aljaz/Auto-GPT
1
0
Fork 0
mirror of https://github.com/aljazceru/Auto-GPT.git synced 2026-02-01 04:14:24 +01:00
Code Issues Packages Projects Releases Wiki Activity

fix(agent/llm): Include bad response in parse-fix prompt in OpenAIProvider.create_chat_completion

Browse Source 
Apparently I forgot to also append the response that caused the parse error before throwing it back to the LLM and letting it fix its mistake(s).
This commit is contained in:
Reinier van der Leer
2024-02-14 11:17:34 +01:00
parent 88bbdfc7fc
commit 22aba6dd8a
1 changed files with 1 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
autogpts/autogpt/autogpt/core/resource/model_providers/openai.py
View File

@@ -466,6 +466,7 @@ class OpenAIProvider(
extras={"assistant_msg": assistant_msg, "i_attempt": attempts}, extras={"assistant_msg": assistant_msg, "i_attempt": attempts},
) )
if attempts < self._configuration.fix_failed_parse_tries: if attempts < self._configuration.fix_failed_parse_tries:
model_prompt.append(assistant_msg)
model_prompt.append( model_prompt.append(
ChatMessage.system(f"ERROR PARSING YOUR RESPONSE:\n\n{e}") ChatMessage.system(f"ERROR PARSING YOUR RESPONSE:\n\n{e}")
) )