From d4fbe990520fd079dc99e8120a0d08ff1076ef69 Mon Sep 17 00:00:00 2001 From: KCaverly Date: Tue, 12 Sep 2023 21:27:35 -0400 Subject: [PATCH] add eval for gpt-engineer --- crates/semantic_index/eval/gpt-engineer.json | 114 +++++++++++++++++++ 1 file changed, 114 insertions(+) create mode 100644 crates/semantic_index/eval/gpt-engineer.json diff --git a/crates/semantic_index/eval/gpt-engineer.json b/crates/semantic_index/eval/gpt-engineer.json new file mode 100644 index 0000000000000000000000000000000000000000..d7c08cd505bda5f70bc6bdb05c967bc005522b97 --- /dev/null +++ b/crates/semantic_index/eval/gpt-engineer.json @@ -0,0 +1,114 @@ +{ + "repo": "https://github.com/AntonOsika/gpt-engineer.git", + "commit": "7735a6445bae3611c62f521e6464c67c957f87c2", + "assertions": [ + { + "query": "How do I contribute to this project?", + "matches": [ + ".github/CONTRIBUTING.md:1", + "ROADMAP.md:48" + ] + }, + { + "query": "What version of the openai package is active?", + "matches": [ + "pyproject.toml:14" + ] + }, + { + "query": "Ask user for clarification", + "matches": [ + "gpt-engineer/steps.py:69" + ] + }, + { + "query": "generate tests for python code", + "matches": [ + "gpt-engineer/steps.py:153" + ] + }, + { + "query": "get item from database based on key", + "matches": [ + "gpt-engineer/db.py:42", + "gpt-engineer/db.py:68" + ] + }, + { + "query": "prompt user to select files", + "matches": [ + "gpt-engineer/file_selector.py:171", + "gpt-engineer/file_selector.py:306", + "gpt-engineer/file_selector.py:289", + "gpt-engineer/file_selector.py:234" + ] + }, + { + "query": "send to rudderstack", + "matches": [ + "gpt-engineer/collect.py:11", + "gpt-engineer/collect.py:38" + ] + }, + { + "query": "parse code blocks from chat messages", + "matches": [ + "gpt-engineer/chat_to_files.py:10", + "docs/intro/chat_parsing.md:1" + ] + }, + { + "query": "how do I use the docker cli?", + "matches": [ + "docker/README.md:1" + ] + }, + { + "query": "ask the user if the code ran successfully?", + "matches": [ + "gpt-engineer/learning.py:54" + ] + }, + { + "query": "how is consent granted by the user?", + "matches": [ + "gpt-engineer/learning.py:107", + "gpt-engineer/learning.py:130", + "gpt-engineer/learning.py:152" + ] + }, + { + "query": "what are all the different steps the agent can take?", + "matches": [ + "docs/intro/steps_module.md:1", + "gpt-engineer/steps.py:391" + ] + }, + { + "query": "ask the user for clarification?", + "matches": [ + "gpt-engineer/steps.py:69" + ] + }, + { + "query": "what models are available?", + "matches": [ + "gpt-engineer/ai.py:315", + "gpt-engineer/ai.py:341", + "docs/open-models.md:1" + ] + }, + { + "query": "what is the current focus of the project?", + "matches": [ + "ROADMAP.md:11" + ] + }, + { + "query": "does the agent know how to fix code?", + "matches": [ + "gpt-engineer/steps.py:367" + ] + } + ] +}