Serialize all the chains!

langchain/chains/api/base.py

@@ -3,7 +3,7 @@
 
 from typing import Any, Dict, List, Optional
 
-from pydantic import BaseModel, root_validator
+from pydantic import BaseModel, Field, root_validator
 
 from langchain.chains.api.prompt import API_RESPONSE_PROMPT, API_URL_PROMPT
 from langchain.chains.base import Chain
@@ -18,7 +18,7 @@ class APIChain(Chain, BaseModel):
 
     api_request_chain: LLMChain
     api_answer_chain: LLMChain
-    requests_wrapper: RequestsWrapper
+    requests_wrapper: RequestsWrapper = Field(exclude=True)
     api_docs: str
     question_key: str = "question"  #: :meta private:
     output_key: str = "output"  #: :meta private:
@@ -102,3 +102,7 @@ def from_llm_and_api_docs(
             api_docs=api_docs,
             **kwargs,
         )
+
+    @property
+    def _chain_type(self) -> str:
+        return "api_chain"

langchain/chains/combine_documents/map_reduce.py

@@ -168,3 +168,7 @@ def combine_docs(
             extra_return_dict = {}
         output, _ = self.combine_document_chain.combine_docs(result_docs, **kwargs)
         return output, extra_return_dict
+
+    @property
+    def _chain_type(self) -> str:
+        return "map_reduce_documents_chain"

langchain/chains/combine_documents/map_rerank.py

@@ -111,3 +111,7 @@ def combine_docs(self, docs: List[Document], **kwargs: Any) -> Tuple[str, dict]:
         if self.return_intermediate_steps:
             extra_info["intermediate_steps"] = results
         return output[self.answer_key], extra_info
+
+    @property
+    def _chain_type(self) -> str:
+        return "map_rerank_documents_chain"

langchain/chains/combine_documents/refine.py

@@ -113,3 +113,7 @@ def combine_docs(self, docs: List[Document], **kwargs: Any) -> Tuple[str, dict]:
         else:
             extra_return_dict = {}
         return res, extra_return_dict
+
+    @property
+    def _chain_type(self) -> str:
+        return "refine_documents_chain"

langchain/chains/combine_documents/stuff.py

@@ -83,3 +83,7 @@ def combine_docs(self, docs: List[Document], **kwargs: Any) -> Tuple[str, dict]:
         inputs = self._get_inputs(docs, **kwargs)
         # Call predict on the LLM.
         return self.llm_chain.predict(**inputs), {}
+
+    @property
+    def _chain_type(self) -> str:
+        return "stuff_documents_chain"

langchain/chains/hyde/base.py

@@ -69,3 +69,7 @@ def from_llm(
         prompt = PROMPT_MAP[prompt_key]
         llm_chain = LLMChain(llm=llm, prompt=prompt)
         return cls(base_embeddings=base_embeddings, llm_chain=llm_chain)
+
+    @property
+    def _chain_type(self) -> str:
+        return "hyde_chain"

langchain/chains/llm_bash/base.py

@@ -73,3 +73,7 @@ def _call(self, inputs: Dict[str, str]) -> Dict[str, str]:
         else:
             raise ValueError(f"unknown format from LLM: {t}")
         return {self.output_key: output}
+
+    @property
+    def _chain_type(self) -> str:
+        return "llm_bash_chain"

langchain/chains/llm_checker/base.py

@@ -97,3 +97,7 @@ def _call(self, inputs: Dict[str, str]) -> Dict[str, str]:
         )
         output = question_to_checked_assertions_chain({"question": question})
         return {self.output_key: output["revised_statement"]}
+
+    @property
+    def _chain_type(self) -> str:
+        return "llm_checker_chain"

langchain/chains/llm_math/base.py

@@ -68,3 +68,7 @@ def _call(self, inputs: Dict[str, str]) -> Dict[str, str]:
         else:
             raise ValueError(f"unknown format from LLM: {t}")
         return {self.output_key: answer}
+
+    @property
+    def _chain_type(self) -> str:
+        return "llm_math_chain"

langchain/chains/llm_requests.py

@@ -18,7 +18,9 @@ class LLMRequestsChain(Chain, BaseModel):
     """Chain that hits a URL and then uses an LLM to parse results."""
 
     llm_chain: LLMChain
-    requests_wrapper: RequestsWrapper = Field(default_factory=RequestsWrapper)
+    requests_wrapper: RequestsWrapper = Field(
+        default_factory=RequestsWrapper, exclude=True
+    )
     text_length: int = 8000
     requests_key: str = "requests_result"  #: :meta private:
     input_key: str = "url"  #: :meta private:
@@ -71,3 +73,7 @@ def _call(self, inputs: Dict[str, str]) -> Dict[str, str]:
         other_keys[self.requests_key] = soup.get_text()[: self.text_length]
         result = self.llm_chain.predict(**other_keys)
         return {self.output_key: result}
+
+    @property
+    def _chain_type(self) -> str:
+        return "llm_requests_chain"