Download source/assets.py from FluidInference/kev-0.6b-coreml: direct link, hf CLI and curl.
- Browser
- Download file 1.13 kB
-
https://huggingface.co/FluidInference/kev-0.6b-coreml/resolve/main/source/assets.py
- Command line
-
hf download hf://FluidInference/kev-0.6b-coreml/source/assets.py
-
curl -L -o assets.py https://huggingface.co/FluidInference/kev-0.6b-coreml/resolve/main/source/assets.py
1.13 kB
| """Download the exact Kev adapter and Qwen base revisions used by this conversion.""" | |
| from __future__ import annotations | |
| import hashlib | |
| import json | |
| from dataclasses import replace | |
| from pathlib import Path | |
| from huggingface_hub import snapshot_download | |
| from kev.checkpoint import Checkpoint, LoadOptions | |
| ROOT = Path(__file__).resolve().parent | |
| LOCK_PATH = ROOT / "assets.lock.json" | |
| def load_model(): | |
| lock = json.loads(LOCK_PATH.read_text()) | |
| checkpoint = lock["checkpoint"] | |
| base = lock["base"] | |
| checkpoint_path = snapshot_download(checkpoint["repo"], revision=checkpoint["revision"]) | |
| for name, expected in checkpoint["files"].items(): | |
| actual = hashlib.sha256((Path(checkpoint_path) / name).read_bytes()).hexdigest() | |
| if actual != expected: | |
| raise ValueError(f"trained Kev file hash mismatch: {name}") | |
| base_path = snapshot_download(base["repo"], revision=base["revision"]) | |
| loader = Checkpoint(checkpoint_path) | |
| loader.meta = replace(loader.meta, base=base_path, base_revision=None) | |
| tokenizer, model = loader.load("cpu", LoadOptions()) | |
| return lock, tokenizer, model | |