LLM/DocsGPT
1
0
Fork 0
mirror of https://github.com/arc53/DocsGPT.git synced 2025-11-29 08:33:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: model registry and capabilities for multi-provider support (#2158)

Browse Source 
* feat: Implement model registry and capabilities for multi-provider support

- Added ModelRegistry to manage available models and their capabilities.
- Introduced ModelProvider enum for different LLM providers.
- Created ModelCapabilities dataclass to define model features.
- Implemented methods to load models based on API keys and settings.
- Added utility functions for model management in model_utils.py.
- Updated settings.py to include provider-specific API keys.
- Refactored LLM classes (Anthropic, OpenAI, Google, etc.) to utilize new model registry.
- Enhanced utility functions to handle token limits and model validation.
- Improved code structure and logging for better maintainability.

* feat: Add model selection feature with API integration and UI component

* feat: Add model selection and default model functionality in agent management

* test: Update assertions and formatting in stream processing tests

* refactor(llm): Standardize model identifier to model_id

* fix tests

---------

Co-authored-by: Alex <a@tushynski.me>
This commit is contained in:
Siddhant Rai
2025-11-14 16:43:19 +05:30
committed by GitHub
parent fbf7cf874b
commit 3f7de867cc
54 changed files with 1388 additions and 226 deletions
Download Patch File Download Diff File Expand all files Collapse all files

78
application/utils.py
View File

@@ -7,6 +7,8 @@ import tiktoken
from flask import jsonify, make_response
from werkzeug.utils import secure_filename
from application.core.model_utils import get_token_limit
from application.core.settings import settings
@@ -75,11 +77,9 @@ def count_tokens_docs(docs):
def calculate_doc_token_budget(
gpt_model: str = "gpt-4o", history_token_limit: int = 2000
model_id: str = "gpt-4o", history_token_limit: int = 2000
) -> int:
total_context = settings.LLM_TOKEN_LIMITS.get(
gpt_model, settings.DEFAULT_LLM_TOKEN_LIMIT
)
total_context = get_token_limit(model_id)
reserved = sum(settings.RESERVED_TOKENS.values())
doc_budget = total_context - history_token_limit - reserved
return max(doc_budget, 1000)
@@ -144,16 +144,13 @@ def get_hash(data):
return hashlib.md5(data.encode(), usedforsecurity=False).hexdigest()
def limit_chat_history(history, max_token_limit=None, gpt_model="docsgpt"):
def limit_chat_history(history, max_token_limit=None, model_id="docsgpt-local"):
"""Limit chat history to fit within token limit."""
from application.core.settings import settings
model_token_limit = get_token_limit(model_id)
max_token_limit = (
max_token_limit
if max_token_limit
and max_token_limit
< settings.LLM_TOKEN_LIMITS.get(gpt_model, settings.DEFAULT_LLM_TOKEN_LIMIT)
else settings.LLM_TOKEN_LIMITS.get(gpt_model, settings.DEFAULT_LLM_TOKEN_LIMIT)
if max_token_limit and max_token_limit < model_token_limit
else model_token_limit
)
if not history:
@@ -205,37 +202,44 @@ def clean_text_for_tts(text: str) -> str:
clean text for Text-to-Speech processing.
"""
# Handle code blocks and links
text = re.sub(r'```mermaid[\s\S]*?```', ' flowchart, ', text) ## ```mermaid...```
text = re.sub(r'```[\s\S]*?```', ' code block, ', text) ## ```code```
text = re.sub(r'\[([^\]]+)\]\([^\)]+\)', r'\1', text) ## [text](url)
text = re.sub(r'!\[([^\]]*)\]\([^\)]+\)', '', text) ## ![alt](url)
text = re.sub(r"```mermaid[\s\S]*?```", " flowchart, ", text) ## ```mermaid...```
text = re.sub(r"```[\s\S]*?```", " code block, ", text) ## ```code```
text = re.sub(r"\[([^\]]+)\]\([^\)]+\)", r"\1", text) ## [text](url)
text = re.sub(r"!\[([^\]]*)\]\([^\)]+\)", "", text) ## ![alt](url)
# Remove markdown formatting
text = re.sub(r'`([^`]+)`', r'\1', text) ## `code`
text = re.sub(r'\{([^}]*)\}', r' \1 ', text) ## {text}
text = re.sub(r'[{}]', ' ', text) ## unmatched {}
text = re.sub(r'\[([^\]]+)\]', r' \1 ', text) ## [text]
text = re.sub(r'[\[\]]', ' ', text) ## unmatched []
text = re.sub(r'(\*\*|__)(.*?)\1', r'\2', text) ## **bold** __bold__
text = re.sub(r'(\*|_)(.*?)\1', r'\2', text) ## *italic* _italic_
text = re.sub(r'^#{1,6}\s+', '', text, flags=re.MULTILINE) ## # headers
text = re.sub(r'^>\s+', '', text, flags=re.MULTILINE) ## > blockquotes
text = re.sub(r'^[\s]*[-\*\+]\s+', '', text, flags=re.MULTILINE) ## - * + lists
text = re.sub(r'^[\s]*\d+\.\s+', '', text, flags=re.MULTILINE) ## 1. numbered lists
text = re.sub(r'^[\*\-_]{3,}\s*$', '', text, flags=re.MULTILINE) ## --- *** ___ rules
text = re.sub(r'<[^>]*>', '', text) ## <html> tags
#Remove non-ASCII (emojis, special Unicode)
text = re.sub(r'[^\x20-\x7E\n\r\t]', '', text)
text = re.sub(r"`([^`]+)`", r"\1", text) ## `code`
text = re.sub(r"\{([^}]*)\}", r" \1 ", text) ## {text}
text = re.sub(r"[{}]", " ", text) ## unmatched {}
text = re.sub(r"\[([^\]]+)\]", r" \1 ", text) ## [text]
text = re.sub(r"[\[\]]", " ", text) ## unmatched []
text = re.sub(r"(\*\*|__)(.*?)\1", r"\2", text) ## **bold** __bold__
text = re.sub(r"(\*|_)(.*?)\1", r"\2", text) ## *italic* _italic_
text = re.sub(r"^#{1,6}\s+", "", text, flags=re.MULTILINE) ## # headers
text = re.sub(r"^>\s+", "", text, flags=re.MULTILINE) ## > blockquotes
text = re.sub(r"^[\s]*[-\*\+]\s+", "", text, flags=re.MULTILINE) ## - * + lists
text = re.sub(r"^[\s]*\d+\.\s+", "", text, flags=re.MULTILINE) ## 1. numbered lists
text = re.sub(
r"^[\*\-_]{3,}\s*$", "", text, flags=re.MULTILINE
) ## --- *** ___ rules
text = re.sub(r"<[^>]*>", "", text) ## <html> tags
#Replace special sequences
text = re.sub(r'-->', ', ', text) ## -->
text = re.sub(r'<--', ', ', text) ## <--
text = re.sub(r'=>', ', ', text) ## =>
text = re.sub(r'::', ' ', text) ## ::
# Remove non-ASCII (emojis, special Unicode)
#Normalize whitespace
text = re.sub(r'\s+', ' ', text)
text = re.sub(r"[^\x20-\x7E\n\r\t]", "", text)
# Replace special sequences
text = re.sub(r"-->", ", ", text) ## -->
text = re.sub(r"<--", ", ", text) ## <--
text = re.sub(r"=>", ", ", text) ## =>
text = re.sub(r"::", " ", text) ## ::
# Normalize whitespace
text = re.sub(r"\s+", " ", text)
text = text.strip()
return text