Commit Graph

17 Commits

Author SHA1 Message Date
Kaushik Deka
415843476c Feat: Add Vision Input Support for LLM with Vision Capabilities (#2848)
* add image feature

* fix-linting

* check model support for images

* add comment

* Add image support to other models

* Add images to chat

* fix linting

* fix test issues

* refactor variable names and import

* fix tests

* fix chat message tests

* fix linting

* add pydantic class message

* use message

* remove redundant comments

* remove redundant comments

* change Message class

* remove unintended change

* fix integration tests using regenerate.sh

* rename image_bas64 to images_url, fix tests

* rename Message.py to message, change reminder append logic, add unit tests

* remove comment, fix error to merge

* codeact_swe_agent

* fix f string

* update eventstream integration tests

* add missing if check in codeact_swe_agent

* update integration tests

* Update frontend/src/components/chat/ChatInput.tsx

* Update frontend/src/components/chat/ChatInput.tsx

* Update frontend/src/components/chat/ChatInput.tsx

* Update frontend/src/components/chat/ChatInput.tsx

* Update frontend/src/components/chat/ChatMessage.tsx

---------

Co-authored-by: tobitege <tobitege@gmx.de>
Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com>
2024-08-04 02:26:22 +08:00
Xingyao Wang
2e60d25eae [Agent, LLM] Make sure codeact agent produce message in u/a/u/a order (#3193)
* make sure codeact agent produce message in u/a/u/a order

* integration tests

* sync message changes to codeact swe

* fix integration tests

---------

Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
2024-08-02 00:17:53 +08:00
Graham Neubig
c897791024 Refactor LLM config (#2953)
* Add max_message_chars to LLM

* Refactor LLM config

* Fix tests

* Made some functions class functions

* Fix regression

* Fixed comments
2024-07-17 09:16:04 -04:00
Anush Kumar V
8f76587e5c docs: updated docstrings using ruff's autofix feature (#2923)
* 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>
2024-07-16 01:35:33 +00:00
Xingyao Wang
e45ddeb2a2 arch: deprecating recall action and search_memory (#2900)
* deprecating recall action

* fix integration tests

* fix integration tests

* remove search memory
2024-07-12 19:23:21 +00:00
Boxuan Li
c68478f470 Customize LLM config per agent (#2756)
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)
2024-07-09 22:05:54 -07:00
Engel Nyst
d37b2973b2 Refactoring: event stream based agent history (#2709)
* add to event stream sync

* remove async from tests

* small logging spam fix

* remove swe agent

* arch refactoring: use history from the event stream

* refactor agents

* monologue agent

* ruff

* planner agent

* micro-agents

* refactor history in evaluations

* evals history refactoring

* adapt evals and tests

* unit testing stuck

* testing micro agents, event stream

* fix planner agent

* fix tests

* fix stuck after rename

* fix test

* small clean up

* fix merge

* fix merge issue

* fix integration tests

* Update agenthub/dummy_agent/agent.py

* fix tests

* rename more clearly; add todo; clean up
2024-07-07 21:04:23 +00:00
Shimada666
d22ff73905 Make the sandbox Python runtime completely transparent (#2796)
* Make the sandbox Python runtime completely independent

* fix source bashrc

* add pip install instruction for ipython to fix intergration tests for codeact swe

* update integration tests

* change flake8 command to (maybe) fix sandbox tests?

* make lint support both unittest & sandbox

* fix agnostic image build error

* refactor build script

---------

Co-authored-by: Xingyao Wang <xingyao6@illinois.edu>
Co-authored-by: tobitege <tobitege@gmx.de>
2024-07-06 15:22:44 +02:00
Leo
c2f557edde refactor: multiple code improvements (#2771) 2024-07-04 18:51:22 +08:00
Engel Nyst
e24c52d060 Small refactoring of obs truncation (#2701)
* refactor truncate_content a bit to be usable by all agents

* adjust doc
2024-06-30 12:12:08 +02:00
Boxuan Li
e45b311c35 Remove MAX_CHARS traffic control (#2694)
* Remove MAX_CHARS limiting

* More cleanup
2024-06-29 12:59:41 -07:00
Engel Nyst
80fe13f4be rename our completion as a drop-in replacement of litellm completion (#2509) 2024-06-19 05:25:25 +02:00
tobitege
823298e0d0 fix: Agentskills enhancements (#2384)
* avoid repeat logging of unneeded messages

* refactored append/edit_file (tests next)

* agentskills and unit test fixes

* testing

* more changes and test prompts

* smaller changes

* final test fixes

* remove dead code from test_agent.py

* reverting unneeded changes

* updated tests, more tweaks to skills

* refactor (#2442)

* chores: fix DelegatorAgent description (#2446)

* change

* change comments

* fix

* stopped container to prevent port issues. (#2447)

* chore: remove useless browsing code in CodeActSWEAgent (#2438)

* remove useless

* fix integration test

* Regenerate test_ipython_module artifacts for CodeActSWEAgent

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Merge remote-tracking branch 'upstream/main' into agent-fileops

* unneeded tweak

* * fix edit_file to not introduce extra newline
* updated docstrings with more details for LLM
* fix legacy typo in prompts causing ]] instead of ]
* several mock files regenerated

* Regen'ed CodeActSWEAgent integration tests

* fix _print_window signature; explicit exception type in _is_valid_path

* splitlines with named param

---------

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: மனோஜ்குமார் பழனிச்சாமி <smartmanoj42857@gmail.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-16 15:06:46 -04:00
Yufan Song
426e429b18 chore: remove useless browsing code in CodeActSWEAgent (#2438)
* remove useless

* fix integration test

* Regenerate test_ipython_module artifacts for CodeActSWEAgent

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-15 10:53:03 +08:00
Yufan Song
7f8c324d3a Refactor CodeActSWEAgent, add response parser (#2368)
* refactor code

* fix

* seperate code

* Update agenthub/codeact_swe_agent/response_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_swe_agent/response_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* Update agenthub/codeact_swe_agent/action_parser.py

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>

* remove browsing

---------

Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-06-13 04:14:05 +00:00
Aaron Xia
42c6b506b5 Lazy launching BrowseEnv / making BrowseEnv optional (#2155)
* feat: lazy launching browser; browser optional for diffrent agents.

* style: lint

* fix: integration test fail due to browser not started.

* fix: run by cli and integration test failed.

* fix: lint

* fix: lint

---------

Co-authored-by: Graham Neubig <neubig@gmail.com>
2024-05-31 16:40:42 -04:00
Xingyao Wang
01ef90205d Add CodeActSWEAgent to remove browsing & github + improvements on agentskills (#2105)
* update swe_bench prompt;
use minimal prompt for codeact;

* upgrade agentskills and update testcases

* update infer prompt

* fix cwd

* add icl for swebench

* also log in_context_example to run infer

* remove extra print

* change prompt to abs path

* update error message to include current file info

* change cwd for jupyter if needed

* update edit error message

* update prompt

* improve git get patch

* update hint string

* default to 50 turns

* revert changes from codeact agent and create new CodeActSWEAgent

* revert changes to codeact

* revert instructions for run infer

* revert instructions for run infer

* update README

* update max iter

* add codeact swe agent

* fix issue for CodeActSWEAgent

* allow specifying max iter in cmdline script

* stop printing

* Update agenthub/codeact_swe_agent/README.md

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>

* Fix prompt regression in jupyter plugin

---------

Co-authored-by: Yufan Song <33971064+yufansong@users.noreply.github.com>
Co-authored-by: Boxuan Li <liboxuan@connect.hku.hk>
2024-05-29 21:19:00 -07:00