EUR/USD synthetic data has \$volume=0 for all rows, causing any VWAP or
volume-weighted factor to produce all-NaN output. Insert a guard after
pd.read_hdf() that replaces zero volume with (\$high - \$low) range proxy
so volume-dependent factors produce meaningful signals.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
LLM sometimes copies the .reset_index(level=N, drop=True) suffix from
groupby().rolling().method() patterns and adds it after .transform(),
but transform() already preserves the original index. The extra
reset_index() drops an index level and causes ValueError: 'cannot reindex
on an axis with duplicate labels' or shape mismatch on assignment.
Detect: any line containing both .transform( and .reset_index(level=..., drop=True)
Fix: strip the .reset_index() suffix from those lines.
Adds 1 new test (test_transform_reset_index_stripped) — total 30 tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. _fix_instrument_column_access: var['instrument'] = EXPR was incorrectly
converted to var.index.get_level_values(1) = EXPR, producing a SyntaxError
('cannot assign to function call'). Added (?!\s*=) negative lookahead to
skip assignment targets.
2. _fix_groupby_column_on_multiindex: groupby(['instrument','date']) on a
reset_index() variable was converted to groupby([var.index.get_level_values...])
but reset_index() produces a plain RangeIndex, not a MultiIndex, causing
AttributeError: 'RangeIndex' has no attribute 'normalize'. Added reset_vars
guard to skip variables produced by reset_index().
Adds 1 new test (test_assignment_target_not_touched) — total 29 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. groupby(level=['instrument','date']) → get_level_values() — string level
names like 'date' don't exist in the (datetime, instrument) MultiIndex;
replaced with get_level_values(0).normalize() + get_level_values(1).
2. groupby(level=['date','instrument']) — symmetric fix for reversed order.
3. groupby(level=['instrument']) → groupby(level=1) — single string level.
4. groupby(level=N)['col'].apply(lambda) → transform(lambda) — apply() on a
grouped Series prepends an extra index level, causing index shape mismatch
when assigned back; transform() preserves the original index.
5. df.loc[instrument] DateParseError fix (instrument_loc_multiindex) — already
committed, adding supporting tests for groupby(level=['instrument','date']).
Adds 5 new tests (TestGroupbyLevelStringNames, TestGroupbyApplyToTransform)
— total 28 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When LLM iterates over instruments via get_level_values('instrument').unique()
and then does df.loc[instrument], pandas tries to parse the instrument string
('EURUSD') as a datetime against level-0 of the (datetime, instrument) index,
raising DateParseError.
Fix: detect loop variables bound to get_level_values(1) or get_level_values('instrument')
and replace DF.loc[loop_var] (read) with DF.xs(loop_var, level=1). Assignment
write-backs are left untouched to avoid complex rewrites.
Adds 4 new tests (TestInstrumentLocMultiindex) — total 23 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
LLM-generated code often accesses df['instrument'] as a column, but
'instrument' is an index level (level 1) in the MultiIndex DataFrame.
Replace with df.index.get_level_values(1) except when the variable
was created via reset_index() (where the column actually exists).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
LLM generates invalid Python by putting keyword args inside lists:
df.groupby([level=1, 'date']) ← SyntaxError
Also fixes the regex for the chained groupby Pattern A/B which had
an unescaped ')' causing re.error that silently reverted the fix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The fixer was raising min_periods to match window size, which causes
all-NaN output for intraday factors with 96 bars/day — window=240 means
zero valid bars per day, window=60 means 61% NaN per day. Critics were
consistently flagging this as incorrect for intraday factors. The LLM
now controls its own min_periods.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
LLM learns from feedback to use groupby(level=1) for instrument, then
chains .groupby('date') to add the date dimension — but DataFrameGroupBy
has no .groupby() method, causing AttributeError at runtime.
Replace the invalid chain with a correct two-level groupby using
index.get_level_values(), consistent with the existing instrument+date fix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The previous fixer converted groupby(['instrument','date']) → groupby(level=1),
stripping the date level. This caused intraday calculations (VWAP, rolling-std,
cumsum) to accumulate across trading days instead of resetting daily, producing
all-NaN factor output — causing 100% failure rate on intraday factors.
New behaviour: capture the DataFrame variable name and emit:
var.groupby([var.index.get_level_values(1),
var.index.get_level_values(0).normalize()])
which groups by (instrument, day) as originally intended.
Adds test/qlib/test_auto_fixer.py covering all fixer cases.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- monte_carlo_trade_pvalue(): shuffles trade P&L N times, returns fraction of
permuted sequences that beat real total return (p<0.05 = genuine edge)
- walk_forward_rolling(): multiple IS/OOS windows (IS=3yr, OOS=1yr, step=1yr),
computes wf_oos_sharpe_mean, wf_oos_consistency (% profitable windows)
- backtest_signal_riskmgmt(): new wf_rolling and mc_n_permutations params
- Strategy generator: enables both (200 MC permutations), adds mc_ok and wf_ok
to acceptance filter (mc_p<0.20, wf_consistency>=50%)
- Rebacktest script: enables both, stores all wf_*/mc_* fields in write-back
- 6 new tests covering MC pvalue, disabled-by-default, zero-trades edge case,
rolling WF key presence and consistency range
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Covers backtest_signal_riskmgmt leverage caps, zero-signal, IS/OOS split keys,
bar counts, OOS independence from IS losses, and Monte Carlo permutation
tests (marked slow, excluded from default pytest run).
Also excludes slow-marked tests from default addopts in pyproject.toml.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add vbt_backtest.py as single source of truth for all metric formulas
(Sharpe, drawdown, IC, transaction costs) — backtest_engine.py and
strategy_orchestrator.py now delegate to it
- Add LLMUnavailableError to exception.py; rd_loop.py catches it at the
proposal stage and raises LoopResumeError to avoid corrupting trace
history with None hypotheses
- Guard record() against None exp/hypothesis so loop resets leave
trace.hist in a consistent state
- Refactor strategy_orchestrator and optuna_optimizer to use unified
backtest path; remove duplicate metric calculation code
- Add predix_rebacktest_unified.py script for offline re-evaluation
- Update tests and README
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- MLTrainer class with feature matrix builder from top factors by IC
- LightGBM training with time-series split (80/20) and early stopping
- Feature importance analysis (gain-based) with ranking
- Model persistence: model.txt + metadata.json + feature_importance.json + CSV
- Feedback generation for factor generation loop
- Model loading from disk
- Full pipeline: load factors -> train -> save -> generate feedback
- 46 unit tests covering all features
- lightgbm and scipy added to requirements.txt
Created rdagent/scenarios/qlib/local/data_loader.py:
- OHLCV loading with thread-safe caching
- Factor metadata loading (sorted by IC)
- Factor time-series loading with alignment
- Feature matrix builder
- Randomized factor selection for diverse strategies
- 11 tests passing
Note: data_loader.py is in local/ (closed source)
Test file is public to validate interface.
- TestCLIModelSelection: 8 tests for predix.py CLI
* predix module imports
* fin_quant --model option
* predix quant --model and --log-file options
* OpenRouter API key validation
* TeeWriter existence check
* health and status commands
- TestLoggingTeeWriter: 2 tests for TeeWriter
* Multi-stream writing
* Broken stream handling
All 103 integration tests pass (93 + 10 new).
Also fix predix.py logging:
- Add --log-file flag (default: fin_quant.log)
- TeeWriter writes to both console AND file
- Works for both local and openrouter backends
- Remove duplicate DB save from quant.py (keep only in factor_runner)
- Add explicit DB path creation with mkdir -p
- Add JSON factor summaries to results/factors/
- Add debug logging for result structure
- Fix logger.debug -> logger.info (RDAgentLog compatibility)
- Update tests to match new architecture (240/240 passing)
- Enhance extract_results.py with progress indicators
Implement automatic trading protection system to prevent excessive losses:
PROTECTIONS (100% original code, NOT copied from Freqtrade):
- Max Drawdown Protection: Blocks trading when DD > 15% (configurable)
- Cooldown Period: 4h mandatory rest after 5% loss
- Stoploss Guard: Detects stoploss clusters (>5 per day)
- Low Performance Filter: Filters factors with Sharpe < 0.5, Win Rate < 40%
ARCHITECTURE:
- Base protection interface with common utilities
- 4 specialized protection implementations
- ProtectionManager orchestrates all active protections
- Time-based blocking with automatic expiry
TESTS (32 total, ALL PASS):
- 25 unit tests in test/backtesting/test_protections.py
- 7 integration tests in test/integration/test_all_features.py
- Tests cover: normal operation, edge cases, error handling
DOCUMENTATION:
- Update QWEN.md with development guidelines for AI assistant
* Mandatory rules: Update QWEN.md, README, requirements.txt, tests
* Pre-commit checklist
* Example workflow
- Update README.md with protection system features
- Update project structure with new modules
All code is 100% original - NO license issues with Freqtrade GPLv3.
- Remove optional code quality hooks (black, isort, ruff, mypy, toml-sort)
* These blocked commits when tools not installed
* Users can run them manually when needed
- Keep only MANDATORY hooks:
* Integration Tests (60 tests, ~7.5s)
* Bandit Security Scan
- Both MUST pass before every commit
- Remove api_key parameter from generate_api_config()
- Update API_CONFIG_TEMPLATE to read TEST_API_KEY from environment at runtime
- Pass TEST_API_KEY via Docker env vars instead of writing to config file
- Fixes py/clear-text-storage-sensitive-data vulnerability
- API key is now read from os.environ.get('TEST_API_KEY') at runtime
- Remove api_key parameter from function signature
- API key is now exclusively read from TEST_API_KEY env var
- Complete removal of API key handling from config generation
Security improvements:
- No API key parameters passed through function calls
- Reduces risk of accidental logging or exposure
- Consistent with security best practices
- Read API key from environment variable instead of config file
- Prevents accidental exposure of API keys in code/config
- Security best practice: secrets should not be stored in files
Security improvements:
- API keys read from TEST_API_KEY environment variable
- Empty string as fallback (will fail gracefully if not set)
- No secrets stored in test configuration files
- Fix Path.cwd() mocking to use correct module path
- Add Path.resolve() mocking for proper path validation testing
- Fix PermissionError handling test with proper mock
- All 14 security tests now pass
Fixes broken tests from previous commit that had incorrect mocking.
New models in models/local/:
- transformer_factor.py: Transformer with self-attention
- tcn_factor.py: Temporal Convolutional Network (multi-scale)
- patchtst_factor.py: PatchTST (SOTA for time-series)
- cnn_lstm_hybrid.py: CNN+LSTM with attention
Features:
- All models support sequence and tabular data
- Automatic device selection (CPU/GPU)
- Training with Adam optimizer + LR scheduler
- Save/load functionality
- Production-ready code
Dependencies installed:
- xgboost
- lightgbm
- torch (PyTorch)
Usage:
from rdagent.components.model_loader import load_model
model = load_model('transformer_factor') # Auto-loads your local version!
- Replace hardcoded API_KEY with os.getenv('TEST_API_KEY')
- Replace hardcoded HF_TOKEN with os.getenv('TEST_HF_TOKEN')
- Replace hardcoded API_BASE with os.getenv('TEST_API_BASE')
- Replace hardcoded MODEL with os.getenv('TEST_MODEL')
- Add test credentials patterns to .gitignore
- Fixes GitHub Security Alert #8 (py/clear-text-storage-sensitive-data)
Sensitive data is now loaded from environment variables instead of clear text.
Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>
* feat: rdkit for chemcotbench
* update qwen2.5&llama3.1 context
* fix: force failure on validation error and remove try/except in validator
* feat: unified error sample extraction (with test scripts)
* feat: set conda cache with .env
* feat: skip data eval if data pass in last evo
* fix: rm redundant param
* fix ui bug
* refactor: centralize assign_code_list_to_evo in MultiProcessEvolvingStrategy
* feat: add test_params.yaml generation and workspace cleanup improvements for finetune
* refactor: replace get_clear_ws_cmd with clear_workspace and update prompts for hard check criteria
* add bioprobench dataset
* fix: handle commas in training config extraction and refactor prompt includes
* bioprobench description
* add bioprobench readme
* feat: merge lora adapter for blackwell gpu
* feat: support for multi benchmarks in one job
* change dfficult aware content for training
* update difficulty-aware and logging principles
* fix: resolve variable name conflict in FTRunnerEvaluator
* set job id accuracy to minute
* feat(ui): display one selected metric per benchmark
* feat: store sota exp, and fix ws_ckp bug
* fix: truncate data.json in feedback
* fix: opencompass data for conda env
* fix: save only the last model
* feat: set log path and ws path
* fix: set overwrite_cache to avoid lock contention(through injecting params)
* feat: redirect stdout to file in localenv
* add pickle cache to dataset desc
* fix CI
* fix: remove redundant wrapper
* feat: set python_unbuffered
* move redirect stdout to env run
* fix a small bug
* move model folder
* feat(ui): display benchmark baseline
* fix: enrich scenario and benchmark description
* fix: rewrite runner eval to accept easier
* feat: compare with baseline when no SOTA
* update tablebench readme
* fix: switch back to single benchmark (for baseline)
* feat(ui): add ws path in ui
* refactor: update SOTA tracking to use DAG traversal and parent selection
* fix: prioritize local_selection in trace and refactor sibling retrieval logic
* refactor: unify error handling in feedback generation and update workspace injection
* feat: add skip_loop_error_stepname to control error skip step in LoopBase
* fix: set local_selection to NEW_ROOT for experiments without parent
* feat: set different ports for jobs
* feat: set different ports for jobs
* feat: add upper data size limit for LLM fine-tuning and update related prompts
* fix: replace get_truncated_stdout() with stdout for consistent output handling
* refactor: remove data.json from cache and workspace logic, focus on script-based reuse
* fix: rm target_scenario
* feat: add selective cache extraction and custom cache key for data processing
* fix(ui): bug when displaying tablebench
* fix: filter config in dataset_info.json
* feat: add test set, set valid set
* feat(ui): update test score, and set color for final decision
* feat: add test score for baseline and update ui
* fix: use [-100:] as test range
* feat: update data_stats in runner
* feat: wait for opencompass init when run multi jobs
* fix: adjust test&valid split
* feat: force to generate COT(with <think> token), and add answer format in scenarios.json
* feat: improve ui
* fix: unify benchmark volume mounts and set extra_volumes for conda env
* fix(ui): number color
* fix: update GPU memory handling to use total memory in GB and streamline code
* fix: set use_cot_postprocessor
* feat: add env_dict to config classes and merge env vars in Env run
* fix: let coder obey proposal
* fix(ui): direction bug and update chemcot core metirc
* fix: set consistent benchmark mount points and env vars for docker and conda
* fix: addintional target for LoRA
* feat: workspace dir log for benchmark running
* fix: tableInstruct path bug and update benchmark description
* feat: timeout for whole job
* fix: align FinanceIQ import to opencompass
* feat: use llm_judge for FinanceIQ
* feat: switch to turn on <think> or not
* feat: using scripts to redirect stdout, and run in different windows
* feat: sync litellm log
* fix: gpu memory format
* fix: escape special characters in benchmark desc
* fix: set data processing timeout to 1h
* feat: set valid_loss and save_best_model
* fix: inject timeout and stage
* fix: loss history extract logic
* feat: inject output dir
* feat: inject eval batch size
* feat: inject save_total_limit
* feat: update data prompt
* fix: escape shell special characters
* fix: tablebench visualization UI
* fix: move implementation validation to coder, and ignore injected params
* docs: add README for RL-PostTraining evaluation system
* Add AutoRL-Bench evaluation framework for RL post-training
* Add architecture documentation
* docs: update architecture and interface documentation for AutoRL-Bench
* improve doc
* fix
* refactor: YAML配置驱动
* feat: add RL Docker env, workspace test, and update project structure
* feat: 重命名 autorl_bench, 新增 RLWorkspace, 配置 Docker extra_volumes
* Add eval-only AutoRL-Bench pipeline
* sturcture clean
* docs: add autorl_bench README
* feat(rl): Implement RL post-training agent scaffold and example
* refactor: simplify RL scenario classes and update RL CoSTEER integration
* feat(rl): 调通 scaffold，mock 数据跑完 5 步循环
* feat(rl): 接入 LLM 生成代码，支持 model_path 传递
* feat(rl): Docker 执行框架，RLWorkspace.run() + RLPostTrainingRunner
* feat(rl): LLM 生成假设/反馈，完整 loop 跑通
* feat: add RL post-training entry point with configurable options
* refactor: simplify RL proposal and trace classes, update config and docs
* Update rl eval autorl_bench layout
* Update RL workflow and evaluation setup
* Integrate AutoRL-Bench evaluation in RL workflow
* feat(rl): 添加 --base-model/--benchmark CLI 参数，简化 RLTask
* feat(rl): Docker 环境动态选择 + example_agent 完整训练评测流程（无llm）
* fix(rl): 修复 feedback 传递 + 添加 verl 依赖
* refactor: remove unused validate in BenchmarkAdapter and add core utils module
* feat(rl): UI
* Refactor autorl_bench layout and docker entrypoint
* autorl_bench: add aider autoloop tool
* feat(rl): environment docker
* refactor: simplify aider autoloop tooling
* chore: update misc files
* feat(rl): yaml-driven dataset download & auto-download on startup
* feat(rl): yaml-driven dataset download & auto-download on startup
* Refactor RL eval runner and clean up
* Simplify RL eval runner and env
* rl: include litellm in RL docker image
* feat(rl): unified resource path & model repo_id structure
* feat(rl): refactor eval with OpenCompass & add training code template
* feat(rl): refactor eval with OpenCompass & add training code template
* feat(rl): delete test bench
* docs: add benchmark interface notes and TODOs for unified evaluation
* feat(rl): unified benchmark eval interface + shared configs
* feat(rl): 优雅
* feat(rl): prompt prososal+coder improve
* feat(rl): fix eval
* fix(rl): docker
* fix(rl): eval
* v 1.0 tmep
* benchmark v1.0
* benchmark v1.1
* benchmark v1.1: grading日志+代码去重
* benchmark v1.1: grading日志+代码去重
* benchmark v1.1: grading日志+代码去重+task description
* benchmark v1.2: fix
* benchmark v1.3: fix，example-agent ok，rdagent test，openhands develop
* benchmark v1.4: fix，example-agent ok，rdagent ok，openhands develop
* benchmark : add alfworld
* benchmark : update readme
* benchmark : update readme
* benchmark :
* chore: add eval bypass block and mark TODO in grading server
* benchmark
* benchmark
* benchmark
* benchmark
* alfworld
* alfworld
* benchmark
* rdagent
* rdagent
* benchmark
* benchmark：ui
* benchmark：delete docker + log
* 1
* alfworld
* ui
* alfworld
* readme
* alfworld
* parallex
* alfworld
* run
* eval gpu
* alfworld
* alfworld
* fix conda init in start.sh for non-interactive shells
Fallback to common miniconda paths when conda is not in PATH.
Fixes B200 pod startup failure (conda: command not found).
Made-with: Cursor
* simplify start.sh: read TRAINING_PYTHON from .env
No more conda detection logic. Just set TRAINING_PYTHON in .env.
Fallback to conda only if not set.
Made-with: Cursor
* use OPENHANDS_PYTHON from .env to run agent
start.sh now uses OPENHANDS_PYTHON for main.py execution,
since the parent process may be in a different conda env.
Made-with: Cursor
* feat: register OpenCode agent into autorl_bench framework
- Add agents/opencode/ with config.yaml, start.sh, README.md
- Include opencode-rl pipeline code (pipeline/, runner_fsm/, benchmarks/)
- Merge opencode-rl dependencies into autorl_bench requirements.txt
- Remove separate venv requirement, share main environment
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* Update opencode agent, benchmarks, and eval configs
- Sync opencode-rl runner_fsm with latest simplifications
- Add smith benchmarks integration
- Update opencompass configs and server with GPU support + error handling
* Update OpenCode agent docs for external opencode-rl integration
- Document external repo architecture (opencode-rl as independent plugin)
- Add setup instructions for cloning and configuring opencode-rl
- Add architecture diagram showing RD-Agent ↔ opencode-rl interaction
- Document OPENCODE_RL_ROOT for custom paths
* feat: add smith benchmark discovery and per-sample evaluator
- Add smith/ module for dynamic benchmark discovery from rl-smith
- Add PerSampleEvaluator for per-sample scoring via vLLM
- Update utils.py to support script-based data download for smith benchmarks
- Update opencode agent config
* enforce RL-only in instructions.md; remove embedded opencode-rl
- instructions.md: prohibit SFT, require RL (GRPO/PPO) for all benchmarks
- remove agents/opencode/opencode-rl/ (runtime uses external OPENCODE_RL_ROOT)
Made-with: Cursor
* comment out OpenCode-only deps in requirements.txt
openai, httpx, python-dotenv, tenacity are for OpenCode agent's
separate environment. Keep peft and pydantic as shared deps.
Made-with: Cursor
* refactor: extract _kill_process_group, narrow exception catches
- run.py: replace 2x nested 3-level try/except with shared
_kill_process_group() using loop + specific exceptions
- server.py: except Exception → except (RuntimeError, ValueError, OSError)
- utils.py: except Exception → except requests.ConnectionError
Made-with: Cursor
* move kill_process_group to core/utils for reuse
Extract from run.py into core/utils.py so other runners
can also use it. Exported via core/__init__.py.
Made-with: Cursor
* add comments to run.py for workspace isolation and signal handling
Made-with: Cursor
* remove OpenCode-only deps from requirements.txt entirely
Made-with: Cursor
* allow SFT in instructions, RL as ultimate goal
Made-with: Cursor
* add workspace isolation rules to instructions.md
Use relative paths, forbid cd outside workspace, ignore symlink targets.
Made-with: Cursor
* update opencode start.sh: use OPENCODE_PYTHON, add PATH for opencode CLI, remove unsupported args
Made-with: Cursor
* opencode start.sh: pass --run-dir to use AutoRL-Bench workspace
Ensures OpenCode-FSM-Runner writes outputs into the workspace prepared
by AutoRL-Bench instead of creating its own runs/ directory.
Made-with: Cursor
* opencode start.sh: prepend training env bin to PATH
Ensures LLM agent bash calls (e.g. python3 -c "from trl import ...")
resolve to the correct training environment, instead of relying on
parent shell conda activation.
Made-with: Cursor
* opencode start.sh: restore --max-retries and --eval-timeout for opencode-rl
Made-with: Cursor
* add humaneval benchmark
* Replace import * cleanup hack with explicit imports in OpenCompass config
- Resolve dataset variable names via importlib before generating config,
so the template uses `from xxx import datasets` instead of `import *`
- Remove the fragile runtime cleanup hack that set leaked modules to None
- Increase OpenCompass timeout from 3600s to 7200s
- Fix score parsing to average across multiple subdatasets
* refine opencompass config file generating
* add humaneval benchmark dependency instructions
human-eval package requires clone from open-compass/human-eval with
a one-line patch to relax assertion for partial evaluation (test split only).
Made-with: Cursor
* fix: sanitize user-provided paths in RL UI (CodeQL)
* fix: resolve user path relative to safe root (CodeQL)
* fix: use Copilot-suggested path sanitization pattern (CodeQL)
* fix: normalize and reject absolute user paths (CodeQL)
* Fix training params, vLLM OOM cleanup, OpenCompass score parsing, and baseline cache logic
* fix: add setuptools<75 to requirements for opencompass pkg_resources dependency
uv venv does not include setuptools by default, causing OpenCompass baseline
evaluation to fail with "No module named 'pkg_resources'".
Made-with: Cursor
* webshop
* feat(autorl_bench): improve smith benchmark integration and evaluator robustness
- Add smith benchmark docs to README: usage examples, discovery mechanism, SMITH_BENCH_DIR
- Improve PerSampleEvaluator: vLLM GPU cleanup, test_range slicing
- Refactor server.py: extract grading server from utils
- Fix OpenCompass score parsing and baseline cache logic
* fix: add smart fallback for OpenCompass dataset variable resolution
When build_dataset_imports_explicit() fails to import an OpenCompass
dataset module (common in grading server subprocess), it now guesses
the correct variable name from the module path convention instead of
falling back to empty names (which causes import * and breaks BBH
due to leaked file handle objects).
* revert: restore opencompass.py to pre-modification state
Revert vLLM pid cleanup, dash-value checks, and metric-based
score parsing added in 31caff2f and bb32e555.
* keep metric-aware score parsing in opencompass; add baseline column to UI
- opencompass.py: retain metric-type filtering (accuracy/score) instead
of naive averaging, avoids polluting scores with pass/timeout counters
- ui.py: add Baseline column to Agent Summary table
Made-with: Cursor
* fix: handle non-string answers in extract_answer to prevent TypeError
arc_agi and other benchmarks can have non-string answer fields (e.g. lists),
which caused a crash in re.search(). Adding str() coercion fixes this.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* update deepsearch qa tasks
* fix: benchmark evaluation reliability (B1-B4)
- B1: auto-detect LoRA adapters and enable vLLM LoRA mode (read base_model from adapter_config.json)
- B2: serialize evaluations with threading.Lock to prevent GPU contention
- B3: cache eval results by model_path to deduplicate concurrent submissions
- B4: propagate error details from OpenCompass to agent (non-numeric scores, load failures)
Made-with: Cursor
* fix(B1): reject LoRA adapter submissions with clear merge instructions
- opencompass.py: detect adapter_config.json and return error with
merge_and_unload() instructions instead of broken vLLM LoRA mode
- instructions.md: add requirement to submit full merged models
- opencompass_template.yaml: remove unused is_lora/lora_path params
Made-with: Cursor
* update chat completion
* update
* update deepsearch
* md
* codex + benchmark update
* codex + benchmark update
* codex + benchmark update
* codex
* codex
* fix: grading server cache key includes mtime to detect model overwrites
Previously cache used only resolved_path, so overwritten models at the
same path returned stale scores. Now cache key = path@max_mtime so
re-evaluation is triggered when model files change.
Made-with: Cursor
* feat: add gemini/claude agent scaffolds, fix codex binary path
- codex/start.sh: use CODEX_BIN env var instead of bare 'codex'
- Add gemini/ and claude/ agent directories with config.yaml and start.sh
Made-with: Cursor
* chore: remove copied human_readable_trace.py from PostTrainBench
Made-with: Cursor
* update evaluation
* benchmark
* Fix log cleanup and OpenHands env
* fix: webshop env pth problem
* benchmark alpacaeval
* style(rl): apply auto-lint fixes
* fix(rl): address CI and CodeQL issues
* fix(rl): make autorl bench imports CI-safe
---------
Co-authored-by: Qizheng Li <jenssenlee@163.com>
Co-authored-by: Young <afe.young@gmail.com>
Co-authored-by: Bowen Xian <xianbowen@outlook.com>
Co-authored-by: chelsea97 <zhuowbrown@gmail.com>
Co-authored-by: Xu Yang <peteryang@vip.qq.com>
Co-authored-by: sakura657 <yctangcse@gmail.com>
Co-authored-by: shatianming5 <tianming.sha@stonybrook.edu>
Co-authored-by: Yeyuqing0913 <shatianming4@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
TPTBusiness