Add train loss

README.md

@@ -109,12 +109,21 @@ curl http://<your-node-url>:8000/v1/chat/completions \
 ```
 
 ## result
+Prompt: "Explain Machine Learning to me in a nutshell."
 <div align=center>
-    <img src="./doc/results.pngcd ../"/>
+    <img src="./doc/results.png"/>
 </div>
 
 ### 精度
-DCU与GPU精度一致，推理框架：pytorch。
+训练框架：[Llama-Factory](https://developer.sourcefind.cn/codes/OpenDAS/llama-factory)
+
+训练脚本：[mistral_lora_sft.yaml](llama-factory/train_lora/mistral_lora_sft.yaml)
+
+|   device   |   train_loss   |
+|:----------:|:-------:|
+| DCU K100AI | 0.8139 |
+|  GPU A800  | 0.497 |
+
 
 ## 应用场景
 ### 算法类别

llama-factory/train_full/mistral_full_sft.yaml

@@ -23,6 +23,7 @@ save_steps: 500
 plot_loss: true
 overwrite_output_dir: true
 save_only_model: false
+report_to: none  # choices: [none, wandb, tensorboard, swanlab, mlflow]
 
 ### train
 per_device_train_batch_size: 1

llama-factory/train_lora/mistral_lora_sft.yaml

@@ -26,6 +26,7 @@ save_steps: 500
 plot_loss: true
 overwrite_output_dir: true
 save_only_model: false
+report_to: none  # choices: [none, wandb, tensorboard, swanlab, mlflow]
 
 ### train
 per_device_train_batch_size: 1