gmakstutis/LightRAG
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Add mandatory dimension parameter handling for Jina API compliance

Browse source
This commit is contained in:
yangdx 2025-11-07 21:08:34 +08:00
parent d8a6355e41
commit c14f25b7f8
3 changed files with 36 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
env.example
View file

@ -244,9 +244,10 @@ OLLAMA_LLM_NUM_CTX=32768
# EMBEDDING_TIMEOUT=30 # EMBEDDING_TIMEOUT=30
### Control whether to send embedding_dim parameter to embedding API ### Control whether to send embedding_dim parameter to embedding API
### Set to 'true' to enable dynamic dimension adjustment (only works for OpenAI and Jina) ### IMPORTANT: Jina ALWAYS sends dimension parameter (API requirement) - this setting is ignored for Jina
### Set to 'false' (default) to disable sending dimension parameter ### For OpenAI: Set to 'true' to enable dynamic dimension adjustment
### Note: This is automatically ignored for backends that don't support dimension parameter ### For OpenAI: Set to 'false' (default) to disable sending dimension parameter
### Note: Automatically ignored for backends that don't support dimension parameter (e.g., Ollama)
# EMBEDDING_SEND_DIM=false # EMBEDDING_SEND_DIM=false
EMBEDDING_BINDING=ollama EMBEDDING_BINDING=ollama

36
lightrag/api/lightrag_server.py
View file

@ -643,9 +643,7 @@ def create_app(args):
elif binding == "jina": elif binding == "jina":
from lightrag.llm.jina import jina_embed from lightrag.llm.jina import jina_embed
return await jina_embed( return await jina_embed(texts, base_url=host, api_key=api_key)
texts, base_url=host, api_key=api_key
)
else: # openai and compatible else: # openai and compatible
from lightrag.llm.openai import openai_embed from lightrag.llm.openai import openai_embed
@ -691,7 +689,7 @@ def create_app(args):
# Create embedding function with optimized configuration # Create embedding function with optimized configuration
import inspect import inspect
# Create the optimized embedding function # Create the optimized embedding function
optimized_embedding_func = create_optimized_embedding_function( optimized_embedding_func = create_optimized_embedding_function(
config_cache=config_cache, config_cache=config_cache,
@ -701,27 +699,33 @@ def create_app(args):
api_key=args.embedding_binding_api_key, api_key=args.embedding_binding_api_key,
args=args, # Pass args object for fallback option generation args=args, # Pass args object for fallback option generation
) )
# Check environment variable for sending dimensions # Check environment variable for sending dimensions
embedding_send_dim = os.getenv("EMBEDDING_SEND_DIM", "false").lower() == "true" embedding_send_dim = os.getenv("EMBEDDING_SEND_DIM", "false").lower() == "true"
# Check if the function signature has embedding_dim parameter # Check if the function signature has embedding_dim parameter
# Note: Since optimized_embedding_func is an async function, inspect its signature # Note: Since optimized_embedding_func is an async function, inspect its signature
sig = inspect.signature(optimized_embedding_func) sig = inspect.signature(optimized_embedding_func)
has_embedding_dim_param = 'embedding_dim' in sig.parameters has_embedding_dim_param = "embedding_dim" in sig.parameters
# Determine send_dimensions value # Determine send_dimensions value based on binding type
# Only send dimensions if both conditions are met: # Jina REQUIRES dimension parameter (forced to True)
# 1. EMBEDDING_SEND_DIM environment variable is true # OpenAI and others: controlled by EMBEDDING_SEND_DIM environment variable
# 2. The function has embedding_dim parameter if args.embedding_binding == "jina":
send_dimensions = embedding_send_dim and has_embedding_dim_param # Jina API requires dimension parameter - always send it
send_dimensions = has_embedding_dim_param
dimension_control = "forced (Jina API requirement)"
else:
# For OpenAI and other bindings, respect EMBEDDING_SEND_DIM setting
send_dimensions = embedding_send_dim and has_embedding_dim_param
dimension_control = f"env_var={embedding_send_dim}"
logger.info( logger.info(
f"Embedding configuration: send_dimensions={send_dimensions} " f"Embedding configuration: send_dimensions={send_dimensions} "
f"(env_var={embedding_send_dim}, has_param={has_embedding_dim_param}, " f"({dimension_control}, has_param={has_embedding_dim_param}, "
f"binding={args.embedding_binding})" f"binding={args.embedding_binding})"
) )
# Create EmbeddingFunc with send_dimensions attribute # Create EmbeddingFunc with send_dimensions attribute
embedding_func = EmbeddingFunc( embedding_func = EmbeddingFunc(
embedding_dim=args.embedding_dim, embedding_dim=args.embedding_dim,

19
lightrag/utils.py
View file

@ -353,24 +353,29 @@ class EmbeddingFunc:
embedding_dim: int embedding_dim: int
func: callable func: callable
max_token_size: int | None = None # deprecated keep it for compatible only max_token_size: int | None = None # deprecated keep it for compatible only
send_dimensions: bool = False # Control whether to send embedding_dim to the function send_dimensions: bool = (
False # Control whether to send embedding_dim to the function
)
async def __call__(self, *args, **kwargs) -> np.ndarray: async def __call__(self, *args, **kwargs) -> np.ndarray:
# Only inject embedding_dim when send_dimensions is True # Only inject embedding_dim when send_dimensions is True
if self.send_dimensions: if self.send_dimensions:
# Check if user provided embedding_dim parameter # Check if user provided embedding_dim parameter
if 'embedding_dim' in kwargs: if "embedding_dim" in kwargs:
user_provided_dim = kwargs['embedding_dim'] user_provided_dim = kwargs["embedding_dim"]
# If user's value differs from class attribute, output warning # If user's value differs from class attribute, output warning
if user_provided_dim is not None and user_provided_dim != self.embedding_dim: if (
user_provided_dim is not None
and user_provided_dim != self.embedding_dim
):
logger.warning( logger.warning(
f"Ignoring user-provided embedding_dim={user_provided_dim}, " f"Ignoring user-provided embedding_dim={user_provided_dim}, "
f"using declared embedding_dim={self.embedding_dim} from decorator" f"using declared embedding_dim={self.embedding_dim} from decorator"
) )
# Inject embedding_dim from decorator # Inject embedding_dim from decorator
kwargs['embedding_dim'] = self.embedding_dim kwargs["embedding_dim"] = self.embedding_dim
return await self.func(*args, **kwargs) return await self.func(*args, **kwargs)