From 62295c25e9933af67ea6be96fa37712298bc9a99 Mon Sep 17 00:00:00 2001
From: Demetrios Agourakis <demetrios@agourakis.med.br>
Date: Tue, 22 Sep 2026 17:18:14 -0300
Subject: [PATCH] fix(scripts): upload README.md in upload_hf.py

The upload-file listing explicitly excluded README.md ("README usually
already uploaded; skip leftovers handled below"), but nothing below
handled leftovers -- the README is simply never uploaded. Given a
checkpoint directory of README.md, config.json, and model.safetensors,
only config.json and model.safetensors were recorded as uploaded.

Removes the exclusion and extracts the listing into files_to_upload()
so the file-selection rule (dotfiles and .bak_vision backups excluded,
everything else included) is independently testable without touching
the network.

Fixes #103.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
---
 scripts/upload_hf.py    | 21 +++++++++++++++------
 tests/test_upload_hf.py | 42 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 57 insertions(+), 6 deletions(-)
 create mode 100644 tests/test_upload_hf.py

diff --git a/scripts/upload_hf.py b/scripts/upload_hf.py
index 56f10e4..3329d02 100644
--- a/scripts/upload_hf.py
+++ b/scripts/upload_hf.py
@@ -76,6 +76,20 @@ def _fix(url, endpoint):
 }
 
 
+def files_to_upload(local_dir: str) -> list[str]:
+    """Every regular file in ``local_dir`` worth uploading, sorted.
+
+    Excludes dotfiles and ``.bak_vision`` backups left by
+    scripts/strip_vision_weights.py; everything else — including
+    README.md — is uploaded.
+    """
+    names = sorted(
+        f for f in os.listdir(local_dir)
+        if not f.startswith(".") and ".bak_vision" not in f
+    )
+    return [n for n in names if os.path.isfile(os.path.join(local_dir, n))]
+
+
 def main():
     tier = sys.argv[1] if len(sys.argv) > 1 else "edge0-8b"
     env_name, repo_id = REPOS[tier]
@@ -89,12 +103,7 @@ def main():
         raise SystemExit(f"checkpoint directory not found: {local_dir}")
     api = HfApi()
     print(f"uploading {local_dir} -> {repo_id} via {os.environ['HF_ENDPOINT']} (no xet)", flush=True)
-    names = sorted(
-        f for f in os.listdir(local_dir)
-        if not f.startswith(".") and ".bak_vision" not in f and f != "README.md"
-    )
-    # README usually already uploaded; skip leftovers handled below
-    files = [n for n in names if os.path.isfile(os.path.join(local_dir, n))]
+    files = files_to_upload(local_dir)
     for i, name in enumerate(files):
         path = os.path.join(local_dir, name)
         print(f"[{i+1}/{len(files)}] {name} ({os.path.getsize(path)} bytes)", flush=True)
diff --git a/tests/test_upload_hf.py b/tests/test_upload_hf.py
new file mode 100644
index 0000000..270ed4d
--- /dev/null
+++ b/tests/test_upload_hf.py
@@ -0,0 +1,42 @@
+"""scripts/upload_hf.py file-selection logic.
+
+Pure filesystem logic, no MLX and no network — the upload itself
+(``HfApi.upload_file``) is out of scope for unit tests.
+"""
+
+from __future__ import annotations
+
+import importlib.util
+import sys
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[1]
+
+
+def _load_upload_hf():
+    """Import scripts/upload_hf.py without running main() or touching
+    the network — the module patches httpx at import time, which is
+    fine (pure monkeypatch, no I/O)."""
+    spec = importlib.util.spec_from_file_location(
+        "upload_hf", ROOT / "scripts" / "upload_hf.py")
+    mod = importlib.util.module_from_spec(spec)
+    sys.modules[spec.name] = mod
+    spec.loader.exec_module(mod)
+    return mod
+
+
+def test_files_to_upload_includes_readme(tmp_path):
+    for name in ("README.md", "config.json", "model.safetensors"):
+        (tmp_path / name).write_text("x")
+    mod = _load_upload_hf()
+    assert mod.files_to_upload(str(tmp_path)) == [
+        "README.md", "config.json", "model.safetensors"]
+
+
+def test_files_to_upload_excludes_dotfiles_and_vision_backups(tmp_path):
+    for name in ("README.md", "model.safetensors",
+                 "model.safetensors.bak_vision", ".DS_Store"):
+        (tmp_path / name).write_text("x")
+    mod = _load_upload_hf()
+    assert mod.files_to_upload(str(tmp_path)) == [
+        "README.md", "model.safetensors"]