Download eval/local_intelligence/run_hf_matrix.py from arach/training-lab: direct link, hf CLI and curl.
- Browser
- Download file 2.34 kB
-
https://huggingface.co/arach/training-lab/resolve/main/eval/local_intelligence/run_hf_matrix.py
- Command line
-
hf download hf://arach/training-lab/eval/local_intelligence/run_hf_matrix.py
-
curl -L -o run_hf_matrix.py https://huggingface.co/arach/training-lab/resolve/main/eval/local_intelligence/run_hf_matrix.py
2.34 kB
| #!/usr/bin/env python3 | |
| from __future__ import annotations | |
| import argparse | |
| import subprocess | |
| import sys | |
| from pathlib import Path | |
| DEFAULT_MODELS = [ | |
| "google/gemma-4-E4B-it", | |
| "nvidia/Llama-3.1-Nemotron-Nano-4B-v1.1", | |
| "Qwen/Qwen3.5-4B-Instruct-2507", | |
| ] | |
| def parse_args() -> argparse.Namespace: | |
| parser = argparse.ArgumentParser(description="Run the local-intelligence eval pack across a small HF model matrix.") | |
| parser.add_argument("--models", nargs="*", default=DEFAULT_MODELS) | |
| parser.add_argument("--limit", type=int) | |
| parser.add_argument("--tier") | |
| parser.add_argument("--trust-remote-code", action="store_true") | |
| parser.add_argument("--max-new-tokens", type=int, default=256) | |
| parser.add_argument("--dtype", default="auto") | |
| parser.add_argument("--device-map", default="auto") | |
| parser.add_argument("--attn-implementation") | |
| parser.add_argument("--results-dir") | |
| parser.add_argument("--keep-going", action="store_true") | |
| return parser.parse_args() | |
| def main() -> int: | |
| args = parse_args() | |
| script = Path(__file__).with_name("run_hf_transformers.py") | |
| failures = 0 | |
| for model in args.models: | |
| cmd = [ | |
| sys.executable, | |
| str(script), | |
| "--model", | |
| model, | |
| "--output-prefix", | |
| "hf-matrix", | |
| "--max-new-tokens", | |
| str(args.max_new_tokens), | |
| "--dtype", | |
| args.dtype, | |
| "--device-map", | |
| args.device_map, | |
| ] | |
| if args.limit is not None: | |
| cmd += ["--limit", str(args.limit)] | |
| if args.tier: | |
| cmd += ["--tier", args.tier] | |
| if args.attn_implementation: | |
| cmd += ["--attn-implementation", args.attn_implementation] | |
| if args.trust_remote_code: | |
| cmd.append("--trust-remote-code") | |
| if args.results_dir: | |
| result_name = model.lower().replace("/", "-").replace(".", "-") + ".json" | |
| cmd += ["--output", str(Path(args.results_dir) / result_name)] | |
| print(f"\n=== Running {model} ===") | |
| result = subprocess.run(cmd) | |
| if result.returncode != 0: | |
| failures += 1 | |
| if not args.keep_going: | |
| return result.returncode | |
| return 1 if failures else 0 | |
| if __name__ == "__main__": | |
| raise SystemExit(main()) | |