How can I retry individual JSON parsing failures with Langchain's RunnableEach?

I am using Langchain's RunnableEach to execute around 25k parallel requests to OpenAI. Each response should be JSON formatted with a boolean flag and a motivation string. However, I recently encountered an error with one input where the LLM returned malformed JSON, causing a JSONDecodeError. My code is as follows:

PYTHONCopied
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.runnables.base import RunnableEach
from langchain_openai import ChatOpenAI

class Classification(BaseModel):
   flag: bool = Field(description="my descr", enum=[True, False])
   answer: str = Field(description="my descr")

llm = ChatOpenAI(temperature=0, model="gpt-4o-mini", api_key=model_key).with_structured_output(
            Classification)

prompt = my_prompt

tagging_chain = prompt | llm
runnable_each = RunnableEach(bound=tagging_chain)
input_list = [{"input": val} for val in mydata]
res = runnable_each.invoke(input_list)

The error message I received is:

Function Classification arguments:

{"flag":false,"motivation":"The passage does not contain any information relevant to products

are not valid JSON. Received JSONDecodeError Unterminated string starting at: line 1 column 34 (char 33)

I understand that the malformed response from the LLM leads to a JSON parsing error. My main questions are:

• Can such errors be handled so that only the individual failing request is impacted, rather than stopping the entire batch?
• In the update, I saw that wrapping the chain with retry logic using .with_retry() (as shown below) might be a solution:

PYTHONCopied
tagging_chain = prompt | llm.with_retry()
runnable_each = RunnableEach(bound=tagging_chain)

Would this retry mechanism apply only to the individual failing request, or would it affect the entire input sequence?

Thanks in advance for your help!