* Refactor GPTAssistantAgent constructor to handle
instructions and overwrite_instructions flag
- Ensure that `system_message` is always consistent with `instructions`
- Ensure provided instructions are always used
- Add option to permanently modify the instructions of the assistant
* Improve default behavior
* Add a test; add method to delete assistant
* Add a new test for overwriting instructions
* Add test case for when no instructions are given for existing assistant
* Add pytest markers to test_gpt_assistant.py
* add test in workflow
* update
* fix test_client_stream
* comment out test_hierarchy_
* Add basic gptassistant notebook
- also improve logging in gpt assistant
* Update notebook/agentchat_oai_assistant_twoagents_basic.ipynb
Co-authored-by: Qingyun Wu <qingyun.wu@psu.edu>
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
Co-authored-by: kevin666aa <yrwu000627@gmail.com>
Co-authored-by: Qingyun Wu <qingyun.wu@psu.edu>
* Refactor GPTAssistantAgent constructor to handle
instructions and overwrite_instructions flag
- Ensure that `system_message` is always consistent with `instructions`
- Ensure provided instructions are always used
- Add option to permanently modify the instructions of the assistant
* Improve default behavior
* Add a test; add method to delete assistant
* Add a new test for overwriting instructions
* Add test case for when no instructions are given for existing assistant
* Add pytest markers to test_gpt_assistant.py
* add test in workflow
* update
* fix test_client_stream
* comment out test_hierarchy_
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
Co-authored-by: kevin666aa <yrwu000627@gmail.com>
* Handled possible IndexError in ConversableAgent.last_message method with more clear error message and added test in test_conversable_agent.py.
* Fix code formatting issues.
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* Notebook showing how to use select speaker to control conversation flow.
* pytest associated with notebook.
* Added llm_config to assistant and user proxy agent, and clarified why we set use_cache to false, as requested in the review.
* Added a @pytest.mark.skipif decorator like other tests to run it only in one py version, 3.10
* Fixed config warning.
* Removd llm_config to UserProxyAgent
* Fixed minor typos.
* Reran outputs
* Remopved llm_config from user_proxy_agent
* Colab Badge link updated.
* pre-commit formatting changes.
* Fixed base_url
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* LMM Code added
* LLaVA notebook update
* Test cases and Notebook modified for OpenAI v1
* Move LMM into contrib
To resolve test issues and deploy issues
In the future, we can install pillow by default, and then move back
LMM agents into agentchat
* LMM test setup update
* try...except... clause for LMM tests
* disable patch for llava agent test
To resolve dependencies issue for build
* Add LMM Blog
* Change docstring for LMM agents
* Docstring update patch
* llava: insert reply at position 1 now
So, it can still handle human_input_mode
and max_consecutive_reply
* Resolve comments
Fixing: typos, blogs, yml, and add OpenAIWrapper
* Signature typo fix for LMM agent: system_message
* Update LMM "content" from latest OpenAI release
Reference https://platform.openai.com/docs/guides/vision
* update LMM test according to latest OpenAI release
* Fully support GPT-4V now
1. Add a notebook for GPT-4V. LLava notebook also updated.
2. img_utils updated
3. GPT-4V formatter now return base64 image with mime type
4. Infer mime type directly from b64 image content (while loading
without suffix)
5. Test cases modified according to all the related changes.
* GPT-4V link updated in blog
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* Initial commit of the autogen testbed environment.
* Fixed some typos in the Testbed README.md
* Added some stricter termination logic to the two_agent scenario, and swiched the logo task from finding Autogen's logo, to finding Microsoft's (it's easier)
* Added documentation to testbed code in preparation for PR
* Added a variation of HumanEval to the Testbed. It is also a reasonable example of how to integrate other benchmarks.
* Removed ChatCompletion.start_logging and related features. Added an explicit TERMINATE output to HumanEval to save 1 turn in each conversation.
* Added metrics utils script for HumanEval
* Updated the requirements in the README.
* Added documentation for HumanEval csv schemas
* Standardized on how the OAI_CONFIG_LIST is handled.
* Removed dot-slash from 'includes' path for cross-platform compatibility
* Missed a file.
* Updated readme to include known-working versions.
* Adding async support to get_human_input
* Adjust code for Code formatting testing fail
* Adjust the test_async_get_human_input.py to run async on test
* Adjust the test_async_get_human_input.py for pre-commit-check error
* Adjust the test_async_get_human_input.py for pre-commit-check error v2
* Adjust remove unnecessary register_reply
* Adjust test to use asyncio call
* Adjust go back to not use asyncio
* async run group chat
* conversible agent allow async functions to generate reply
* test for async execution
---------
Co-authored-by: Qingyun Wu <qingyun.wu@psu.edu>
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* Update FAQ with workaround for Issue #251
* Update website/docs/FAQ.md
* Update website/docs/FAQ.md
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
* Update Installation.md
Replace autogen->pyautogen in env setup to avoid confusion
Related issue: #211
* Update Installation.md
Add deactivation instructions
* Update website/docs/Installation.md
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
---------
Co-authored-by: Chi Wang <wang.chi@microsoft.com>
