Repository navigation
Add Docker-free multi-turn SQLite text-to-SQL workflow (BIRD) - #635
Open
flyingdoubleG wants to merge 2 commits into
Open
flyingdoubleG wants to merge 2 commits into
flyingdoubleG wants to merge 2 commits into
Conversation
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The workflow called chat_async with n=1 only, so per-task rollout_args such as the eval taskset's temperature 0.0 in the example config were ignored. Forward asdict(task.rollout_args) like the frozen_lake workflow does, keeping n=1 since each turn needs a single response. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This branch has not been deployed
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 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Closes #634
Adds
sqlite_sql_workflow(trinity/common/workflows/envs/sqlite_sql/):sqlite_sql_env.py: pure-Python environment (no Trinity imports): read-only execution with anSQLite authorizer, single statement, timeout and row cap;
<action>/<answer>parsing thatignores tags inside
<think>; result rendering; BIRD/Spider execution-accuracy reward.sqlite_sql_workflow.py:MultiTurnWorkflowloop (schema + question + evidence prompt, up tomax_env_stepsturns, optionalauto_submit), experience info withenv_rounds,submitted,num_queries,num_query_errors,gold_error,gold_empty.sqlite_sql_workflow.examples/grpo_sqlite_sql/: README, data script (BIRD -> taskset), config (BIRD train, Mini-Dev eval).How to test:
pytest tests/common/sqlite_sql_env_test.py(13 tests: parsing, explore-then-answer,wrong answer, writes / ATTACH / PRAGMA / multi-statement rejected without touching the DB, timeout,
invalid format, broken/empty reference). I also exercised the workflow loop with a scripted model
(explore -> answer, no submission with/without auto_submit, write attempt as error observation).
Checklist