jerpint commited on
Commit
f8c09da
1 Parent(s): 4a4bbfd

update README, add better docs

Browse files
Files changed (2) hide show
  1. README.md +4 -4
  2. app.py +19 -5
README.md CHANGED
@@ -24,9 +24,9 @@ RAGTheDocs is an open-source library that allows you to deploy retrieval augment
24
  ![image](https://github.com/jerpint/buster/assets/18450628/0c89038c-c3af-4c1f-9d3b-9b4d83db4910)
25
 
26
  3) Set your environment variables:
27
- * OPENAI_API_KEY: Needed for the app to work, e.g. `sk-...`
28
- * READTHEDOCS_URL: The url of the website you are interested in scraping
29
- * READTHEDOCS_VERSION: This is important only if there exist multiple versions of the docs (e.g. "en/v0.2.7" or "en/latest"). If left empty, it will scrape all available versions.
30
 
31
  **WARNING** This library is experimental and automatically calls OpenAI APIs for you. Use at your own risk! ⚠️
32
 
@@ -35,6 +35,6 @@ RAGTheDocs is an open-source library that allows you to deploy retrieval augment
35
 
36
  - **Web Scraping and embeddings:** RAGtheDocs automatically scrapes and embeds documentation from any website generated by ReadTheDocs/Sphinx using OpenAI embeddings
37
 
38
- - **RAG Interface:** It comes built-in with a gradio UI for users to interact with [Buster 🤖](github.com/jerpint/buster) our RAG agent.
39
 
40
  - **Customization Options:** Tailor RAGtheDocs to your needs with customizable settings and options.
 
24
  ![image](https://github.com/jerpint/buster/assets/18450628/0c89038c-c3af-4c1f-9d3b-9b4d83db4910)
25
 
26
  3) Set your environment variables:
27
+ * `OPENAI_API_KEY`: Needed for the app to work, e.g. `sk-...`
28
+ * `READTHEDOCS_URL`: The url of the website you are interested in scraping
29
+ * `READTHEDOCS_VERSION`: This is important only if there exist multiple versions of the docs (e.g. "en/v0.2.7" or "en/latest"). If left empty, it will scrape all available versions.
30
 
31
  **WARNING** This library is experimental and automatically calls OpenAI APIs for you. Use at your own risk! ⚠️
32
 
 
35
 
36
  - **Web Scraping and embeddings:** RAGtheDocs automatically scrapes and embeds documentation from any website generated by ReadTheDocs/Sphinx using OpenAI embeddings
37
 
38
+ - **RAG Interface:** It comes built-in with a gradio UI for users to interact with [Buster 🤖](https://github.com/jerpint/buste) our RAG agent.
39
 
40
  - **Customization Options:** Tailor RAGtheDocs to your needs with customizable settings and options.
app.py CHANGED
@@ -22,6 +22,7 @@ homepage_url = os.getenv("READTHEDOCS_URL") # e.g. "https://orion.readthedocs.io
22
  target_version = os.getenv("READTHEDOCS_VERSION") # e.g. "en/stable"
23
 
24
  # scrape and embed content from readthedocs website
 
25
  scrape_rtd(
26
  homepage_url=homepage_url, save_directory="outputs/", target_version=target_version
27
  )
@@ -101,7 +102,24 @@ def chat(chat_history: ChatHistory) -> Tuple[ChatHistory, Completion]:
101
  demo = gr.Blocks()
102
  with demo:
103
  with gr.Row():
104
- gr.Markdown("<h3><center>RAGTheDocs</center></h3>")
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
105
 
106
  chatbot = gr.Chatbot()
107
 
@@ -122,10 +140,6 @@ with demo:
122
  inputs=question,
123
  )
124
 
125
- gr.Markdown(
126
- "This app uses [Buster 🤖](github.com/jerpint/buster) and ChatGPT to search the docs for relevant info and answer questions."
127
- )
128
-
129
  response = gr.State()
130
 
131
  # fmt: off
 
22
  target_version = os.getenv("READTHEDOCS_VERSION") # e.g. "en/stable"
23
 
24
  # scrape and embed content from readthedocs website
25
+ # comment out if already embedded locally to avoid extra costs
26
  scrape_rtd(
27
  homepage_url=homepage_url, save_directory="outputs/", target_version=target_version
28
  )
 
102
  demo = gr.Blocks()
103
  with demo:
104
  with gr.Row():
105
+ gr.Markdown("<h1><center>RAGTheDocs</center></h1>")
106
+
107
+ gr.Markdown(
108
+ """
109
+ ## About
110
+ RAGTheDocs allows you to ask questions about any documentation hosted on readthedocs.
111
+ Simply clone this space and point it to the right URL!
112
+
113
+ Try it out by asking a question below about [orion](https://orion.readthedocs.io/), an open-source hyperparameter optimization library.
114
+
115
+ ## How it works
116
+ This app uses [Buster 🤖](https://github.com/jerpint/buster) and ChatGPT to search the docs for relevant info and
117
+ answer questions.
118
+ View the code on the [project homepage](https://github.com/jerpint/RAGTheDocs)
119
+ """
120
+
121
+ )
122
+
123
 
124
  chatbot = gr.Chatbot()
125
 
 
140
  inputs=question,
141
  )
142
 
 
 
 
 
143
  response = gr.State()
144
 
145
  # fmt: off