mbudisic commited on
Commit
354fd7d
Β·
1 Parent(s): c97ecaf

Chainlit.md updated with README.md info

Browse files
Files changed (1) hide show
  1. chainlit.md +80 -8
chainlit.md CHANGED
@@ -1,14 +1,86 @@
1
- # Welcome to Chainlit! πŸš€πŸ€–
 
 
 
 
 
 
 
 
 
2
 
3
- Hi there, Developer! πŸ‘‹ We're excited to have you on board. Chainlit is a powerful tool designed to help you prototype, debug and share applications built on top of LLMs.
4
 
5
- ## Useful Links πŸ”—
6
 
7
- - **Documentation:** Get started with our comprehensive [Chainlit Documentation](https://docs.chainlit.io) πŸ“š
8
- - **Discord Community:** Join our friendly [Chainlit Discord](https://discord.gg/k73SQ3FyUh) to ask questions, share your projects, and connect with other developers! πŸ’¬
9
 
10
- We can't wait to see what you create with Chainlit! Happy coding! πŸ’»πŸ˜Š
11
 
12
- ## Welcome screen
13
 
14
- To modify the welcome screen, edit the `chainlit.md` file at the root of your project. If you do not want a welcome screen, just leave this file empty.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ title: PsTuts RAG
3
+ emoji: πŸ’»
4
+ colorFrom: yellow
5
+ colorTo: indigo
6
+ sdk: docker
7
+ pinned: false
8
+ license: mit
9
+ short_description: Agentic RAG that interrogates the PsTuts dataset.
10
+ ---
11
 
12
+ Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
13
 
14
+ # πŸ€– PsTuts RAG System
15
 
16
+ An agentic RAG system for the PsTuts dataset that provides AI-powered answers to Adobe Photoshop questions using video tutorial transcripts.
 
17
 
18
+ ## 🎨 Visual Theme
19
 
20
+ The application features a beautiful **sepia-toned color scheme** that gives it a vintage, artistic feel perfect for Adobe Photoshop tutorials:
21
 
22
+ - πŸŒ… **Light Theme** (default): Warm cream and tan colors reminiscent of old photography
23
+ - πŸŒ™ **Dark Theme**: Rich coffee and amber tones for comfortable nighttime usage
24
+ - πŸ–ΌοΈ **Sepia Filter**: Subtle sepia treatment on images for visual consistency
25
+ - ⚑ **Smooth Transitions**: Elegant animations when switching between themes
26
+
27
+ Users can toggle between light and dark variants using the theme switcher in the interface.
28
+
29
+ ## πŸš€ Getting Started
30
+
31
+ 1. Install dependencies:
32
+
33
+ ```bash
34
+ # Basic installation (includes Jupyter support)
35
+ pip install -e .
36
+
37
+ # With development tools
38
+ pip install -e ".[dev]"
39
+
40
+ # With web server components
41
+ pip install -e ".[web]"
42
+
43
+ # With additional extras (numpy, ragas, tavily)
44
+ pip install -e ".[extras]"
45
+
46
+ # Full installation with all features
47
+ pip install -e ".[dev,web,extras]"
48
+ ```
49
+
50
+ 2. Run the app:
51
+ ```bash
52
+ chainlit run app.py
53
+ ```
54
+
55
+ 3. Open your browser and navigate to `http://localhost:8000`
56
+
57
+ ## πŸ’‘ Features
58
+
59
+ - Retrieval-augmented generation (RAG) for Photoshop tutorials
60
+ - Multi-agent system with team supervisor
61
+ - Web search integration via Tavily
62
+ - Semantic chunking for better context retrieval
63
+ - Interactive chat interface through Chainlit
64
+
65
+ ## βš™οΈ Configuration Options
66
+
67
+ You can customize the behavior of PsTuts RAG using environment variables. Set these in your shell, `.env` file, or deployment environment. Here are the available options:
68
+
69
+ | Env Var | Description |
70
+ |---------|-------------|
71
+ | `EVA_WORKFLOW_NAME` | 🏷️ Name of the EVA workflow. Default: `EVA_workflow` |
72
+ | `EVA_LOG_LEVEL` | πŸͺ΅ Logging level for EVA. Default: `INFO` |
73
+ | `TRANSCRIPT_GLOB` | πŸ“„ Glob pattern for transcript JSON files (supports multiple files separated by `:`). Default: `data/test.json` |
74
+ | `EMBEDDING_MODEL` | 🧊 Name of the embedding model to use (default: custom fine-tuned snowflake model). Default: `mbudisic/snowflake-arctic-embed-s-ft-pstuts` |
75
+ | `EVA_STRIP_THINK` | πŸ’­ If set (present in env), strips 'think' steps from EVA output. |
76
+ | `EMBEDDING_API` | πŸ”Œ API provider for embeddings (`OPENAI`, `HUGGINGFACE`, or `OLLAMA`). Default: `HUGGINGFACE` |
77
+ | `LLM_API` | πŸ€– API provider for LLM (`OPENAI`, `HUGGINGFACE`, or `OLLAMA`). Default: `OLLAMA` |
78
+ | `MAX_RESEARCH_LOOPS` | πŸ” Maximum number of research loops to perform. Default: `3` |
79
+ | `LLM_TOOL_MODEL` | πŸ› οΈ Name of the LLM model to use for tool calling. Default: `smollm2:1.7b-instruct-q2_K` |
80
+ | `N_CONTEXT_DOCS` | πŸ“š Number of context documents to retrieve for RAG. Default: `2` |
81
+ | `EVA_SEARCH_PERMISSION` | 🌐 Permission for search (`yes`, `no`, or `ask`). Default: `no` |
82
+ | `EVA_DB_PERSIST` | πŸ’Ύ Path or flag for DB persistence. Default: unset |
83
+ | `EVA_REINITIALIZE` | πŸ”„ If true, reinitializes EVA DB. Default: `False` |
84
+ | `THREAD_ID` | 🧡 Thread ID for the current session. Default: unset |
85
+
86
+ Set these variables to control model selection, logging, search permissions, and more. For advanced usage, see the developer documentation.