38cdacbcb6
<!-- .github/pull_request_template.md --> ## Description Resolve Gemini Adapter issues: 1. resolve embedding batch issue, 2. Resolve slowness because gemini tokenizer was sending word per word to Googles API to count tokens (using OpenAI's local tokenizer to count tokens for Gemini now) 3. Update deprecated library and move to instructor ## Type of Change <!-- Please check the relevant option --> - [x] Bug fix (non-breaking change that fixes an issue) - [ ] New feature (non-breaking change that adds functionality) - [ ] Breaking change (fix or feature that would cause existing functionality to change) - [ ] Documentation update - [ ] Code refactoring - [ ] Performance improvement - [ ] Other (please specify): ## Pre-submission Checklist <!-- Please check all boxes that apply before submitting your PR --> - [ ] **I have tested my changes thoroughly before submitting this PR** - [ ] **This PR contains minimal changes necessary to address the issue/feature** - [ ] My code follows the project's coding standards and style guidelines - [ ] I have added tests that prove my fix is effective or that my feature works - [ ] I have added necessary documentation (if applicable) - [ ] All new and existing tests pass - [ ] I have searched existing PRs to ensure this change hasn't been submitted already - [ ] I have linked any relevant issues in the description - [ ] My commits have clear and descriptive messages ## DCO Affirmation I affirm that all code in every commit of this pull request conforms to the terms of the Topoteretes Developer Certificate of Origin.
69 lines
2.6 KiB
Python
69 lines
2.6 KiB
Python
from typing import Optional
|
|
from functools import lru_cache
|
|
from pydantic_settings import BaseSettings, SettingsConfigDict
|
|
|
|
|
|
class EmbeddingConfig(BaseSettings):
|
|
"""
|
|
Manage configuration settings for embedding operations, including provider, model
|
|
details, API configuration, and tokenizer settings.
|
|
|
|
Public methods:
|
|
- to_dict: Serialize the configuration settings to a dictionary.
|
|
"""
|
|
|
|
embedding_provider: Optional[str] = "openai"
|
|
embedding_model: Optional[str] = "openai/text-embedding-3-large"
|
|
embedding_dimensions: Optional[int] = 3072
|
|
embedding_endpoint: Optional[str] = None
|
|
embedding_api_key: Optional[str] = None
|
|
embedding_api_version: Optional[str] = None
|
|
embedding_max_completion_tokens: Optional[int] = 8191
|
|
embedding_batch_size: Optional[int] = None
|
|
huggingface_tokenizer: Optional[str] = None
|
|
model_config = SettingsConfigDict(env_file=".env", extra="allow")
|
|
|
|
def model_post_init(self, __context) -> None:
|
|
# If embedding batch size is not defined use 2048 as default for OpenAI and 100 for all other embedding models
|
|
if not self.embedding_batch_size and self.embedding_provider.lower() == "openai":
|
|
self.embedding_batch_size = 2048
|
|
elif not self.embedding_batch_size:
|
|
self.embedding_batch_size = 100
|
|
|
|
def to_dict(self) -> dict:
|
|
"""
|
|
Serialize all embedding configuration settings to a dictionary.
|
|
|
|
Returns:
|
|
--------
|
|
|
|
- dict: A dictionary containing the embedding configuration settings.
|
|
"""
|
|
return {
|
|
"embedding_provider": self.embedding_provider,
|
|
"embedding_model": self.embedding_model,
|
|
"embedding_dimensions": self.embedding_dimensions,
|
|
"embedding_endpoint": self.embedding_endpoint,
|
|
"embedding_api_key": self.embedding_api_key,
|
|
"embedding_api_version": self.embedding_api_version,
|
|
"embedding_max_completion_tokens": self.embedding_max_completion_tokens,
|
|
"huggingface_tokenizer": self.huggingface_tokenizer,
|
|
}
|
|
|
|
|
|
@lru_cache
|
|
def get_embedding_config():
|
|
"""
|
|
Retrieve a cached instance of the EmbeddingConfig class.
|
|
|
|
This function returns an instance of EmbeddingConfig with default settings. It uses
|
|
memoization to cache the result, ensuring that subsequent calls return the same instance
|
|
without re-initialization, improving performance and resource utilization.
|
|
|
|
Returns:
|
|
--------
|
|
|
|
- EmbeddingConfig: An instance of EmbeddingConfig containing the embedding
|
|
configuration settings.
|
|
"""
|
|
return EmbeddingConfig()
|