gavindiaz/NexQuant
1
0
Fork 0
mirror of https://github.com/NicolasBohn/NexQuant.git synced 2026-08-09 13:00:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

add feedback to workspace and ds runner base on costeer

Browse Source
This commit is contained in:
Xu Yang
2025-02-11 20:50:19 +08:00
committed by GitHub
parent 18de5eea43
commit c1085921b5
25 changed files with 383 additions and 231 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rdagent/components/coder/data_science/workflow/__init__.py
+9 -16
View File
@@ -41,21 +41,14 @@ class WorkflowMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
if queried_knowledge is not None
else []
)
latest_code_feedback = [
knowledge.feedback
for knowledge in queried_former_failed_knowledge[0]
if knowledge.implementation.file_dict.get("main.py") is not None
and knowledge.implementation.file_dict.get("main.py") == workspace.file_dict.get("main.py")
]
if len(latest_code_feedback) > 0:
queried_former_failed_knowledge = (
[
knowledge
for knowledge in queried_former_failed_knowledge[0]
if knowledge.implementation.file_dict.get("main.py") != workspace.file_dict.get("main.py")
],
queried_former_failed_knowledge[1],
)
queried_former_failed_knowledge = (
[
knowledge
for knowledge in queried_former_failed_knowledge[0]
if knowledge.implementation.file_dict.get("main.py") != workspace.file_dict.get("main.py")
],
queried_former_failed_knowledge[1],
)
# 2. code
system_prompt = T(".prompts:workflow_coder.system").r(
@@ -71,7 +64,7 @@ class WorkflowMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
ensemble_code=workspace.file_dict["ensemble.py"],
latest_code=workspace.file_dict.get("main.py"),
workflow_spec=workspace.file_dict["spec/workflow.md"],
latest_code_feedback=latest_code_feedback[0] if len(latest_code_feedback) > 0 else None,
latest_code_feedback=workspace.feedback,
)
for _ in range(5):
rdagent/components/coder/data_science/workflow/eval.py
+6 -6
View File
@@ -97,21 +97,21 @@ class WorkflowGeneralCaseSpecEvaluator(CoSTEEREvaluator):
if not submission_fp.exists():
stdout += "\nSubmission file (submission.csv) is not generated."
else:
base_check_code = (DIRNAME / "eval_tests" / "submission_check.txt").read_text()
implementation.inject_files(**{"submission_check.py": base_check_code})
base_check_code = (DIRNAME / "eval_tests" / "submission_format_test.txt").read_text()
implementation.inject_files(**{"submission_format_test.py": base_check_code})
# stdout += "----Submission Check 1-----\n"
stdout += implementation.execute(env=de, entry="python submission_check.py")
stdout += implementation.execute(env=de, entry="python submission_format_test.py")
# MLEBench Check
# !!! Since we are running on a sampled dataset, mlebench check is not required.
# mle_check_code = (
# (DIRNAME / "eval_tests" / "mle_submission_check.txt")
# (DIRNAME / "eval_tests" / "mle_submission_format_test.txt")
# .read_text()
# .replace("<competition_id>", self.scen.competition)
# )
# implementation.inject_files(**{"mle_submission_check.py": mle_check_code})
# implementation.inject_files(**{"mle_submission_format_test.py": mle_check_code})
# stdout += "----Submission Check 2-----\n"
# stdout += implementation.execute(env=mde, entry=f"python mle_submission_check.py")
# stdout += implementation.execute(env=mde, entry=f"python mle_submission_format_test.py")
system_prompt = T(".prompts:workflow_eval.system").r(
scenario=self.scen.get_scenario_all_desc(),
rdagent/components/coder/data_science/workflow/eval_tests/submission_check.txt → rdagent/components/coder/data_science/workflow/eval_tests/submission_format_test.txt
View File