Claude/docker deployment guide 015g1cum b lu mc ytpeq9cdeo8 #13149
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Add detailed Docker deployment instructions - Include 3 deployment methods: pre-built images, build from source, and docker-compose - Add configuration examples and troubleshooting section - Provide Vietnamese language guide for local users
- Analyze Rasa's Vietnamese language capabilities (75-90% accuracy) - Provide 3 pipeline configurations: basic, advanced, and PyVi-based - Include training data examples and best practices - Add performance benchmarks and troubleshooting tips - Recommend PyVi tokenizer for optimal Vietnamese NLP
- Explain training module purpose and functionality (not code details) - Cover 4 types of training: full, NLU-only, core-only, and incremental - Document training process from data preparation to model output - Include 7 key features: fingerprinting, augmentation, conflict detection, etc. - Provide real-world examples for flight booking chatbot - Add best practices and troubleshooting guide
- Explain that Rasa does NOT use LLMs (GPT-4, Claude, etc.) - Document Rasa's custom transformers (DIETClassifier, TEDPolicy) - Describe optional pre-trained models (BERT, GPT-2, RoBERTa) - Compare pre-trained language models vs large language models - Provide detailed architecture and feature comparison - Include decision tree for choosing Rasa vs LLM chatbots - Show 4 methods to integrate Rasa with LLMs - Add real-world examples and use case recommendations
…with Rasa+LLM integration - Explain Classification model (Rasa): label selection, predictable, fast - Explain Generative model (LLM): text generation, flexible, creative - Detailed comparison with examples and analogies - 3 real-world approaches: pure Rasa, pure LLM, hybrid (recommended) - 4 integration methods: LLM fallback, response enhancement, intent clarification, hybrid router - Include code examples, metrics tracking, cost optimization - Add best practices: caching, monitoring, prompt engineering, A/B testing - Provide decision framework for choosing the right approach
…/Claude chatbot Tạo module tích hợp LLM hoàn chỉnh với các tính năng: - Hỗ trợ OpenAI GPT-4 và Anthropic Claude - LLM Fallback khi Rasa confidence thấp - Response Enhancement cho câu trả lời tự nhiên hơn - Intent Clarification cho các intent mơ hồ - Custom Rasa actions tích hợp LLM - Example hybrid bot với cấu hình đầy đủ - Hướng dẫn chi tiết bằng tiếng Việt Module files: - actions/llm/providers.py: OpenAI + Claude providers - actions/llm/fallback.py: LLM fallback handler - actions/llm/response_enhancer.py: Natural response generation - actions/llm/intent_clarifier.py: Intent disambiguation - actions/actions_llm.py: Custom Rasa actions - examples/hybrid_bot/: Complete example bot - LLM_INTEGRATION_GUIDE_VI.md: Comprehensive documentation
… demos Tạo hệ thống demo hoàn chỉnh cho chatbot bán sản phẩm IT: Web Frontend (frontend/): - index.html: Giao diện hiện đại với hero section, categories, features - style.css: Responsive design, gradient backgrounds, animations - script.js: Chat widget tích hợp Rasa + fallback demo mode Tính năng Frontend: ✅ 6 danh mục sản phẩm: Máy tính, Máy chủ, Switch, Camera, Phần mềm ✅ Chat widget chuyên nghiệp với typing indicator, quick replies ✅ Product cards, price formatting tiếng Việt ✅ Demo mode standalone (không cần Rasa server) ✅ Click-to-demo examples cho từng tính năng IT Store Chatbot (examples/it_store_bot/): - domain.yml: 16 intents cho tư vấn sản phẩm IT - nlu.yml: 200+ training examples tiếng Việt - actions_it_store.py: 10 custom actions với product database - config.yml: Pipeline tối ưu cho tiếng Việt Database Demo: - 4 laptop (gaming + office): ASUS ROG, MSI, Dell, HP - 3 máy chủ: Dell PowerEdge, HP ProLiant, Lenovo - 2 switch: Cisco Catalyst, HPE OfficeConnect - 2 camera: Hikvision, Dahua - 3 phần mềm: Windows, Office Tính năng Chatbot: 🔍 Tìm kiếm sản phẩm theo loại/giá/specs ⚖️ So sánh sản phẩm chi tiết 💡 Tư vấn dựa trên use case 💰 Báo giá + khuyến mãi 🛡️ Chính sách bảo hành 🛠️ Hỗ trợ kỹ thuật Documentation: - IT_STORE_DEMO_GUIDE_VI.md: Hướng dẫn demo đầy đủ * 5 kịch bản demo cho khách hàng * 2 phương án chạy (standalone/full Rasa) * Hướng dẫn tùy chỉnh & triển khai * Tips demo hiệu quả Phù hợp cho: Demo bán hàng, POC, training sales team
|
|
…n tools Tạo hệ thống hướng dẫn và tools đầy đủ cho việc tạo training data: 📚 Main Guide (TRAINING_DATA_GUIDE_VI.md - 1000+ dòng): - Giải thích 3 loại training data: NLU, Stories, Rules - Hướng dẫn chi tiết cách viết intents, entities, examples - Best practices với ví dụ ✅ DO và ❌ DON'T - Synonyms, Lookup Tables, Regex patterns - Stories syntax (basic, với entities, OR, checkpoint) - Rules syntax (simple, conditional, forms) - 7 ví dụ thực tế (đặt đồ ăn, support, e-commerce...) - Hướng dẫn test & validate - Checklist before deploy 📝 Templates (training_templates/): 1. ecommerce_nlu_template.yml (200+ examples) - 20+ intents cho e-commerce - Entities: product_type, brand, price_range, use_case... - Synonyms, lookups, regex patterns 2. ecommerce_stories_template.yml (30+ stories) - Basic flows (greet → search → goodbye) - Product discovery & comparison - Purchase flows (cart → checkout) - Order tracking & cancellation - Error handling 3. QUICK_REFERENCE.md - Cheat sheet syntax (NLU, Stories, Rules) - Best practices summary - Common commands - Common errors & fixes 🛠️ Tools (tools/): 1. generate_training_data.py - Auto-generate examples từ templates - Entity annotator tự động - Data augmentation (synonyms, typos) - Built-in validator 2. validate_training_data.py - Kiểm tra quality: min/max examples, duplicates - Detect intent similarity/overlap - Entity usage analysis - Beautiful validation report - Exit codes cho CI/CD Features: ✅ Hướng dẫn từ cơ bản đến nâng cao (tiếng Việt) ✅ Templates production-ready cho e-commerce ✅ Auto-generate & validate tools ✅ Best practices với examples cụ thể ✅ Quick reference để tra cứu nhanh ✅ Workflow đề xuất cho beginners Use Cases: - Học cách tạo training data từ đầu - Copy templates để start project nhanh - Validate data quality trước khi train - Auto-generate data cho testing
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed changes:
Status (please check what you already did):
black(please check Readme for instructions)