docs: update experiment documentation to include Adaptive-RAG and RQRAG

- Added sections for Adaptive-RAG and RQRAG in both English and Chinese documentation. - Included additional steps for downloading models and modifying function parameters for both methods.
RUC-NLPIR · Jan 11, 2025 · 264d8c7 · 264d8c7
1 parent 0862158
commit 264d8c7
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 0 deletions.
diff --git a/docs/reproduce_experiment.md b/docs/reproduce_experiment.md
@@ -35,6 +35,8 @@ Table of Contents:
 - [SKR](#skr)
 - [Self-RAG](#self-rag)
 - [Spring](#spring)
+- [Adaptive-RAG](#adaptive-rag)
+- [RQRAG](#rqrag)
 
 #### AAR
 
@@ -106,6 +108,13 @@ If the official open-source checkpoint is released in the future, we will update
 - Additional Step1: Download classifier model from huggingface repo (**not official**): [illuminoplanet/combined_flan_t5_xl_classifier](https://huggingface.co/illuminoplanet/combined_flan_t5_xl_classifier)
 - Additional Step2: Modify the `model_path` in `adaptive` function.
 
+#### RQRAG
+
+This method requires downloading the RQRAG model.
+
+- Additional Step1: Download RQRAG model from huggingface repo: [zorowin123/rq_rag_llama2_7B](https://huggingface.co/zorowin123/rq_rag_llama2_7B)
+- Additional Step2: Modify the `generator_model_path` in the `rqrag` function.
+
 ### 3. Run methods
 
 Run the experiment on the NQ dataset using the following command.

diff --git a/docs/reproduce_experiment_zh.md b/docs/reproduce_experiment_zh.md
@@ -33,6 +33,8 @@
 - [SKR](#skr)
 - [Self-RAG](#self-rag)
 - [Spring](#spring)
+- [Adaptive-RAG](#adaptive-rag)
+- [RQRAG](#rqrag)
 
 #### AAR
 
@@ -104,6 +106,13 @@
 - 额外步骤1：从 Huggingface 仓库下载分类器模型（**非官方**）：[illuminoplanet/combined_flan_t5_xl_classifier](https://huggingface.co/illuminoplanet/combined_flan_t5_xl_classifier)
 - 额外步骤2：修改 `adaptive` 函数中的 `model_path`。
 
+#### RQRAG
+
+这种方法需要下载 RQRAG 模型。
+
+- 额外步骤1：从 Huggingface 仓库下载 RQRAG 模型：[zorowin123/rq_rag_llama2_7B](https://huggingface.co/zorowin123/rq_rag_llama2_7B)
+- 额外步骤2：在 `rqrag` 函数中修改 `generator_model_path`
+
 
 ### 3. 运行方法