[pull] main from agentscope-ai:main#44
Merged
pull[bot] merged 1 commit intopan-x-c:mainfrom Mar 24, 2026
Merged
Conversation
) * feat(tuner): naive adaption of dspy * feat(tuner): implement naive prompt tuning with dspy integration * refactor(tuner): add config and logs for prompt tuning * fix(tuner): remove trailing whitespace and unused code * feat(tuner): export prompt tuning config and function in module init * feat(tuner): add training summary * docs(tuner): add comprehensive docstrings for prompt tuning module * refactor(tuner): rename Dataset to DatasetConfig and TunerChatModel to TunerModelConfig * feat(tuner): add model selection optimizer * fix(tuner): avoid sharing agent context in all prompt tuning dataset * feat(tuner): add local file loading support for training datasets * fix(tuner): fix dataset loading by using data_files parameter * feat(tuner): add local file loading support for evaluation datasets * feat(tuner): use deepcopy for agent to prevent shared state during optimization * feat(tuner): simplify prompt-tune example prompt * refactor(tuner): change prompt tuning to optimize prompt strings instead of agent instances * docs(tuner): add comprehensive prompt tuning guide and relocate example * chore: clear code * docs(tuner): clarify that example uses GSM8K subset in README * chore(tuner): format code * refactor(tuner): make prompt_tune internal modules private * chore(tuner): format code * feat(tuner): add model select * fix(tuner): improve model selection and built-in judges robustness * refactor(tuner): rename and encapsulate prompt tuning internals * fix(tuner): specify response type as WorkflowOutput in judge functions * fix(tuner): correct response usage extraction in avg_token_consumption_judge * refactor(tuner): simplify prompt tuning interface with minimal workflow signature * refactor(tuner): return optimization metrics in tune_prompt * chore(deps): reorganize dependencies for tuner support * fix(tuner): modify model selection logic according to pr comments, mainly to record metrics and unify optimization direction * fix: fix issues for getting token usage; add examples for model selection * add naive implementation of model selection tuning function, including metrics by time and token usage * refactor(tuner): simplify WorkflowType and update prompt tuning integration * chore(deps): update datasets and trinity-rft * refactor(tuner): simplify JudgeType and remove auxiliary_models parameter from prompt tuning * refactor(tuner): validate functions and simplify invocation in model selection * refactor(tuner): clean up imports and improve code formatting * docs(tuner): update docstrings for clarity in prompt tuning and its example * fix(tuner): remove duplicated code * fix(tuner): add concurrent execution for efficiency in model selection * chore: format code * chore: remove empty code * docs(tuner): update example name * refactor(tuner): model select use response any * docs(tuner): update tuning docs * deps: switch trinity to seperate tuner-gpu deps * docs(tuner): apply suggestion from @pan-x-c Co-authored-by: Xuchen Pan <32844285+pan-x-c@users.noreply.github.com> * refactor(tuner): update bleu example * docs(tuner): update url * chore: pre-commit check * backup: clean wrongly commit user configs * fix(tuner): minor typo of dependencies * fix(tuner): fix for pre-commit flake8 * fix(tuner): fix for pre-commit pylint * fix(tuner): minor fixes * fix(tuner): pre-commit issues * fix(tuner): pre-commit issues * fix(tuner): minor fixes * fix(tuner): minor fixes * fix(tuner): pre-commit black * fix(tuner): trailing coma fix * fix(tuner): typo fix * fix(tuner): pyproject fix * fix(tuner): fix black pre-commit * chore(tuner): format code --------- Co-authored-by: wce <wce@pku.edu.cn> Co-authored-by: 天置 <wce@pku> Co-authored-by: Xuchen Pan <32844285+pan-x-c@users.noreply.github.com>
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
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
1 participant
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.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )