github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-04-28 03:51:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

community[minor]: Allow passing allow_dangerous_deserialization when loading LLM chain (#18894)

Browse Source 
### Issue
Recently, the new `allow_dangerous_deserialization` flag was introduced
for preventing unsafe model deserialization that relies on pickle
without user's notice (#18696). Since then some LLMs like Databricks
requires passing in this flag with true to instantiate the model.

However, this breaks existing functionality to loading such LLMs within
a chain using `load_chain` method, because the underlying loader
function
[load_llm_from_config](f96dd57501/libs/langchain/langchain/chains/loading.py (L40))
 (and load_llm) ignores keyword arguments passed in. 

### Solution
This PR fixes this issue by propagating the
`allow_dangerous_deserialization` argument to the class loader iff the
LLM class has that field.

---------

Co-authored-by: Eugene Yurtsev <eyurtsev@gmail.com>
Co-authored-by: Bagatur <22008038+baskaryan@users.noreply.github.com>
This commit is contained in:
Yuki Watanabe 2024-03-27 00:07:55 +09:00 committed by GitHub
parent d7c14cb6f9
commit cfecbda48b
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 130 additions and 64 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
libs/community/langchain_community/llms/loading.py
View File

@ -1,15 +1,17 @@
"""Base interface for loading large language model APIs."""
import json
from pathlib import Path
from typing import Union
from typing import Any, Union
import yaml
from langchain_core.language_models.llms import BaseLLM
from langchain_community.llms import get_type_to_cls_dict
_ALLOW_DANGEROUS_DESERIALIZATION_ARG = "allow_dangerous_deserialization"
def load_llm_from_config(config: dict) -> BaseLLM:
def load_llm_from_config(config: dict, **kwargs: Any) -> BaseLLM:
"""Load LLM from Config Dict."""
if "_type" not in config:
raise ValueError("Must specify an LLM Type in config")
@ -21,11 +23,17 @@ def load_llm_from_config(config: dict) -> BaseLLM:
raise ValueError(f"Loading {config_type} LLM not supported")
llm_cls = type_to_cls_dict[config_type]()
return llm_cls(**config)
load_kwargs = {}
if _ALLOW_DANGEROUS_DESERIALIZATION_ARG in llm_cls.__fields__:
load_kwargs[_ALLOW_DANGEROUS_DESERIALIZATION_ARG] = kwargs.get(
_ALLOW_DANGEROUS_DESERIALIZATION_ARG, False
)
return llm_cls(**config, **load_kwargs)
def load_llm(file: Union[str, Path]) -> BaseLLM:
"""Load LLM from file."""
def load_llm(file: Union[str, Path], **kwargs: Any) -> BaseLLM:
# Convert file to Path object.
if isinstance(file, str):
file_path = Path(file)
@ -41,4 +49,4 @@ def load_llm(file: Union[str, Path]) -> BaseLLM:
else:
raise ValueError("File type must be json or yaml")
# Load the LLM from the config now.
return load_llm_from_config(config)
return load_llm_from_config(config, **kwargs)

26
libs/community/tests/unit_tests/llms/test_databricks.py
View File

@ -1,4 +1,5 @@
"""test Databricks LLM"""
from pathlib import Path
from typing import Any, Dict
import pytest
@ -8,6 +9,8 @@ from langchain_community.llms.databricks import (
Databricks,
_load_pickled_fn_from_hex_string,
)
from langchain_community.llms.loading import load_llm
from tests.integration_tests.llms.utils import assert_llm_equality
class MockDatabricksServingEndpointClient:
@ -55,3 +58,26 @@ def test_serde_transform_input_fn(monkeypatch: MonkeyPatch) -> None:
request = {"prompt": "What is the meaning of life?"}
fn = _load_pickled_fn_from_hex_string(params["transform_input_fn"])
assert fn(**request) == transform_input(**request)
def test_saving_loading_llm(monkeypatch: MonkeyPatch, tmp_path: Path) -> None:
monkeypatch.setattr(
"langchain_community.llms.databricks._DatabricksServingEndpointClient",
MockDatabricksServingEndpointClient,
)
monkeypatch.setenv("DATABRICKS_HOST", "my-default-host")
monkeypatch.setenv("DATABRICKS_TOKEN", "my-default-token")
llm = Databricks(
endpoint_name="chat", temperature=0.1, allow_dangerous_deserialization=True
)
llm.save(file_path=tmp_path / "databricks.yaml")
# Loading without allowing_dangerous_deserialization=True should raise an error.
with pytest.raises(ValueError, match="This code relies on the pickle module."):
load_llm(tmp_path / "databricks.yaml")
loaded_llm = load_llm(
tmp_path / "databricks.yaml", allow_dangerous_deserialization=True
)
assert_llm_equality(llm, loaded_llm)

148
libs/langchain/langchain/chains/loading.py
View File

@ -37,9 +37,9 @@ def _load_llm_chain(config: dict, **kwargs: Any) -> LLMChain:
"""Load LLM chain from config dict."""
if "llm" in config:
llm_config = config.pop("llm")
llm = load_llm_from_config(llm_config)
llm = load_llm_from_config(llm_config, **kwargs)
elif "llm_path" in config:
llm = load_llm(config.pop("llm_path"))
llm = load_llm(config.pop("llm_path"), **kwargs)
else:
raise ValueError("One of `llm` or `llm_path` must be present.")
@ -59,9 +59,9 @@ def _load_hyde_chain(config: dict, **kwargs: Any) -> HypotheticalDocumentEmbedde
"""Load hypothetical document embedder chain from config dict."""
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
if "embeddings" in kwargs:
@ -78,9 +78,9 @@ def _load_hyde_chain(config: dict, **kwargs: Any) -> HypotheticalDocumentEmbedde
def _load_stuff_documents_chain(config: dict, **kwargs: Any) -> StuffDocumentsChain:
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
@ -107,9 +107,9 @@ def _load_map_reduce_documents_chain(
) -> MapReduceDocumentsChain:
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
@ -118,12 +118,14 @@ def _load_map_reduce_documents_chain(
if "reduce_documents_chain" in config:
reduce_documents_chain = load_chain_from_config(
config.pop("reduce_documents_chain")
config.pop("reduce_documents_chain"), **kwargs
)
elif "reduce_documents_chain_path" in config:
reduce_documents_chain = load_chain(config.pop("reduce_documents_chain_path"))
reduce_documents_chain = load_chain(
config.pop("reduce_documents_chain_path"), **kwargs
)
else:
reduce_documents_chain = _load_reduce_documents_chain(config)
reduce_documents_chain = _load_reduce_documents_chain(config, **kwargs)
return MapReduceDocumentsChain(
llm_chain=llm_chain,
@ -138,14 +140,22 @@ def _load_reduce_documents_chain(config: dict, **kwargs: Any) -> ReduceDocuments
if "combine_documents_chain" in config:
combine_document_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_document_chain_config)
combine_documents_chain = load_chain_from_config(
combine_document_chain_config, **kwargs
)
elif "combine_document_chain" in config:
combine_document_chain_config = config.pop("combine_document_chain")
combine_documents_chain = load_chain_from_config(combine_document_chain_config)
combine_documents_chain = load_chain_from_config(
combine_document_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
elif "combine_document_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_document_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_document_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -158,11 +168,11 @@ def _load_reduce_documents_chain(config: dict, **kwargs: Any) -> ReduceDocuments
collapse_documents_chain = None
else:
collapse_documents_chain = load_chain_from_config(
collapse_document_chain_config
collapse_document_chain_config, **kwargs
)
elif "collapse_documents_chain_path" in config:
collapse_documents_chain = load_chain(
config.pop("collapse_documents_chain_path")
config.pop("collapse_documents_chain_path"), **kwargs
)
elif "collapse_document_chain" in config:
collapse_document_chain_config = config.pop("collapse_document_chain")
@ -170,11 +180,11 @@ def _load_reduce_documents_chain(config: dict, **kwargs: Any) -> ReduceDocuments
collapse_documents_chain = None
else:
collapse_documents_chain = load_chain_from_config(
collapse_document_chain_config
collapse_document_chain_config, **kwargs
)
elif "collapse_document_chain_path" in config:
collapse_documents_chain = load_chain(
config.pop("collapse_document_chain_path")
config.pop("collapse_document_chain_path"), **kwargs
)
return ReduceDocumentsChain(
@ -190,17 +200,17 @@ def _load_llm_bash_chain(config: dict, **kwargs: Any) -> Any:
llm_chain = None
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
# llm attribute is deprecated in favor of llm_chain, here to support old configs
elif "llm" in config:
llm_config = config.pop("llm")
llm = load_llm_from_config(llm_config)
llm = load_llm_from_config(llm_config, **kwargs)
# llm_path attribute is deprecated in favor of llm_chain_path,
# its to support old configs
elif "llm_path" in config:
llm = load_llm(config.pop("llm_path"))
llm = load_llm(config.pop("llm_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
if "prompt" in config:
@ -217,9 +227,9 @@ def _load_llm_bash_chain(config: dict, **kwargs: Any) -> Any:
def _load_llm_checker_chain(config: dict, **kwargs: Any) -> LLMCheckerChain:
if "llm" in config:
llm_config = config.pop("llm")
llm = load_llm_from_config(llm_config)
llm = load_llm_from_config(llm_config, **kwargs)
elif "llm_path" in config:
llm = load_llm(config.pop("llm_path"))
llm = load_llm(config.pop("llm_path"), **kwargs)
else:
raise ValueError("One of `llm` or `llm_path` must be present.")
if "create_draft_answer_prompt" in config:
@ -264,17 +274,17 @@ def _load_llm_math_chain(config: dict, **kwargs: Any) -> LLMMathChain:
llm_chain = None
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
# llm attribute is deprecated in favor of llm_chain, here to support old configs
elif "llm" in config:
llm_config = config.pop("llm")
llm = load_llm_from_config(llm_config)
llm = load_llm_from_config(llm_config, **kwargs)
# llm_path attribute is deprecated in favor of llm_chain_path,
# its to support old configs
elif "llm_path" in config:
llm = load_llm(config.pop("llm_path"))
llm = load_llm(config.pop("llm_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
if "prompt" in config:
@ -293,9 +303,9 @@ def _load_map_rerank_documents_chain(
) -> MapRerankDocumentsChain:
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
return MapRerankDocumentsChain(llm_chain=llm_chain, **config) # type: ignore[arg-type]
@ -306,9 +316,9 @@ def _load_pal_chain(config: dict, **kwargs: Any) -> Any:
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
return PALChain(llm_chain=llm_chain, **config) # type: ignore[arg-type]
@ -317,18 +327,18 @@ def _load_pal_chain(config: dict, **kwargs: Any) -> Any:
def _load_refine_documents_chain(config: dict, **kwargs: Any) -> RefineDocumentsChain:
if "initial_llm_chain" in config:
initial_llm_chain_config = config.pop("initial_llm_chain")
initial_llm_chain = load_chain_from_config(initial_llm_chain_config)
initial_llm_chain = load_chain_from_config(initial_llm_chain_config, **kwargs)
elif "initial_llm_chain_path" in config:
initial_llm_chain = load_chain(config.pop("initial_llm_chain_path"))
initial_llm_chain = load_chain(config.pop("initial_llm_chain_path"), **kwargs)
else:
raise ValueError(
"One of `initial_llm_chain` or `initial_llm_chain_path` must be present."
)
if "refine_llm_chain" in config:
refine_llm_chain_config = config.pop("refine_llm_chain")
refine_llm_chain = load_chain_from_config(refine_llm_chain_config)
refine_llm_chain = load_chain_from_config(refine_llm_chain_config, **kwargs)
elif "refine_llm_chain_path" in config:
refine_llm_chain = load_chain(config.pop("refine_llm_chain_path"))
refine_llm_chain = load_chain(config.pop("refine_llm_chain_path"), **kwargs)
else:
raise ValueError(
"One of `refine_llm_chain` or `refine_llm_chain_path` must be present."
@ -349,9 +359,13 @@ def _load_refine_documents_chain(config: dict, **kwargs: Any) -> RefineDocuments
def _load_qa_with_sources_chain(config: dict, **kwargs: Any) -> QAWithSourcesChain:
if "combine_documents_chain" in config:
combine_documents_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_documents_chain_config)
combine_documents_chain = load_chain_from_config(
combine_documents_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -369,13 +383,13 @@ def _load_sql_database_chain(config: dict, **kwargs: Any) -> Any:
raise ValueError("`database` must be present.")
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
chain = load_chain_from_config(llm_chain_config)
return SQLDatabaseChain(llm_chain=chain, database=database, **config) # type: ignore[arg-type]
chain = load_chain_from_config(llm_chain_config, **kwargs, **kwargs)
return SQLDatabaseChain(llm_chain=chain, database=database, **config)
if "llm" in config:
llm_config = config.pop("llm")
llm = load_llm_from_config(llm_config)
llm = load_llm_from_config(llm_config, **kwargs)
elif "llm_path" in config:
llm = load_llm(config.pop("llm_path"))
llm = load_llm(config.pop("llm_path"), **kwargs)
else:
raise ValueError("One of `llm` or `llm_path` must be present.")
if "prompt" in config:
@ -396,9 +410,13 @@ def _load_vector_db_qa_with_sources_chain(
raise ValueError("`vectorstore` must be present.")
if "combine_documents_chain" in config:
combine_documents_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_documents_chain_config)
combine_documents_chain = load_chain_from_config(
combine_documents_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -418,9 +436,13 @@ def _load_retrieval_qa(config: dict, **kwargs: Any) -> RetrievalQA:
raise ValueError("`retriever` must be present.")
if "combine_documents_chain" in config:
combine_documents_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_documents_chain_config)
combine_documents_chain = load_chain_from_config(
combine_documents_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -442,9 +464,13 @@ def _load_retrieval_qa_with_sources_chain(
raise ValueError("`retriever` must be present.")
if "combine_documents_chain" in config:
combine_documents_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_documents_chain_config)
combine_documents_chain = load_chain_from_config(
combine_documents_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -464,9 +490,13 @@ def _load_vector_db_qa(config: dict, **kwargs: Any) -> VectorDBQA:
raise ValueError("`vectorstore` must be present.")
if "combine_documents_chain" in config:
combine_documents_chain_config = config.pop("combine_documents_chain")
combine_documents_chain = load_chain_from_config(combine_documents_chain_config)
combine_documents_chain = load_chain_from_config(
combine_documents_chain_config, **kwargs
)
elif "combine_documents_chain_path" in config:
combine_documents_chain = load_chain(config.pop("combine_documents_chain_path"))
combine_documents_chain = load_chain(
config.pop("combine_documents_chain_path"), **kwargs
)
else:
raise ValueError(
"One of `combine_documents_chain` or "
@ -486,12 +516,14 @@ def _load_graph_cypher_chain(config: dict, **kwargs: Any) -> GraphCypherQAChain:
raise ValueError("`graph` must be present.")
if "cypher_generation_chain" in config:
cypher_generation_chain_config = config.pop("cypher_generation_chain")
cypher_generation_chain = load_chain_from_config(cypher_generation_chain_config)
cypher_generation_chain = load_chain_from_config(
cypher_generation_chain_config, **kwargs
)
else:
raise ValueError("`cypher_generation_chain` must be present.")
if "qa_chain" in config:
qa_chain_config = config.pop("qa_chain")
qa_chain = load_chain_from_config(qa_chain_config)
qa_chain = load_chain_from_config(qa_chain_config, **kwargs)
else:
raise ValueError("`qa_chain` must be present.")
@ -506,7 +538,7 @@ def _load_graph_cypher_chain(config: dict, **kwargs: Any) -> GraphCypherQAChain:
def _load_api_chain(config: dict, **kwargs: Any) -> APIChain:
if "api_request_chain" in config:
api_request_chain_config = config.pop("api_request_chain")
api_request_chain = load_chain_from_config(api_request_chain_config)
api_request_chain = load_chain_from_config(api_request_chain_config, **kwargs)
elif "api_request_chain_path" in config:
api_request_chain = load_chain(config.pop("api_request_chain_path"))
else:
@ -515,9 +547,9 @@ def _load_api_chain(config: dict, **kwargs: Any) -> APIChain:
)
if "api_answer_chain" in config:
api_answer_chain_config = config.pop("api_answer_chain")
api_answer_chain = load_chain_from_config(api_answer_chain_config)
api_answer_chain = load_chain_from_config(api_answer_chain_config, **kwargs)
elif "api_answer_chain_path" in config:
api_answer_chain = load_chain(config.pop("api_answer_chain_path"))
api_answer_chain = load_chain(config.pop("api_answer_chain_path"), **kwargs)
else:
raise ValueError(
"One of `api_answer_chain` or `api_answer_chain_path` must be present."
@ -537,9 +569,9 @@ def _load_api_chain(config: dict, **kwargs: Any) -> APIChain:
def _load_llm_requests_chain(config: dict, **kwargs: Any) -> LLMRequestsChain:
if "llm_chain" in config:
llm_chain_config = config.pop("llm_chain")
llm_chain = load_chain_from_config(llm_chain_config)
llm_chain = load_chain_from_config(llm_chain_config, **kwargs)
elif "llm_chain_path" in config:
llm_chain = load_chain(config.pop("llm_chain_path"))
llm_chain = load_chain(config.pop("llm_chain_path"), **kwargs)
else:
raise ValueError("One of `llm_chain` or `llm_chain_path` must be present.")
if "requests_wrapper" in kwargs: