ba2de9bb22
<!-- .github/pull_request_template.md --> ## Description Resolve issue noticed by [RyabykinIlya](https://github.com/RyabykinIlya) where too many HuggingFace requests have been sent due to the embedding engine not working as a singleton per config ## 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. --------- Signed-off-by: Ryabykin Ilya <ryabykinia@sibur.ru> Co-authored-by: greshish <ryabykinia@yandex.ru> Co-authored-by: Ryabykin Ilya <ryabykinia@sibur.ru>
35 lines
1.3 KiB
Python
35 lines
1.3 KiB
Python
from typing import Optional
|
|
from functools import lru_cache
|
|
from pydantic_settings import BaseSettings, SettingsConfigDict
|
|
|
|
|
|
class EmbeddingConfig(BaseSettings):
|
|
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_tokens: Optional[int] = 8191
|
|
huggingface_tokenizer: Optional[str] = None
|
|
model_config = SettingsConfigDict(env_file=".env", extra="allow")
|
|
|
|
def to_dict(self) -> dict:
|
|
"""
|
|
Serialize all embedding configuration settings to a dictionary.
|
|
"""
|
|
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_tokens": self.embedding_max_tokens,
|
|
"huggingface_tokenizer": self.huggingface_tokenizer,
|
|
}
|
|
|
|
|
|
@lru_cache
|
|
def get_embedding_config():
|
|
return EmbeddingConfig()
|