update readme

README.md

@@ -12,7 +12,9 @@ OriGen is a fine-tuned lora model designed for Verilog code generation. It is tr
 
 OriGen_Fix is a fine-tuned lora model designed for fixing syntax errors in Verilog code. It is trained based on OriGen using debug datasets. The datasets can be found in the [origen_dataset_debug](https://huggingface.co/datasets/henryen/origen_dataset_debug).
 
-The models have been uploaded to Hugging Face, and the repository contains the inference scripts. The data generation flow will be released soon.
+There is another dataset called [origen_dataset_description](https://huggingface.co/datasets/henryen/origen_dataset_description). This dataset is in the format of `Description` and `Code`, and [origen_dataset_instruction](https://huggingface.co/datasets/henryen/origen_dataset_instruction) is obtained by converting the `Description` to `Instruction`.
+
+The models and datasets have been uploaded to Hugging Face, and the repository contains the inference scripts. The data generation flow will be released soon.
 
 - **Huggingface**: 
     - https://huggingface.co/henryen/OriGen
@@ -20,6 +22,7 @@ The models have been uploaded to Hugging Face, and the repository contains the i
 - **Dataset**: 
     - https://huggingface.co/datasets/henryen/origen_dataset_instruction
     - https://huggingface.co/datasets/henryen/origen_dataset_debug
+    - https://huggingface.co/datasets/henryen/origen_dataset_description
 - **Repository**: https://github.com/pku-liang/OriGen
 
 ### Evaluation Results