Dev branch for the ToolUseAgent #239
Draft
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
|
Based on your review schedule, I'll hold off on reviewing this PR until it's marked as ready for review. If you'd like me to take a look now, comment
|
TLSDC
requested review from
Copilot,
recursix and
amanjaiswal73892
and removed request for
Copilot
There was a problem hiding this comment.
Pull Request Overview
This PR introduces a new ToolUseAgent and updates benchmarking support by replacing references to bgym.Benchmark and bgym.HighLevelActionSetArgs with their counterparts in agentlab.experiments.benchmark. Key changes include updating import statements and type annotations across study, agent, and benchmark files, as well as adding new benchmark configuration and metadata files.
Reviewed Changes
Copilot reviewed 27 out of 29 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| src/agentlab/experiments/study.py | Updated benchmark and import references to the new Benchmark class. |
| src/agentlab/experiments/reproducibility_util.py | Replaced bgym.Benchmark with the new Benchmark type. |
| src/agentlab/experiments/loop.py | Adjusted imports and commented out redundant TapeAgent-related code. |
| src/agentlab/experiments/benchmark/*.py | Added new benchmark configuration, utility, and metadata files. |
| src/agentlab/agents/tool_use_agent/agent.py | Introduced ToolUseAgent and updated benchmark type references. |
| Other agent and agent_configs files | Updated type annotations for benchmark and HighLevelActionSetArgs references. |
Files not reviewed (2)
- src/agentlab/experiments/benchmark/metadata/assistantbench.csv: Language not supported
- src/agentlab/experiments/benchmark/metadata/miniwob.csv: Language not supported
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.
Comes in combination with this bgym PR:
ServiceNow/BrowserGym#340
Description by Korbit AI
What change is being made?
Introduce a new
ToolUseAgentand supporting benchmark data, and replace existing usage ofbgym.Benchmarkandbgym.HighLevelActionSetArgswith the newly definedBenchmarkandHighLevelActionSetArgsfromagentlab.experiments.benchmark.Why are these changes being made?
These changes are being introduced to expand the functionality of the agent system by adding a
ToolUseAgentwhich leverages tool descriptions to perform actions, while also supporting more refined benchmarking capabilities through the newBenchmarkandHighLevelActionSetArgsclasses which allow for more consistent and modular benchmarking configurations. This improves scalability and ease of future adaptations and improvements in the agent's capabilities and testing environments.