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 @@ 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-900
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# Llama-3.1-8B-Instruct-SFT-900
|
| 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: 0.1053
|
| 21 |
|
|
|
|
| 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-900
|
|
|
|
| 16 |
|
| 17 |
# Llama-3.1-8B-Instruct-SFT-900
|
| 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_900 dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.1053
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.876543209876543,
|
| 3 |
+
"eval_loss": 0.10525013506412506,
|
| 4 |
+
"eval_runtime": 2.2656,
|
| 5 |
+
"eval_samples_per_second": 39.725,
|
| 6 |
+
"eval_steps_per_second": 19.862,
|
| 7 |
+
"total_flos": 4.40152941896663e+16,
|
| 8 |
+
"train_loss": 0.2507400060892105,
|
| 9 |
+
"train_runtime": 715.6643,
|
| 10 |
+
"train_samples_per_second": 11.318,
|
| 11 |
+
"train_steps_per_second": 0.699
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.876543209876543,
|
| 3 |
+
"eval_loss": 0.10525013506412506,
|
| 4 |
+
"eval_runtime": 2.2656,
|
| 5 |
+
"eval_samples_per_second": 39.725,
|
| 6 |
+
"eval_steps_per_second": 19.862
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.876543209876543,
|
| 3 |
+
"total_flos": 4.40152941896663e+16,
|
| 4 |
+
"train_loss": 0.2507400060892105,
|
| 5 |
+
"train_runtime": 715.6643,
|
| 6 |
+
"train_samples_per_second": 11.318,
|
| 7 |
+
"train_steps_per_second": 0.699
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,472 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.10525013506412506,
|
| 3 |
+
"best_model_checkpoint": "saves/Llama-3.1-8B-Instruct/lora/sft-900/checkpoint-450",
|
| 4 |
+
"epoch": 9.876543209876543,
|
| 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": 0.19753086419753085,
|
| 13 |
+
"grad_norm": 5.285682678222656,
|
| 14 |
+
"learning_rate": 9.000000000000001e-07,
|
| 15 |
+
"loss": 1.6663,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.3950617283950617,
|
| 20 |
+
"grad_norm": 7.212312698364258,
|
| 21 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 22 |
+
"loss": 1.6571,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5925925925925926,
|
| 27 |
+
"grad_norm": 6.918689250946045,
|
| 28 |
+
"learning_rate": 2.9e-06,
|
| 29 |
+
"loss": 1.6049,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.7901234567901234,
|
| 34 |
+
"grad_norm": 4.972626209259033,
|
| 35 |
+
"learning_rate": 3.900000000000001e-06,
|
| 36 |
+
"loss": 1.345,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.9876543209876543,
|
| 41 |
+
"grad_norm": 4.738757133483887,
|
| 42 |
+
"learning_rate": 4.9000000000000005e-06,
|
| 43 |
+
"loss": 1.201,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.9876543209876543,
|
| 48 |
+
"eval_loss": 1.0016428232192993,
|
| 49 |
+
"eval_runtime": 2.2723,
|
| 50 |
+
"eval_samples_per_second": 39.607,
|
| 51 |
+
"eval_steps_per_second": 19.803,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.1851851851851851,
|
| 56 |
+
"grad_norm": 3.143040180206299,
|
| 57 |
+
"learning_rate": 4.9950668210706795e-06,
|
| 58 |
+
"loss": 0.8783,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.382716049382716,
|
| 63 |
+
"grad_norm": 2.624129295349121,
|
| 64 |
+
"learning_rate": 4.978038850628855e-06,
|
| 65 |
+
"loss": 0.6789,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.5802469135802468,
|
| 70 |
+
"grad_norm": 2.1852622032165527,
|
| 71 |
+
"learning_rate": 4.94893812399836e-06,
|
| 72 |
+
"loss": 0.4397,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.7777777777777777,
|
| 77 |
+
"grad_norm": 1.8680675029754639,
|
| 78 |
+
"learning_rate": 4.907906416994146e-06,
|
| 79 |
+
"loss": 0.2305,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.9753086419753085,
|
| 84 |
+
"grad_norm": 1.5398958921432495,
|
| 85 |
+
"learning_rate": 4.855143631968242e-06,
|
| 86 |
+
"loss": 0.1407,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.9753086419753085,
|
| 91 |
+
"eval_loss": 0.15125828981399536,
|
| 92 |
+
"eval_runtime": 2.2668,
|
| 93 |
+
"eval_samples_per_second": 39.703,
|
| 94 |
+
"eval_steps_per_second": 19.851,
|
| 95 |
+
"step": 100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.1728395061728394,
|
| 99 |
+
"grad_norm": 2.09860897064209,
|
| 100 |
+
"learning_rate": 4.790906823905599e-06,
|
| 101 |
+
"loss": 0.1319,
|
| 102 |
+
"step": 110
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.3703703703703702,
|
| 106 |
+
"grad_norm": 0.7738865613937378,
|
| 107 |
+
"learning_rate": 4.715508948078037e-06,
|
| 108 |
+
"loss": 0.0906,
|
| 109 |
+
"step": 120
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.567901234567901,
|
| 113 |
+
"grad_norm": 0.3859764039516449,
|
| 114 |
+
"learning_rate": 4.62931733535762e-06,
|
| 115 |
+
"loss": 0.0772,
|
| 116 |
+
"step": 130
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.765432098765432,
|
| 120 |
+
"grad_norm": 0.847908079624176,
|
| 121 |
+
"learning_rate": 4.5327519026175694e-06,
|
| 122 |
+
"loss": 0.0878,
|
| 123 |
+
"step": 140
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.962962962962963,
|
| 127 |
+
"grad_norm": 0.6789230108261108,
|
| 128 |
+
"learning_rate": 4.426283106939474e-06,
|
| 129 |
+
"loss": 0.0885,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.962962962962963,
|
| 134 |
+
"eval_loss": 0.10822762548923492,
|
| 135 |
+
"eval_runtime": 2.2637,
|
| 136 |
+
"eval_samples_per_second": 39.757,
|
| 137 |
+
"eval_steps_per_second": 19.879,
|
| 138 |
+
"step": 150
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.1604938271604937,
|
| 142 |
+
"grad_norm": 0.82857745885849,
|
| 143 |
+
"learning_rate": 4.3104296535936695e-06,
|
| 144 |
+
"loss": 0.0712,
|
| 145 |
+
"step": 160
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.3580246913580245,
|
| 149 |
+
"grad_norm": 0.6130207777023315,
|
| 150 |
+
"learning_rate": 4.185755968959308e-06,
|
| 151 |
+
"loss": 0.0918,
|
| 152 |
+
"step": 170
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.5555555555555554,
|
| 156 |
+
"grad_norm": 0.8012928366661072,
|
| 157 |
+
"learning_rate": 4.052869450695776e-06,
|
| 158 |
+
"loss": 0.0698,
|
| 159 |
+
"step": 180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 3.753086419753086,
|
| 163 |
+
"grad_norm": 1.5906227827072144,
|
| 164 |
+
"learning_rate": 3.912417508562345e-06,
|
| 165 |
+
"loss": 0.0873,
|
| 166 |
+
"step": 190
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 3.950617283950617,
|
| 170 |
+
"grad_norm": 0.6971777677536011,
|
| 171 |
+
"learning_rate": 3.7650844103029093e-06,
|
| 172 |
+
"loss": 0.0743,
|
| 173 |
+
"step": 200
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 3.950617283950617,
|
| 177 |
+
"eval_loss": 0.10680027306079865,
|
| 178 |
+
"eval_runtime": 2.2702,
|
| 179 |
+
"eval_samples_per_second": 39.644,
|
| 180 |
+
"eval_steps_per_second": 19.822,
|
| 181 |
+
"step": 200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 4.148148148148148,
|
| 185 |
+
"grad_norm": 0.9408888816833496,
|
| 186 |
+
"learning_rate": 3.611587947962319e-06,
|
| 187 |
+
"loss": 0.0622,
|
| 188 |
+
"step": 210
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.345679012345679,
|
| 192 |
+
"grad_norm": 0.5946834087371826,
|
| 193 |
+
"learning_rate": 3.452675940875686e-06,
|
| 194 |
+
"loss": 0.0883,
|
| 195 |
+
"step": 220
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.54320987654321,
|
| 199 |
+
"grad_norm": 0.592318594455719,
|
| 200 |
+
"learning_rate": 3.2891225923677565e-06,
|
| 201 |
+
"loss": 0.0546,
|
| 202 |
+
"step": 230
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 4.7407407407407405,
|
| 206 |
+
"grad_norm": 0.7733595967292786,
|
| 207 |
+
"learning_rate": 3.121724717912138e-06,
|
| 208 |
+
"loss": 0.052,
|
| 209 |
+
"step": 240
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 4.938271604938271,
|
| 213 |
+
"grad_norm": 0.7871303558349609,
|
| 214 |
+
"learning_rate": 2.9512978631264006e-06,
|
| 215 |
+
"loss": 0.0855,
|
| 216 |
+
"step": 250
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 4.938271604938271,
|
| 220 |
+
"eval_loss": 0.10621543228626251,
|
| 221 |
+
"eval_runtime": 2.2684,
|
| 222 |
+
"eval_samples_per_second": 39.675,
|
| 223 |
+
"eval_steps_per_second": 19.837,
|
| 224 |
+
"step": 250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.135802469135802,
|
| 228 |
+
"grad_norm": 0.4525160789489746,
|
| 229 |
+
"learning_rate": 2.778672330515814e-06,
|
| 230 |
+
"loss": 0.0887,
|
| 231 |
+
"step": 260
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 5.333333333333333,
|
| 235 |
+
"grad_norm": 1.1445720195770264,
|
| 236 |
+
"learning_rate": 2.604689134322999e-06,
|
| 237 |
+
"loss": 0.058,
|
| 238 |
+
"step": 270
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 5.530864197530864,
|
| 242 |
+
"grad_norm": 1.6000568866729736,
|
| 243 |
+
"learning_rate": 2.4301959031910785e-06,
|
| 244 |
+
"loss": 0.0671,
|
| 245 |
+
"step": 280
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 5.728395061728395,
|
| 249 |
+
"grad_norm": 0.7566145062446594,
|
| 250 |
+
"learning_rate": 2.256042750602127e-06,
|
| 251 |
+
"loss": 0.0563,
|
| 252 |
+
"step": 290
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 5.925925925925926,
|
| 256 |
+
"grad_norm": 0.996746838092804,
|
| 257 |
+
"learning_rate": 2.0830781332097446e-06,
|
| 258 |
+
"loss": 0.0571,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 5.925925925925926,
|
| 263 |
+
"eval_loss": 0.10575470328330994,
|
| 264 |
+
"eval_runtime": 2.2646,
|
| 265 |
+
"eval_samples_per_second": 39.742,
|
| 266 |
+
"eval_steps_per_second": 19.871,
|
| 267 |
+
"step": 300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.1234567901234565,
|
| 271 |
+
"grad_norm": 0.46847003698349,
|
| 272 |
+
"learning_rate": 1.912144717243525e-06,
|
| 273 |
+
"loss": 0.0644,
|
| 274 |
+
"step": 310
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 6.320987654320987,
|
| 278 |
+
"grad_norm": 0.6955059766769409,
|
| 279 |
+
"learning_rate": 1.744075273123889e-06,
|
| 280 |
+
"loss": 0.0574,
|
| 281 |
+
"step": 320
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 6.518518518518518,
|
| 285 |
+
"grad_norm": 1.0270895957946777,
|
| 286 |
+
"learning_rate": 1.5796886182883053e-06,
|
| 287 |
+
"loss": 0.0642,
|
| 288 |
+
"step": 330
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 6.716049382716049,
|
| 292 |
+
"grad_norm": 0.8975910544395447,
|
| 293 |
+
"learning_rate": 1.419785627995044e-06,
|
| 294 |
+
"loss": 0.0584,
|
| 295 |
+
"step": 340
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 6.91358024691358,
|
| 299 |
+
"grad_norm": 1.245640754699707,
|
| 300 |
+
"learning_rate": 1.2651453335394232e-06,
|
| 301 |
+
"loss": 0.063,
|
| 302 |
+
"step": 350
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 6.91358024691358,
|
| 306 |
+
"eval_loss": 0.10544978082180023,
|
| 307 |
+
"eval_runtime": 2.2637,
|
| 308 |
+
"eval_samples_per_second": 39.758,
|
| 309 |
+
"eval_steps_per_second": 19.879,
|
| 310 |
+
"step": 350
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 7.111111111111111,
|
| 314 |
+
"grad_norm": 0.7810999155044556,
|
| 315 |
+
"learning_rate": 1.11652112689164e-06,
|
| 316 |
+
"loss": 0.0585,
|
| 317 |
+
"step": 360
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 7.308641975308642,
|
| 321 |
+
"grad_norm": 1.3329459428787231,
|
| 322 |
+
"learning_rate": 9.746370902468311e-07,
|
| 323 |
+
"loss": 0.0556,
|
| 324 |
+
"step": 370
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 7.506172839506172,
|
| 328 |
+
"grad_norm": 0.9610387086868286,
|
| 329 |
+
"learning_rate": 8.40184468369396e-07,
|
| 330 |
+
"loss": 0.0665,
|
| 331 |
+
"step": 380
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 7.703703703703704,
|
| 335 |
+
"grad_norm": 1.5703972578048706,
|
| 336 |
+
"learning_rate": 7.138183009179922e-07,
|
| 337 |
+
"loss": 0.0513,
|
| 338 |
+
"step": 390
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.901234567901234,
|
| 342 |
+
"grad_norm": 0.517612636089325,
|
| 343 |
+
"learning_rate": 5.961542311581586e-07,
|
| 344 |
+
"loss": 0.0597,
|
| 345 |
+
"step": 400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 7.901234567901234,
|
| 349 |
+
"eval_loss": 0.10566332191228867,
|
| 350 |
+
"eval_runtime": 2.265,
|
| 351 |
+
"eval_samples_per_second": 39.736,
|
| 352 |
+
"eval_steps_per_second": 19.868,
|
| 353 |
+
"step": 400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 8.098765432098766,
|
| 357 |
+
"grad_norm": 0.3458334803581238,
|
| 358 |
+
"learning_rate": 4.87765506610215e-07,
|
| 359 |
+
"loss": 0.0481,
|
| 360 |
+
"step": 410
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 8.296296296296296,
|
| 364 |
+
"grad_norm": 1.3999119997024536,
|
| 365 |
+
"learning_rate": 3.891801862449629e-07,
|
| 366 |
+
"loss": 0.0474,
|
| 367 |
+
"step": 420
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 8.493827160493828,
|
| 371 |
+
"grad_norm": 1.124482274055481,
|
| 372 |
+
"learning_rate": 3.0087856783345916e-07,
|
| 373 |
+
"loss": 0.0455,
|
| 374 |
+
"step": 430
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.691358024691358,
|
| 378 |
+
"grad_norm": 1.1105051040649414,
|
| 379 |
+
"learning_rate": 2.2329084798455747e-07,
|
| 380 |
+
"loss": 0.0609,
|
| 381 |
+
"step": 440
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 8.88888888888889,
|
| 385 |
+
"grad_norm": 1.2002372741699219,
|
| 386 |
+
"learning_rate": 1.567950262702714e-07,
|
| 387 |
+
"loss": 0.0694,
|
| 388 |
+
"step": 450
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 8.88888888888889,
|
| 392 |
+
"eval_loss": 0.10525013506412506,
|
| 393 |
+
"eval_runtime": 2.2672,
|
| 394 |
+
"eval_samples_per_second": 39.696,
|
| 395 |
+
"eval_steps_per_second": 19.848,
|
| 396 |
+
"step": 450
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 9.08641975308642,
|
| 400 |
+
"grad_norm": 0.9672008752822876,
|
| 401 |
+
"learning_rate": 1.0171506364985622e-07,
|
| 402 |
+
"loss": 0.0483,
|
| 403 |
+
"step": 460
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 9.283950617283951,
|
| 407 |
+
"grad_norm": 0.4676777422428131,
|
| 408 |
+
"learning_rate": 5.83193041645802e-08,
|
| 409 |
+
"loss": 0.0691,
|
| 410 |
+
"step": 470
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 9.481481481481481,
|
| 414 |
+
"grad_norm": 2.014739751815796,
|
| 415 |
+
"learning_rate": 2.681916759252917e-08,
|
| 416 |
+
"loss": 0.0612,
|
| 417 |
+
"step": 480
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 9.679012345679013,
|
| 421 |
+
"grad_norm": 1.5374245643615723,
|
| 422 |
+
"learning_rate": 7.368119432699383e-09,
|
| 423 |
+
"loss": 0.0461,
|
| 424 |
+
"step": 490
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 9.876543209876543,
|
| 428 |
+
"grad_norm": 0.8854678273200989,
|
| 429 |
+
"learning_rate": 6.092323651313293e-11,
|
| 430 |
+
"loss": 0.0593,
|
| 431 |
+
"step": 500
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 9.876543209876543,
|
| 435 |
+
"eval_loss": 0.10525856912136078,
|
| 436 |
+
"eval_runtime": 2.2702,
|
| 437 |
+
"eval_samples_per_second": 39.645,
|
| 438 |
+
"eval_steps_per_second": 19.822,
|
| 439 |
+
"step": 500
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 9.876543209876543,
|
| 443 |
+
"step": 500,
|
| 444 |
+
"total_flos": 4.40152941896663e+16,
|
| 445 |
+
"train_loss": 0.2507400060892105,
|
| 446 |
+
"train_runtime": 715.6643,
|
| 447 |
+
"train_samples_per_second": 11.318,
|
| 448 |
+
"train_steps_per_second": 0.699
|
| 449 |
+
}
|
| 450 |
+
],
|
| 451 |
+
"logging_steps": 10,
|
| 452 |
+
"max_steps": 500,
|
| 453 |
+
"num_input_tokens_seen": 0,
|
| 454 |
+
"num_train_epochs": 10,
|
| 455 |
+
"save_steps": 50,
|
| 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": 4.40152941896663e+16,
|
| 469 |
+
"train_batch_size": 2,
|
| 470 |
+
"trial_name": null,
|
| 471 |
+
"trial_params": null
|
| 472 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|