feat: Add screenshot downsampling and refine LLM config/calls #24
Conversation
This file contains hidden or 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
- Adds OMNIPARSER_DOWNSAMPLE_FACTOR config variable to allow scaling screenshots before parsing (default 1.0). Implemented via new utils.downsample_image function called from visual_state.py. - Adds LLM_PROVIDER and ANTHROPIC_DEFAULT_MODEL config variables. - Updates completions.py to use config, add tenacity retries, and restore/use format_chat_messages (guarded by new DEBUG_FULL_PROMPTS config flag). - Fixes test failures in test_visual_state.py related to mock data and patches. Agent planning sequence for calculator task now appears correct with Sonnet and downsample_factor=1.0, although perception (~7-11s) and planning (~5-9s) steps remain slow. Downsampling factor < 1.0 can be enabled via .env for performance testing, but may affect accuracy.
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Outdated suggestions cannot be applied.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.
screenshots before parsing (default 1.0). Implemented via new
utils.downsample_image function called from visual_state.py.
restore/use format_chat_messages (guarded by new DEBUG_FULL_PROMPTS config flag).
Agent planning sequence for calculator task now appears correct with Sonnet
and downsample_factor=1.0, although perception (~7-11s) and planning
(~5-9s) steps remain slow. Downsampling factor < 1.0 can be enabled via
.env for performance testing, but may affect accuracy.