End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +472 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: sft_mc_filtered
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# sft_mc_filtered
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 2.3079
|
| 21 |
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: sft_mc_filtered
|
|
|
|
| 16 |
|
| 17 |
# sft_mc_filtered
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the identity and the data_mc_filtered datasets.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 2.3079
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 2.3078930377960205,
|
| 4 |
+
"eval_runtime": 1.0328,
|
| 5 |
+
"eval_samples_per_second": 16.46,
|
| 6 |
+
"eval_steps_per_second": 2.905,
|
| 7 |
+
"total_flos": 37534574837760.0,
|
| 8 |
+
"train_loss": 0.26039194798469545,
|
| 9 |
+
"train_runtime": 1475.1742,
|
| 10 |
+
"train_samples_per_second": 5.403,
|
| 11 |
+
"train_steps_per_second": 0.339
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 2.3078930377960205,
|
| 4 |
+
"eval_runtime": 1.0328,
|
| 5 |
+
"eval_samples_per_second": 16.46,
|
| 6 |
+
"eval_steps_per_second": 2.905
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 37534574837760.0,
|
| 4 |
+
"train_loss": 0.26039194798469545,
|
| 5 |
+
"train_runtime": 1475.1742,
|
| 6 |
+
"train_samples_per_second": 5.403,
|
| 7 |
+
"train_steps_per_second": 0.339
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,472 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2,
|
| 13 |
+
"grad_norm": 32.215486418057836,
|
| 14 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 15 |
+
"loss": 1.507,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.4,
|
| 20 |
+
"grad_norm": 17.53584966854256,
|
| 21 |
+
"learning_rate": 4.000000000000001e-06,
|
| 22 |
+
"loss": 1.0158,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.6,
|
| 27 |
+
"grad_norm": 10.21777506024786,
|
| 28 |
+
"learning_rate": 6e-06,
|
| 29 |
+
"loss": 0.9148,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.8,
|
| 34 |
+
"grad_norm": 11.037533581774264,
|
| 35 |
+
"learning_rate": 8.000000000000001e-06,
|
| 36 |
+
"loss": 0.7659,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.0,
|
| 41 |
+
"grad_norm": 12.348125604423814,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 0.8433,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0,
|
| 48 |
+
"eval_loss": 1.0960203409194946,
|
| 49 |
+
"eval_runtime": 1.0335,
|
| 50 |
+
"eval_samples_per_second": 16.448,
|
| 51 |
+
"eval_steps_per_second": 2.903,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.2,
|
| 56 |
+
"grad_norm": 11.850091156248704,
|
| 57 |
+
"learning_rate": 9.987820251299121e-06,
|
| 58 |
+
"loss": 0.6045,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.4,
|
| 63 |
+
"grad_norm": 8.774567976123997,
|
| 64 |
+
"learning_rate": 9.951340343707852e-06,
|
| 65 |
+
"loss": 0.6436,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.6,
|
| 70 |
+
"grad_norm": 7.56423224860242,
|
| 71 |
+
"learning_rate": 9.890738003669029e-06,
|
| 72 |
+
"loss": 0.6194,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.8,
|
| 77 |
+
"grad_norm": 9.761428972107632,
|
| 78 |
+
"learning_rate": 9.806308479691595e-06,
|
| 79 |
+
"loss": 0.6762,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.0,
|
| 84 |
+
"grad_norm": 8.275718136211415,
|
| 85 |
+
"learning_rate": 9.698463103929542e-06,
|
| 86 |
+
"loss": 0.6237,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 2.0,
|
| 91 |
+
"eval_loss": 1.131184697151184,
|
| 92 |
+
"eval_runtime": 1.0288,
|
| 93 |
+
"eval_samples_per_second": 16.524,
|
| 94 |
+
"eval_steps_per_second": 2.916,
|
| 95 |
+
"step": 100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.2,
|
| 99 |
+
"grad_norm": 10.243082528915453,
|
| 100 |
+
"learning_rate": 9.567727288213005e-06,
|
| 101 |
+
"loss": 0.3926,
|
| 102 |
+
"step": 110
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.4,
|
| 106 |
+
"grad_norm": 6.692595611616738,
|
| 107 |
+
"learning_rate": 9.414737964294636e-06,
|
| 108 |
+
"loss": 0.386,
|
| 109 |
+
"step": 120
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.6,
|
| 113 |
+
"grad_norm": 7.767268297417072,
|
| 114 |
+
"learning_rate": 9.24024048078213e-06,
|
| 115 |
+
"loss": 0.4027,
|
| 116 |
+
"step": 130
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.8,
|
| 120 |
+
"grad_norm": 8.931326769204444,
|
| 121 |
+
"learning_rate": 9.045084971874738e-06,
|
| 122 |
+
"loss": 0.397,
|
| 123 |
+
"step": 140
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 3.0,
|
| 127 |
+
"grad_norm": 8.32608967253919,
|
| 128 |
+
"learning_rate": 8.83022221559489e-06,
|
| 129 |
+
"loss": 0.391,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 3.0,
|
| 134 |
+
"eval_loss": 1.2465544939041138,
|
| 135 |
+
"eval_runtime": 1.0305,
|
| 136 |
+
"eval_samples_per_second": 16.498,
|
| 137 |
+
"eval_steps_per_second": 2.911,
|
| 138 |
+
"step": 150
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.2,
|
| 142 |
+
"grad_norm": 7.073212488180378,
|
| 143 |
+
"learning_rate": 8.596699001693257e-06,
|
| 144 |
+
"loss": 0.2198,
|
| 145 |
+
"step": 160
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.4,
|
| 149 |
+
"grad_norm": 7.2442332680693875,
|
| 150 |
+
"learning_rate": 8.345653031794292e-06,
|
| 151 |
+
"loss": 0.2356,
|
| 152 |
+
"step": 170
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.6,
|
| 156 |
+
"grad_norm": 6.16850982087996,
|
| 157 |
+
"learning_rate": 8.078307376628292e-06,
|
| 158 |
+
"loss": 0.2416,
|
| 159 |
+
"step": 180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 3.8,
|
| 163 |
+
"grad_norm": 5.844877152864581,
|
| 164 |
+
"learning_rate": 7.795964517353734e-06,
|
| 165 |
+
"loss": 0.2504,
|
| 166 |
+
"step": 190
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 4.0,
|
| 170 |
+
"grad_norm": 6.614175675387944,
|
| 171 |
+
"learning_rate": 7.500000000000001e-06,
|
| 172 |
+
"loss": 0.2529,
|
| 173 |
+
"step": 200
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 4.0,
|
| 177 |
+
"eval_loss": 1.4002876281738281,
|
| 178 |
+
"eval_runtime": 1.0309,
|
| 179 |
+
"eval_samples_per_second": 16.49,
|
| 180 |
+
"eval_steps_per_second": 2.91,
|
| 181 |
+
"step": 200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 4.2,
|
| 185 |
+
"grad_norm": 4.846835703254587,
|
| 186 |
+
"learning_rate": 7.191855733945388e-06,
|
| 187 |
+
"loss": 0.1487,
|
| 188 |
+
"step": 210
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.4,
|
| 192 |
+
"grad_norm": 5.61752695846374,
|
| 193 |
+
"learning_rate": 6.873032967079562e-06,
|
| 194 |
+
"loss": 0.1411,
|
| 195 |
+
"step": 220
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.6,
|
| 199 |
+
"grad_norm": 7.574897231650108,
|
| 200 |
+
"learning_rate": 6.545084971874738e-06,
|
| 201 |
+
"loss": 0.1586,
|
| 202 |
+
"step": 230
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 4.8,
|
| 206 |
+
"grad_norm": 4.565228907443193,
|
| 207 |
+
"learning_rate": 6.209609477998339e-06,
|
| 208 |
+
"loss": 0.2038,
|
| 209 |
+
"step": 240
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 5.0,
|
| 213 |
+
"grad_norm": 4.885717790285905,
|
| 214 |
+
"learning_rate": 5.8682408883346535e-06,
|
| 215 |
+
"loss": 0.148,
|
| 216 |
+
"step": 250
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.0,
|
| 220 |
+
"eval_loss": 1.46064031124115,
|
| 221 |
+
"eval_runtime": 1.03,
|
| 222 |
+
"eval_samples_per_second": 16.505,
|
| 223 |
+
"eval_steps_per_second": 2.913,
|
| 224 |
+
"step": 250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.2,
|
| 228 |
+
"grad_norm": 7.108983510814077,
|
| 229 |
+
"learning_rate": 5.522642316338268e-06,
|
| 230 |
+
"loss": 0.0645,
|
| 231 |
+
"step": 260
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 5.4,
|
| 235 |
+
"grad_norm": 5.454792099444044,
|
| 236 |
+
"learning_rate": 5.174497483512506e-06,
|
| 237 |
+
"loss": 0.084,
|
| 238 |
+
"step": 270
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 5.6,
|
| 242 |
+
"grad_norm": 3.0763686972511572,
|
| 243 |
+
"learning_rate": 4.825502516487497e-06,
|
| 244 |
+
"loss": 0.0773,
|
| 245 |
+
"step": 280
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 5.8,
|
| 249 |
+
"grad_norm": 4.791359217511534,
|
| 250 |
+
"learning_rate": 4.477357683661734e-06,
|
| 251 |
+
"loss": 0.1003,
|
| 252 |
+
"step": 290
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.0,
|
| 256 |
+
"grad_norm": 3.746744930874563,
|
| 257 |
+
"learning_rate": 4.131759111665349e-06,
|
| 258 |
+
"loss": 0.0818,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 6.0,
|
| 263 |
+
"eval_loss": 1.5259884595870972,
|
| 264 |
+
"eval_runtime": 1.0326,
|
| 265 |
+
"eval_samples_per_second": 16.463,
|
| 266 |
+
"eval_steps_per_second": 2.905,
|
| 267 |
+
"step": 300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.2,
|
| 271 |
+
"grad_norm": 7.111673793819084,
|
| 272 |
+
"learning_rate": 3.790390522001662e-06,
|
| 273 |
+
"loss": 0.049,
|
| 274 |
+
"step": 310
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 6.4,
|
| 278 |
+
"grad_norm": 2.936988902922916,
|
| 279 |
+
"learning_rate": 3.4549150281252635e-06,
|
| 280 |
+
"loss": 0.0428,
|
| 281 |
+
"step": 320
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 6.6,
|
| 285 |
+
"grad_norm": 1.8852801169974247,
|
| 286 |
+
"learning_rate": 3.12696703292044e-06,
|
| 287 |
+
"loss": 0.0508,
|
| 288 |
+
"step": 330
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 6.8,
|
| 292 |
+
"grad_norm": 2.313238846786626,
|
| 293 |
+
"learning_rate": 2.8081442660546126e-06,
|
| 294 |
+
"loss": 0.0394,
|
| 295 |
+
"step": 340
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 7.0,
|
| 299 |
+
"grad_norm": 3.084443875902426,
|
| 300 |
+
"learning_rate": 2.5000000000000015e-06,
|
| 301 |
+
"loss": 0.0352,
|
| 302 |
+
"step": 350
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.0,
|
| 306 |
+
"eval_loss": 2.0276832580566406,
|
| 307 |
+
"eval_runtime": 1.0308,
|
| 308 |
+
"eval_samples_per_second": 16.492,
|
| 309 |
+
"eval_steps_per_second": 2.91,
|
| 310 |
+
"step": 350
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 7.2,
|
| 314 |
+
"grad_norm": 3.5661235534617943,
|
| 315 |
+
"learning_rate": 2.204035482646267e-06,
|
| 316 |
+
"loss": 0.0272,
|
| 317 |
+
"step": 360
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 7.4,
|
| 321 |
+
"grad_norm": 0.4070183405976748,
|
| 322 |
+
"learning_rate": 1.9216926233717087e-06,
|
| 323 |
+
"loss": 0.0257,
|
| 324 |
+
"step": 370
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 7.6,
|
| 328 |
+
"grad_norm": 2.699699532592376,
|
| 329 |
+
"learning_rate": 1.6543469682057105e-06,
|
| 330 |
+
"loss": 0.0259,
|
| 331 |
+
"step": 380
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 7.8,
|
| 335 |
+
"grad_norm": 1.0226717900156348,
|
| 336 |
+
"learning_rate": 1.4033009983067454e-06,
|
| 337 |
+
"loss": 0.0179,
|
| 338 |
+
"step": 390
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.0,
|
| 342 |
+
"grad_norm": 1.2731011142101376,
|
| 343 |
+
"learning_rate": 1.1697777844051105e-06,
|
| 344 |
+
"loss": 0.0197,
|
| 345 |
+
"step": 400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 8.0,
|
| 349 |
+
"eval_loss": 2.1508045196533203,
|
| 350 |
+
"eval_runtime": 1.0303,
|
| 351 |
+
"eval_samples_per_second": 16.5,
|
| 352 |
+
"eval_steps_per_second": 2.912,
|
| 353 |
+
"step": 400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 8.2,
|
| 357 |
+
"grad_norm": 0.9183553687922849,
|
| 358 |
+
"learning_rate": 9.549150281252633e-07,
|
| 359 |
+
"loss": 0.009,
|
| 360 |
+
"step": 410
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 8.4,
|
| 364 |
+
"grad_norm": 1.051244265434313,
|
| 365 |
+
"learning_rate": 7.597595192178702e-07,
|
| 366 |
+
"loss": 0.0074,
|
| 367 |
+
"step": 420
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 8.6,
|
| 371 |
+
"grad_norm": 0.8106413095614442,
|
| 372 |
+
"learning_rate": 5.852620357053651e-07,
|
| 373 |
+
"loss": 0.0115,
|
| 374 |
+
"step": 430
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.8,
|
| 378 |
+
"grad_norm": 1.4089090443500107,
|
| 379 |
+
"learning_rate": 4.322727117869951e-07,
|
| 380 |
+
"loss": 0.0142,
|
| 381 |
+
"step": 440
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 9.0,
|
| 385 |
+
"grad_norm": 0.2897517823285663,
|
| 386 |
+
"learning_rate": 3.015368960704584e-07,
|
| 387 |
+
"loss": 0.0129,
|
| 388 |
+
"step": 450
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 9.0,
|
| 392 |
+
"eval_loss": 2.2827517986297607,
|
| 393 |
+
"eval_runtime": 1.0366,
|
| 394 |
+
"eval_samples_per_second": 16.401,
|
| 395 |
+
"eval_steps_per_second": 2.894,
|
| 396 |
+
"step": 450
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 9.2,
|
| 400 |
+
"grad_norm": 0.581294302728187,
|
| 401 |
+
"learning_rate": 1.9369152030840553e-07,
|
| 402 |
+
"loss": 0.009,
|
| 403 |
+
"step": 460
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 9.4,
|
| 407 |
+
"grad_norm": 0.07224896290446714,
|
| 408 |
+
"learning_rate": 1.0926199633097156e-07,
|
| 409 |
+
"loss": 0.0071,
|
| 410 |
+
"step": 470
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 9.6,
|
| 414 |
+
"grad_norm": 0.9382264066330144,
|
| 415 |
+
"learning_rate": 4.865965629214819e-08,
|
| 416 |
+
"loss": 0.0058,
|
| 417 |
+
"step": 480
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 9.8,
|
| 421 |
+
"grad_norm": 0.7841805435055503,
|
| 422 |
+
"learning_rate": 1.2179748700879013e-08,
|
| 423 |
+
"loss": 0.0104,
|
| 424 |
+
"step": 490
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 10.0,
|
| 428 |
+
"grad_norm": 0.5841804580391363,
|
| 429 |
+
"learning_rate": 0.0,
|
| 430 |
+
"loss": 0.0066,
|
| 431 |
+
"step": 500
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 10.0,
|
| 435 |
+
"eval_loss": 2.3078930377960205,
|
| 436 |
+
"eval_runtime": 1.0306,
|
| 437 |
+
"eval_samples_per_second": 16.495,
|
| 438 |
+
"eval_steps_per_second": 2.911,
|
| 439 |
+
"step": 500
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 10.0,
|
| 443 |
+
"step": 500,
|
| 444 |
+
"total_flos": 37534574837760.0,
|
| 445 |
+
"train_loss": 0.26039194798469545,
|
| 446 |
+
"train_runtime": 1475.1742,
|
| 447 |
+
"train_samples_per_second": 5.403,
|
| 448 |
+
"train_steps_per_second": 0.339
|
| 449 |
+
}
|
| 450 |
+
],
|
| 451 |
+
"logging_steps": 10,
|
| 452 |
+
"max_steps": 500,
|
| 453 |
+
"num_input_tokens_seen": 0,
|
| 454 |
+
"num_train_epochs": 10,
|
| 455 |
+
"save_steps": 500,
|
| 456 |
+
"stateful_callbacks": {
|
| 457 |
+
"TrainerControl": {
|
| 458 |
+
"args": {
|
| 459 |
+
"should_epoch_stop": false,
|
| 460 |
+
"should_evaluate": false,
|
| 461 |
+
"should_log": false,
|
| 462 |
+
"should_save": true,
|
| 463 |
+
"should_training_stop": true
|
| 464 |
+
},
|
| 465 |
+
"attributes": {}
|
| 466 |
+
}
|
| 467 |
+
},
|
| 468 |
+
"total_flos": 37534574837760.0,
|
| 469 |
+
"train_batch_size": 2,
|
| 470 |
+
"trial_name": null,
|
| 471 |
+
"trial_params": null
|
| 472 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|