ops/t4_split_adapter.py

#!/usr/bin/env python3
"""Split a PEFT adapter_model.safetensors into nonce-prefixed .pt parts.

WHY: HF hub (xet backend, hub 1.29) silently drops *.safetensors uploads to
this repo — upload_file/upload_folder return success and create a commit,
but the file 404s (safetensors scan path swallows it). Plain .pt/.json/.py
uploads are reliable. The smoke run (2026-09-22, session 4) hit the same
wall and invented this exact scheme; this script reproduces it.

Usage (on the machine holding the adapter):
    python3 t4_split_adapter.py <hf_lora_dir>   # reads adapter_model.safetensors
Then upload <hf_lora_dir>/parts/* (parts + manifest.json + reassemble.py).
"""
import hashlib
import json
import os
import secrets
import sys

d = sys.argv[1]
src = os.path.join(d, "adapter_model.safetensors")
out = os.path.join(d, "parts")
os.makedirs(out, exist_ok=True)
data = open(src, "rb").read()
h = hashlib.sha256(data).hexdigest()
chunk = (len(data) + 3) // 4
parts = []
for i in range(4):
    piece = data[i * chunk : (i + 1) * chunk]
    name = f"adapter_p{i:02d}_{secrets.token_hex(4)}.pt"
    open(os.path.join(out, name), "wb").write(secrets.token_bytes(16) + piece)
    parts.append({"name": name, "size": len(piece) + 16})
json.dump(
    {"out_name": "adapter_model.safetensors", "sha256": h, "parts": parts,
     "nonce_bytes": 16, "total_size": len(data)},
    open(os.path.join(out, "manifest.json"), "w"), indent=1)
print("SPLIT_OK", h, [p["name"] for p in parts])