Buckets:

download
raw
4.72 kB
#!/usr/bin/env python3
"""
Generate comprehensive YAML documentation for all Kozmoai components.
This script creates detailed documentation files for every component.
"""
import os
import yaml
from pathlib import Path
# Base directory for documentation
DOCS_DIR = Path("src/backend/base/kozmoai/components/docs")
DOCS_DIR.mkdir(parents=True, exist_ok=True)
# Model components documentation
MODELS = {
"groq_model": {
"component_name": "GroqModel",
"display_name": "Groq",
"category": "models",
"overview": {
"summary": "Ultra-fast inference with Groq's LPU technology",
"description": """Groq provides the fastest LLM inference available through their Language Processing Unit (LPU) technology.
Available models:
- Llama 3.1 (8B, 70B, 405B): Meta's latest models
- Mixtral 8x7B: Mistral's mixture-of-experts model
- Gemma 2 (9B, 27B): Google's efficient models
Key advantages:
- Extremely fast inference (500+ tokens/second)
- Low latency responses
- Cost-effective pricing
- Support for function calling
- High throughput
Use cases:
- Real-time chat applications
- High-volume API services
- Latency-sensitive applications
- Interactive user experiences"""
},
"features": [
"Ultra-fast inference with LPU technology",
"Support for Llama, Mixtral, and Gemma models",
"Function calling capabilities",
"Streaming responses",
"Low latency (sub-second responses)",
"Cost-effective pricing",
"High throughput capacity"
]
},
"mistral_model": {
"component_name": "MistralModel",
"display_name": "Mistral AI",
"category": "models",
"overview": {
"summary": "European AI with strong multilingual and coding capabilities",
"description": """Mistral AI provides powerful open-source and proprietary models with excellent performance.
Available models:
- Mistral Large: Most capable model
- Mistral Medium: Balanced performance
- Mistral Small: Fast and efficient
- Mixtral 8x7B/8x22B: Mixture-of-experts models
- Codestral: Specialized for code generation
Key strengths:
- Strong multilingual support (especially European languages)
- Excellent coding capabilities
- Function calling support
- JSON mode for structured outputs
- Competitive pricing
- European data sovereignty
Use cases:
- Multilingual applications
- Code generation and review
- European market applications
- Function calling and tool use
- Structured data extraction"""
}
},
"cohere_model": {
"component_name": "CohereModel",
"display_name": "Cohere",
"category": "models",
"overview": {
"summary": "Enterprise-focused AI with strong RAG capabilities",
"description": """Cohere provides enterprise-grade language models optimized for business applications.
Available models:
- Command R+: Most capable for complex tasks
- Command R: Balanced performance
- Command: Fast and efficient
- Command Light: Lightweight option
Key features:
- Excellent RAG (Retrieval Augmented Generation) performance
- Built-in reranking capabilities
- Strong multilingual support (100+ languages)
- Enterprise security and compliance
- Grounded generation with citations
- Function calling support
Use cases:
- Enterprise search and Q&A
- Document analysis with citations
- Multilingual applications
- Customer support automation
- Content generation with sources"""
}
}
}
def create_model_doc(filename, data):
"""Create a comprehensive model documentation file."""
doc_path = DOCS_DIR / filename
# Build complete documentation structure
full_doc = {
"component_name": data["component_name"],
"display_name": data["display_name"],
"category": data["category"],
"version": "1.0.0",
"overview": data["overview"],
"features": data.get("features", []),
"inputs": {},
"outputs": {
"language_model": {
"display_name": "Language Model",
"type": "LanguageModel",
"description": f"Configured {data['display_name']} model ready for use"
}
},
"examples": [],
"troubleshooting": [],
"external_links": []
}
with open(doc_path, 'w') as f:
yaml.dump(full_doc, f, default_flow_style=False, sort_keys=False, allow_unicode=True)
print(f"Created: {filename}")
# Generate all model documentation
for filename, data in MODELS.items():
create_model_doc(f"{filename}.yaml", data)
print(f"\nGenerated {len(MODELS)} model documentation files")

Xet Storage Details

Size:
4.72 kB
·
Xet hash:
6a6cb95bee1fef2210a915cf494be9a64ee9cd59369c98d08a3989f4b19431a4

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.