add cli auto-prompt analyze - prompt analyzer in cli #288
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
kxtran
reviewed
Aug 28, 2024
log10/cli/autoprompt.py
Outdated
Comment on lines
48
to
57
| prompt = "\n\n".join([f'{m["role"]}: {m["content"]}' for m in prompt_json if m["role"] in ALLOWED_ROLES]) | ||
| elif isinstance(prompt_json, dict) and "request" in prompt_json and "messages" in prompt_json["request"]: | ||
| # prompt is a log10 completion | ||
| prompt = "\n\n".join( | ||
| [ | ||
| f'{m["role"]}: {m["content"]}' | ||
| for m in prompt_json["request"]["messages"] | ||
| if m["role"] in ALLOWED_ROLES | ||
| ] | ||
| ) |
There was a problem hiding this comment.
we can extract the building prompt in its own function, wdyt?
log10/cli/autoprompt.py
Outdated
| if isinstance(prompt_json, list) and all("role" in item and "content" in item for item in prompt_json): | ||
| # prompt is a list of messages | ||
| prompt = "\n\n".join([f'{m["role"]}: {m["content"]}' for m in prompt_json if m["role"] in ALLOWED_ROLES]) | ||
| elif isinstance(prompt_json, dict) and "request" in prompt_json and "messages" in prompt_json["request"]: |
There was a problem hiding this comment.
nit:
Suggested change
| elif isinstance(prompt_json, dict) and "request" in prompt_json and "messages" in prompt_json["request"]: | |
| elif isinstance(prompt_json, dict) and "messages" in prompt_json.get("request", {}): |
log10/cli/autoprompt.py
Outdated
| from log10.prompt_analyzer import PromptAnalyzer, display_prompt_analyzer_suggestions | ||
|
|
||
|
|
||
| # ignor "tool" and "funciton" roles |
There was a problem hiding this comment.
Suggested change
| # ignor "tool" and "funciton" roles | |
| # ignore "tool" and "function" roles |
| Analyze a prompt or messages and provide suggestions on how to improve it. | ||
|
|
||
| Options: | ||
| -p, --prompt TEXT The prompt to analyze. Provide a string or a file |
There was a problem hiding this comment.
should we have a separate flag for file?
There was a problem hiding this comment.
we are checking the file internally and keeping the option simple for users. Similar to what we have for other cli option, like --ids for benchmark_models cli.
|
|
||
| # try to parse the prompt as a json | ||
| try: | ||
| prompt_json = json.loads(prompt) |
There was a problem hiding this comment.
if prompt is empty after parsing the text or reading from the file, should we exit immediately?
kxtran
approved these changes
Aug 28, 2024
* string * openai chat request messages * log10 completion format
wenzhe-log10
force-pushed
the
wenzhe/eng-1060-cli-auto-prompt
branch
from
ee52421 to
5f9bdb5
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Three input format allowed:
[{"role": "user", "content": "Hello"}, {"role": "assistant", "content": "Hi"}]Doc:
e.g.
log10 auto-prompt analyze -p "this is a test prompt for autoprompt cli"or pass a file
log10 auto-prompt analyze -p test.prompt