End of training
Browse files- README.md +3 -2
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +429 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ library_name: peft
|
|
| 4 |
license: llama3.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: Llama-3.1-8B-Instruct-SFT-800
|
|
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# Llama-3.1-8B-Instruct-SFT-800
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 4 |
license: llama3.1
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- lora
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: Llama-3.1-8B-Instruct-SFT-800
|
|
|
|
| 16 |
|
| 17 |
# Llama-3.1-8B-Instruct-SFT-800
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on the bct_non_cot_sft_800 dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.1235
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 0.12346285581588745,
|
| 4 |
+
"eval_runtime": 2.0598,
|
| 5 |
+
"eval_samples_per_second": 38.839,
|
| 6 |
+
"eval_steps_per_second": 19.419,
|
| 7 |
+
"total_flos": 4.027787393354957e+16,
|
| 8 |
+
"train_loss": 0.26821523129940034,
|
| 9 |
+
"train_runtime": 642.399,
|
| 10 |
+
"train_samples_per_second": 11.208,
|
| 11 |
+
"train_steps_per_second": 0.7
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 0.12346285581588745,
|
| 4 |
+
"eval_runtime": 2.0598,
|
| 5 |
+
"eval_samples_per_second": 38.839,
|
| 6 |
+
"eval_steps_per_second": 19.419
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 4.027787393354957e+16,
|
| 4 |
+
"train_loss": 0.26821523129940034,
|
| 5 |
+
"train_runtime": 642.399,
|
| 6 |
+
"train_samples_per_second": 11.208,
|
| 7 |
+
"train_steps_per_second": 0.7
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,429 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.12346285581588745,
|
| 3 |
+
"best_model_checkpoint": "saves/Llama-3.1-8B-Instruct/lora/sft-800/checkpoint-200",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 450,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2222222222222222,
|
| 13 |
+
"grad_norm": 7.030238628387451,
|
| 14 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 15 |
+
"loss": 1.6678,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.4444444444444444,
|
| 20 |
+
"grad_norm": 6.172452449798584,
|
| 21 |
+
"learning_rate": 2.1111111111111114e-06,
|
| 22 |
+
"loss": 1.5986,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.6666666666666666,
|
| 27 |
+
"grad_norm": 6.666406631469727,
|
| 28 |
+
"learning_rate": 3.2222222222222227e-06,
|
| 29 |
+
"loss": 1.549,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.8888888888888888,
|
| 34 |
+
"grad_norm": 4.0742573738098145,
|
| 35 |
+
"learning_rate": 4.333333333333334e-06,
|
| 36 |
+
"loss": 1.4175,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.1111111111111112,
|
| 41 |
+
"grad_norm": 3.02299427986145,
|
| 42 |
+
"learning_rate": 4.998796669702378e-06,
|
| 43 |
+
"loss": 1.1189,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.1111111111111112,
|
| 48 |
+
"eval_loss": 0.9741103053092957,
|
| 49 |
+
"eval_runtime": 2.0516,
|
| 50 |
+
"eval_samples_per_second": 38.995,
|
| 51 |
+
"eval_steps_per_second": 19.497,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.3333333333333333,
|
| 56 |
+
"grad_norm": 2.574375867843628,
|
| 57 |
+
"learning_rate": 4.9852725038674035e-06,
|
| 58 |
+
"loss": 0.8338,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.5555555555555556,
|
| 63 |
+
"grad_norm": 2.0675790309906006,
|
| 64 |
+
"learning_rate": 4.956801616766033e-06,
|
| 65 |
+
"loss": 0.5267,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.7777777777777777,
|
| 70 |
+
"grad_norm": 2.8309988975524902,
|
| 71 |
+
"learning_rate": 4.913555235663306e-06,
|
| 72 |
+
"loss": 0.4284,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 2.0,
|
| 77 |
+
"grad_norm": 2.8843226432800293,
|
| 78 |
+
"learning_rate": 4.855793449383731e-06,
|
| 79 |
+
"loss": 0.2176,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 2.2222222222222223,
|
| 84 |
+
"grad_norm": 1.2611788511276245,
|
| 85 |
+
"learning_rate": 4.783863644106502e-06,
|
| 86 |
+
"loss": 0.1565,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 2.2222222222222223,
|
| 91 |
+
"eval_loss": 0.15967388451099396,
|
| 92 |
+
"eval_runtime": 2.056,
|
| 93 |
+
"eval_samples_per_second": 38.91,
|
| 94 |
+
"eval_steps_per_second": 19.455,
|
| 95 |
+
"step": 100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.4444444444444446,
|
| 99 |
+
"grad_norm": 1.5975863933563232,
|
| 100 |
+
"learning_rate": 4.698198414144136e-06,
|
| 101 |
+
"loss": 0.117,
|
| 102 |
+
"step": 110
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.6666666666666665,
|
| 106 |
+
"grad_norm": 0.8590288162231445,
|
| 107 |
+
"learning_rate": 4.599312960269375e-06,
|
| 108 |
+
"loss": 0.0813,
|
| 109 |
+
"step": 120
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.888888888888889,
|
| 113 |
+
"grad_norm": 0.4642271101474762,
|
| 114 |
+
"learning_rate": 4.48780199123712e-06,
|
| 115 |
+
"loss": 0.1048,
|
| 116 |
+
"step": 130
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 3.111111111111111,
|
| 120 |
+
"grad_norm": 0.8627158403396606,
|
| 121 |
+
"learning_rate": 4.3643361471360045e-06,
|
| 122 |
+
"loss": 0.1059,
|
| 123 |
+
"step": 140
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 3.3333333333333335,
|
| 127 |
+
"grad_norm": 0.6250473260879517,
|
| 128 |
+
"learning_rate": 4.22965796607999e-06,
|
| 129 |
+
"loss": 0.0611,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 3.3333333333333335,
|
| 134 |
+
"eval_loss": 0.12997671961784363,
|
| 135 |
+
"eval_runtime": 2.0533,
|
| 136 |
+
"eval_samples_per_second": 38.961,
|
| 137 |
+
"eval_steps_per_second": 19.481,
|
| 138 |
+
"step": 150
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.5555555555555554,
|
| 142 |
+
"grad_norm": 1.1554416418075562,
|
| 143 |
+
"learning_rate": 4.084577418496775e-06,
|
| 144 |
+
"loss": 0.0917,
|
| 145 |
+
"step": 160
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.7777777777777777,
|
| 149 |
+
"grad_norm": 0.806210994720459,
|
| 150 |
+
"learning_rate": 3.929967035870346e-06,
|
| 151 |
+
"loss": 0.0933,
|
| 152 |
+
"step": 170
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 4.0,
|
| 156 |
+
"grad_norm": 1.2217227220535278,
|
| 157 |
+
"learning_rate": 3.7667566632339557e-06,
|
| 158 |
+
"loss": 0.0805,
|
| 159 |
+
"step": 180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 4.222222222222222,
|
| 163 |
+
"grad_norm": 0.717496931552887,
|
| 164 |
+
"learning_rate": 3.595927866972694e-06,
|
| 165 |
+
"loss": 0.0649,
|
| 166 |
+
"step": 190
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 4.444444444444445,
|
| 170 |
+
"grad_norm": 0.7854779362678528,
|
| 171 |
+
"learning_rate": 3.4185080315677454e-06,
|
| 172 |
+
"loss": 0.0975,
|
| 173 |
+
"step": 200
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 4.444444444444445,
|
| 177 |
+
"eval_loss": 0.12346285581588745,
|
| 178 |
+
"eval_runtime": 2.0576,
|
| 179 |
+
"eval_samples_per_second": 38.879,
|
| 180 |
+
"eval_steps_per_second": 19.44,
|
| 181 |
+
"step": 200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 4.666666666666667,
|
| 185 |
+
"grad_norm": 0.8865996599197388,
|
| 186 |
+
"learning_rate": 3.2355641807852377e-06,
|
| 187 |
+
"loss": 0.084,
|
| 188 |
+
"step": 210
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.888888888888889,
|
| 192 |
+
"grad_norm": 0.9991048574447632,
|
| 193 |
+
"learning_rate": 3.0481965604697582e-06,
|
| 194 |
+
"loss": 0.0647,
|
| 195 |
+
"step": 220
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 5.111111111111111,
|
| 199 |
+
"grad_norm": 0.482669472694397,
|
| 200 |
+
"learning_rate": 2.8575320215363855e-06,
|
| 201 |
+
"loss": 0.0602,
|
| 202 |
+
"step": 230
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 5.333333333333333,
|
| 206 |
+
"grad_norm": 2.210219383239746,
|
| 207 |
+
"learning_rate": 2.6647172429566963e-06,
|
| 208 |
+
"loss": 0.0848,
|
| 209 |
+
"step": 240
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 5.555555555555555,
|
| 213 |
+
"grad_norm": 0.8471083045005798,
|
| 214 |
+
"learning_rate": 2.470911835496508e-06,
|
| 215 |
+
"loss": 0.0664,
|
| 216 |
+
"step": 250
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.555555555555555,
|
| 220 |
+
"eval_loss": 0.12403944879770279,
|
| 221 |
+
"eval_runtime": 2.0543,
|
| 222 |
+
"eval_samples_per_second": 38.943,
|
| 223 |
+
"eval_steps_per_second": 19.472,
|
| 224 |
+
"step": 250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.777777777777778,
|
| 228 |
+
"grad_norm": 0.664802610874176,
|
| 229 |
+
"learning_rate": 2.2772813676802853e-06,
|
| 230 |
+
"loss": 0.093,
|
| 231 |
+
"step": 260
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 6.0,
|
| 235 |
+
"grad_norm": 0.7576791048049927,
|
| 236 |
+
"learning_rate": 2.0849903559248907e-06,
|
| 237 |
+
"loss": 0.0575,
|
| 238 |
+
"step": 270
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 6.222222222222222,
|
| 242 |
+
"grad_norm": 1.3942710161209106,
|
| 243 |
+
"learning_rate": 1.895195261000831e-06,
|
| 244 |
+
"loss": 0.0708,
|
| 245 |
+
"step": 280
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 6.444444444444445,
|
| 249 |
+
"grad_norm": 0.9209055304527283,
|
| 250 |
+
"learning_rate": 1.7090375329411085e-06,
|
| 251 |
+
"loss": 0.0592,
|
| 252 |
+
"step": 290
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.666666666666667,
|
| 256 |
+
"grad_norm": 1.9221105575561523,
|
| 257 |
+
"learning_rate": 1.5276367462264152e-06,
|
| 258 |
+
"loss": 0.0974,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 6.666666666666667,
|
| 263 |
+
"eval_loss": 0.12409935146570206,
|
| 264 |
+
"eval_runtime": 2.0546,
|
| 265 |
+
"eval_samples_per_second": 38.937,
|
| 266 |
+
"eval_steps_per_second": 19.468,
|
| 267 |
+
"step": 300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.888888888888889,
|
| 271 |
+
"grad_norm": 0.6475366353988647,
|
| 272 |
+
"learning_rate": 1.3520838665324704e-06,
|
| 273 |
+
"loss": 0.0471,
|
| 274 |
+
"step": 310
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 7.111111111111111,
|
| 278 |
+
"grad_norm": 0.598598062992096,
|
| 279 |
+
"learning_rate": 1.1834346895340817e-06,
|
| 280 |
+
"loss": 0.062,
|
| 281 |
+
"step": 320
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 7.333333333333333,
|
| 285 |
+
"grad_norm": 0.4335130751132965,
|
| 286 |
+
"learning_rate": 1.0227034912257222e-06,
|
| 287 |
+
"loss": 0.0548,
|
| 288 |
+
"step": 330
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 7.555555555555555,
|
| 292 |
+
"grad_norm": 1.1936215162277222,
|
| 293 |
+
"learning_rate": 8.708569279463622e-07,
|
| 294 |
+
"loss": 0.0417,
|
| 295 |
+
"step": 340
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 7.777777777777778,
|
| 299 |
+
"grad_norm": 0.7542036771774292,
|
| 300 |
+
"learning_rate": 7.288082227945007e-07,
|
| 301 |
+
"loss": 0.0677,
|
| 302 |
+
"step": 350
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.777777777777778,
|
| 306 |
+
"eval_loss": 0.12392617762088776,
|
| 307 |
+
"eval_runtime": 2.0536,
|
| 308 |
+
"eval_samples_per_second": 38.956,
|
| 309 |
+
"eval_steps_per_second": 19.478,
|
| 310 |
+
"step": 350
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 8.0,
|
| 314 |
+
"grad_norm": 0.44422823190689087,
|
| 315 |
+
"learning_rate": 5.974116733969989e-07,
|
| 316 |
+
"loss": 0.0847,
|
| 317 |
+
"step": 360
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 8.222222222222221,
|
| 321 |
+
"grad_norm": 0.7955084443092346,
|
| 322 |
+
"learning_rate": 4.774575140626317e-07,
|
| 323 |
+
"loss": 0.0621,
|
| 324 |
+
"step": 370
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 8.444444444444445,
|
| 328 |
+
"grad_norm": 1.8487240076065063,
|
| 329 |
+
"learning_rate": 3.6966716321997195e-07,
|
| 330 |
+
"loss": 0.0657,
|
| 331 |
+
"step": 380
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 8.666666666666666,
|
| 335 |
+
"grad_norm": 0.7747776508331299,
|
| 336 |
+
"learning_rate": 2.7468888472207606e-07,
|
| 337 |
+
"loss": 0.07,
|
| 338 |
+
"step": 390
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.88888888888889,
|
| 342 |
+
"grad_norm": 0.8187076449394226,
|
| 343 |
+
"learning_rate": 1.9309388911139427e-07,
|
| 344 |
+
"loss": 0.0486,
|
| 345 |
+
"step": 400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 8.88888888888889,
|
| 349 |
+
"eval_loss": 0.12458845227956772,
|
| 350 |
+
"eval_runtime": 2.0523,
|
| 351 |
+
"eval_samples_per_second": 38.981,
|
| 352 |
+
"eval_steps_per_second": 19.49,
|
| 353 |
+
"step": 400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 9.11111111111111,
|
| 357 |
+
"grad_norm": 2.2537145614624023,
|
| 358 |
+
"learning_rate": 1.2537289829235633e-07,
|
| 359 |
+
"loss": 0.0785,
|
| 360 |
+
"step": 410
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 9.333333333333334,
|
| 364 |
+
"grad_norm": 0.8826271891593933,
|
| 365 |
+
"learning_rate": 7.19331942720991e-08,
|
| 366 |
+
"loss": 0.0592,
|
| 367 |
+
"step": 420
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 9.555555555555555,
|
| 371 |
+
"grad_norm": 0.46676668524742126,
|
| 372 |
+
"learning_rate": 3.309616971855195e-08,
|
| 373 |
+
"loss": 0.0428,
|
| 374 |
+
"step": 430
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.777777777777779,
|
| 378 |
+
"grad_norm": 0.6508820056915283,
|
| 379 |
+
"learning_rate": 9.095395067115764e-09,
|
| 380 |
+
"loss": 0.0591,
|
| 381 |
+
"step": 440
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 10.0,
|
| 385 |
+
"grad_norm": 1.4323163032531738,
|
| 386 |
+
"learning_rate": 7.52138005810643e-11,
|
| 387 |
+
"loss": 0.075,
|
| 388 |
+
"step": 450
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 10.0,
|
| 392 |
+
"eval_loss": 0.12433161586523056,
|
| 393 |
+
"eval_runtime": 2.0514,
|
| 394 |
+
"eval_samples_per_second": 38.997,
|
| 395 |
+
"eval_steps_per_second": 19.498,
|
| 396 |
+
"step": 450
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 10.0,
|
| 400 |
+
"step": 450,
|
| 401 |
+
"total_flos": 4.027787393354957e+16,
|
| 402 |
+
"train_loss": 0.26821523129940034,
|
| 403 |
+
"train_runtime": 642.399,
|
| 404 |
+
"train_samples_per_second": 11.208,
|
| 405 |
+
"train_steps_per_second": 0.7
|
| 406 |
+
}
|
| 407 |
+
],
|
| 408 |
+
"logging_steps": 10,
|
| 409 |
+
"max_steps": 450,
|
| 410 |
+
"num_input_tokens_seen": 0,
|
| 411 |
+
"num_train_epochs": 10,
|
| 412 |
+
"save_steps": 50,
|
| 413 |
+
"stateful_callbacks": {
|
| 414 |
+
"TrainerControl": {
|
| 415 |
+
"args": {
|
| 416 |
+
"should_epoch_stop": false,
|
| 417 |
+
"should_evaluate": false,
|
| 418 |
+
"should_log": false,
|
| 419 |
+
"should_save": true,
|
| 420 |
+
"should_training_stop": true
|
| 421 |
+
},
|
| 422 |
+
"attributes": {}
|
| 423 |
+
}
|
| 424 |
+
},
|
| 425 |
+
"total_flos": 4.027787393354957e+16,
|
| 426 |
+
"train_batch_size": 2,
|
| 427 |
+
"trial_name": null,
|
| 428 |
+
"trial_params": null
|
| 429 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|