* update and polish gptq eval
* fix typo
* Update evaluation/gpqa/README.md
Co-authored-by: Graham Neubig <neubig@gmail.com>
* Update evaluation/gpqa/run_infer.py
Co-authored-by: Graham Neubig <neubig@gmail.com>
* add headless mode to all appropriate agent controller call
* delegate set to error when in headless mode
* try to deduplicate a bit
* make headless_mode default to True and only change it to false for AgentSession
---------
Co-authored-by: Graham Neubig <neubig@gmail.com>
* Updated documentation using ruff's autofix feature
* Updated pyproject.toml to include docstring validations
* Updated documentation using ruff's autofix feature
* Updated pyproject.toml to include docstring validations
* Updated docstrings using ruff's autfix feature
* Deleted opendevin/runtime/utils/soource.py, Keeping in sync with main
---------
Co-authored-by: Graham Neubig <neubig@gmail.com>
Currently, OpenDevin uses a global singleton LLM config and a global singleton agent config. This PR allows customers to configure an LLM config for each agent. A hypothetically useful scenario is to use a cheaper LLM for repo exploration / code search, and a more powerful LLM to actually do the problem solving (CodeActAgent).
Partially solves #2075 (web GUI improvement is not the goal of this PR)
* add event to stream before budget check
* make the budget check before the step
* Update opendevin/controller/agent_controller.py
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
---------
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
1. Add support for rejection action on frontend
2. Show users the reason for rejection
3. Get rid of weird empty box after delegation
4. On web GUI, show customer when a delegation starts and ends
* Fix AgentRejectAction handling
* Add ManagerAgent to integration tests
* Fix regenerate.sh
* Fix merge
* Update README for micro-agents
* Add test reject to regenerate.sh
* regenerate.sh: Add support for running a specific test and/or agent
* Refine reject schema, and allow ManagerAgent to handle reject
* Add test artifacts for test_simple_task_rejection
* Fix manager agent tests
* Fix README
* test_simple_task_rejection: check final agent state
* Integration test: exit if mock prompt not found
* Update test_simple_task_rejection tests
* Fix test_edits test artifacts after prompt update
* Fix ManagerAgent test_edits
* WIP
* Fix tests
* update test_edits for ManagerAgent
* Skip local sandbox for reject test
* Fix test comparison
This PR fixes#1897. In addition, this PR fixes and tweaks a few micro-agents.
For the first time, I am able to use ManagerAgent to complete test_write_simple_script and test_edits tasks in integration tests, so this PR also adds ManagerAgent as part of integration tests. test_write_simple_script involves delegation to CoderAgent while test_edits involves delegation to TypoFixerAgent.
Also for the first time, I am able to use DelegateAgent to complete test_write_simple_script and test_edits tasks in integration tests, so this PR also adds DelegateAgent as part of integration tests. It involves delegation to StudyRepoForTaskAgent, CoderAgent and VerifierAgent.
This PR is a blocker for #1735 and likely #1945.
* feat: add max_budget_per_task configuration to control task cost
* Fix test_arg_parser.py
* Use the config.max_budget_per_task as default value
* Add max_budget_per_task to core/main.py as well
* Update opendevin/controller/agent_controller.py
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
---------
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
* properly log user messages;
format browser action/obs, summarize action, messages properly for logging
* add source to message
* add spaces for printing
* Refactor monologue to use the messages in state history
* add messages, clean up
* fix monologue
* update integration tests
* move private method
* update SWE agent to use the history from State
* integration tests for SWE agent
* rename monologue to initial_thoughts, since that is what it is
* Refactor monologue to use the messages in state history
remove now unused method
* is_stuck update
* fix is_stuck
* unit tests
* fix tests
* Revert "Refactor monologue to use the messages in state history"
This reverts commit 76b4b765ef.
* Override eq for CmdOutputObservation to ignore the pid, compare the actual command only
* Revert "Override eq for CmdOutputObservation to ignore the pid, compare the actual command only"
This reverts commit 6418d856b5.
* improve error info logging
* Move assignment of self.state.error to report_error function
* only log exception to state, but not to user
---------
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
This PR includes three changes:
1) Iteration reminder should start with MAX_ITERATIONS from config rather than default value 100
2) In the first prompt, we should tell the LLM it has `MAX_ITERATIONS - 1` turns left, rather than `MAX_ITERATIONS - 2`
3) Remove legacy ITERATION_REMINDER config
* support returning states at the end of controller
* remove return None
* fix issue of overriding final state
* return the final state on close
* merge AgentState with State
* fix integration test
* add ChangeAgentStateAction to history in attempt to fix integration
* add back set agent state
* update tests
* update tests
* directly return get state
* add back the missing .close()
* Update typo in opendevin/core/main.py
---------
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
