fix(deps): make torch/ML stack opt-in instead of a mandatory install

requirements-amd.txt/requirements-nvidia.txt were pulling a multi-GB
torch wheel by default even though nothing in synapse/ imports torch,
transformers, accelerate, bitsandbytes, or PySide6 - dead weight that
made the pip batch fragile (one failed download could take unrelated
base deps down with it on a slow connection). Split the unused ML/GUI
stack out of requirements-base.txt into a new opt-in requirements-ml.txt,
and dropped the torch lines from the AMD/NVIDIA overlays and generator.

Also: recreate the venv if it exists but pip is missing, instead of
silently reusing a half-built one (ensurepip can fail during venv
creation and leave an interpreter with no pip).

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

bin/sync.py

@@ -61,6 +61,16 @@ def venv_python() -> Path:
     """Path to the Promethean interpreter, creating the venv if it's missing."""
     venv = ROOT / "Promethean"
     py = venv / ("Scripts/python.exe" if os.name == "nt" else "bin/python")
+    if py.exists() and subprocess.run(
+            [str(py), "-m", "pip", "--version"], capture_output=True).returncode:
+        # venv creation can partially succeed: the interpreter gets built but
+        # ensurepip's bootstrap fails (e.g. the matching pythonX.Y-venv package
+        # wasn't installed yet), leaving pip missing. Existence of `py` alone
+        # can't tell a venv like that apart from a good one, so a prior failed
+        # run would otherwise be reused forever instead of getting rebuilt now
+        # that whatever broke ensurepip is fixed.
+        print("Existing venv has no pip - recreating it...")
+        shutil.rmtree(venv)
     if not py.exists():
         result = subprocess.run([sys.executable, "-m", "venv", str(venv)])
         if result.returncode: