Update directory structure for NPU-related files

- Create a new `npu` directory under `assets` to store NPU-related images.
- Move NPU-related `README` files to the `ecosystem` directory.

ecosystem/README_npu.md

@@ -14,8 +14,8 @@
     <div> </div>
   </div>
 
-[![license](./assets/license.svg)](./LICENSE)
-[![evaluation](./assets/compass_support.svg)](https://github.com/internLM/OpenCompass/)
+[![license](../assets/license.svg)](../LICENSE)
+[![evaluation](../assets/compass_support.svg)](https://github.com/internLM/OpenCompass/)
 
 <!-- [![Documentation Status](https://readthedocs.org/projects/internlm/badge/?version=latest)](https://internlm.readthedocs.io/zh_CN/latest/?badge=latest) -->
 
@@ -28,8 +28,8 @@
 [🔗API](https://internlm.intern-ai.org.cn/api/document) |
 [🧩Modelers](https://modelers.cn/spaces/MindSpore-Lab/INTERNLM2-20B-PLAN)
 
-[English](./README_npu.md) |
-[简体中文](./README_npu_zh-CN.md)
+[English](README_npu.md) |
+[简体中文](README_npu_zh-CN.md)
 
 </div>
 
@@ -140,7 +140,7 @@ NPROC_PER_NODE=8 xtuner train internlm3_8b_instruct_lora_oasst1_e10.py --deepspe
 The fine-tuning results are saved in the directory `./work_dirs/internlm3_8b_instruct_lora_oasst1_e10/iter_xxx.pth`.
 The comparison of loss between NPU and GPU is as follows:
 
-![xtuner_training_loss](assets/xtuner_training_loss_compare.png)
+![xtuner_training_loss](../assets/npu/xtuner_training_loss_compare.png)
 
 ### Model Convert
 
@@ -254,11 +254,11 @@ llamafactory-cli train examples/train_full/internlm3_8b_instruct_full_sft.yaml
 
 The loss curve obtained after finetuning is as follows:
 
-![training_loss](assets/lf_training_loss_npu.png)
+![training_loss](../assets/npu/lf_training_loss_npu.png)
 
 The loss curve compared with GPU is as follows:
 
-![training_loss_compare](assets/lf_training_loss_compare.png)
+![training_loss_compare](../assets/npu/lf_training_loss_compare.png)
 
 ## Transformers
 

ecosystem/README_npu_zh-CN.md

@@ -28,8 +28,8 @@
 [🔗API](https://internlm.intern-ai.org.cn/api/document) |
 [🧩魔乐社区](https://modelers.cn/spaces/MindSpore-Lab/INTERNLM2-20B-PLAN)
 
-[English](./README_npu.md) |
-[简体中文](./README_npu_zh-CN.md)
+[English](README_npu.md) |
+[简体中文](README_npu_zh-CN.md)
 
 </div>
 
@@ -139,7 +139,7 @@ NPROC_PER_NODE=8 xtuner train internlm3_8b_instruct_lora_oasst1_e10.py --deepspe
 
 微调后结果保存在`./work_dirs/internlm3_8b_instruct_lora_oasst1_e10/iter_xxx.pth`，NPU与GPU的loss对比如下：
 
-![xtuner_training_loss](assets/xtuner_training_loss_compare.png)
+![xtuner_training_loss](../assets/npu/xtuner_training_loss_compare.png)
 
 ### 模型转换
 
@@ -250,11 +250,11 @@ llamafactory-cli train examples/train_full/internlm3_8b_instruct_full_sft.yaml
 
 微调后得到的loss曲线如下：
 
-![training_loss](assets/lf_training_loss_npu.png)
+![training_loss](../assets/npu/lf_training_loss_npu.png)
 
 与GPU对比的loss曲线如下：
 
-![training_loss_compare](assets/lf_training_loss_compare.png)
+![training_loss_compare](../assets/npu/lf_training_loss_compare.png)
 
 ## Transformers