }, + { + "epoch": 2.14, + "learning_rate": 0.0001, + "loss": 1.6602, + "step": 3026 + }, + { + "epoch": 2.14, + "eval_loss": 1.9871209859848022, + "eval_runtime": 63.2825, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 3026 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.2999, + "step": 3027 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.3978, + "step": 3028 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 2.0221, + "step": 3029 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 0.553, + "step": 3030 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.3624, + "step": 3031 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 0.8701, + "step": 3032 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.8029, + "step": 3033 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.5268, + "step": 3034 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.7262, + "step": 3035 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.7248, + "step": 3036 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.6311, + "step": 3037 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.4855, + "step": 3038 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.9484, + "step": 3039 + }, + { + "epoch": 2.15, + "learning_rate": 0.0001, + "loss": 1.5585, + "step": 3040 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.3885, + "step": 3041 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.5515, + "step": 3042 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 0.9314, + "step": 3043 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.1305, + "step": 3044 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.3944, + "step": 3045 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.6975, + "step": 3046 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.7503, + "step": 3047 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.2851, + "step": 3048 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 2.1027, + "step": 3049 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.5026, + "step": 3050 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 2.1529, + "step": 3051 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.518, + "step": 3052 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.2531, + "step": 3053 + }, + { + "epoch": 2.16, + "learning_rate": 0.0001, + "loss": 1.3982, + "step": 3054 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.2495, + "step": 3055 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.4697, + "step": 3056 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 0.7477, + "step": 3057 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.0271, + "step": 3058 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.6957, + "step": 3059 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.4984, + "step": 3060 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.62, + "step": 3061 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 0.6401, + "step": 3062 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.4909, + "step": 3063 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.6193, + "step": 3064 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.1524, + "step": 3065 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.8591, + "step": 3066 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 0.2211, + "step": 3067 + }, + { + "epoch": 2.17, + "learning_rate": 0.0001, + "loss": 1.3749, + "step": 3068 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 0.5384, + "step": 3069 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.5359, + "step": 3070 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 0.9272, + "step": 3071 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.5419, + "step": 3072 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.4315, + "step": 3073 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 0.8407, + "step": 3074 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 0.3585, + "step": 3075 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.4306, + "step": 3076 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.346, + "step": 3077 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 0.9179, + "step": 3078 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.188, + "step": 3079 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.0594, + "step": 3080 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.0696, + "step": 3081 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 2.1657, + "step": 3082 + }, + { + "epoch": 2.18, + "learning_rate": 0.0001, + "loss": 1.6035, + "step": 3083 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.6132, + "step": 3084 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.3956, + "step": 3085 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.5768, + "step": 3086 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.1886, + "step": 3087 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.2633, + "step": 3088 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 0.6472, + "step": 3089 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.796, + "step": 3090 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.0636, + "step": 3091 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.5257, + "step": 3092 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.2776, + "step": 3093 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 1.5894, + "step": 3094 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 0.9282, + "step": 3095 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 0.3995, + "step": 3096 + }, + { + "epoch": 2.19, + "learning_rate": 0.0001, + "loss": 0.8528, + "step": 3097 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.1667, + "step": 3098 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.9096, + "step": 3099 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.4222, + "step": 3100 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.474, + "step": 3101 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.7124, + "step": 3102 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.407, + "step": 3103 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 0.9068, + "step": 3104 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.2428, + "step": 3105 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 2.191, + "step": 3106 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.6536, + "step": 3107 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.7713, + "step": 3108 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.3776, + "step": 3109 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 1.6124, + "step": 3110 + }, + { + "epoch": 2.2, + "learning_rate": 0.0001, + "loss": 0.7827, + "step": 3111 + }, + { + "epoch": 2.21, + "learning_rate": 0.0001, + "loss": 1.441, + "step": 3112 + }, + { + "epoch": 2.21, + "learning_rate": 0.0001, + "loss": 0.9992, + "step": 3113 + }, + { + "epoch": 2.21, + "learning_rate": 0.0001, + "loss": 1.5193, + "step": 3114 + }, + { + "epoch": 2.21, + "learning_rate": 0.0001, + "loss": 1.3712, + "step": 3115 + }, + { + "epoch": 2.21, + "eval_loss": 1.982452392578125, + "eval_runtime": 63.3933, + "eval_samples_per_second": 0.237, + "eval_steps_per_second": 0.237, + "step": 3115 + } + ], + "logging_steps": 1, + "max_steps": 4233, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 89, + "total_flos": 7.278872267335926e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}