fix(backend): honour VOICEBOX_FORCE_CPU in device selection

The override is documented in gpu-acceleration.mdx and listed as step 1 of
the get_torch_device() precedence in tts-generation.mdx, but grepping the
tree for VOICEBOX_FORCE_CPU matched only those two doc files - nothing read
it. Users whose GPU has no compiled kernels in the bundled PyTorch had no
way to fall back to CPU short of renaming the installed CUDA backend
directory.

Resolve it before torch is imported, so it still works when the installed
build is itself the reason CPU is wanted.

backend/backends/base.py

@@ -6,6 +6,7 @@ voice prompt combination, and model loading progress tracking.
 """
 
 import logging
+import os
 import platform
 from contextlib import contextmanager
 from pathlib import Path
@@ -77,6 +78,13 @@ def is_model_cached(
         return False
 
 
+# Documented escape hatch (docs/content/docs/overview/gpu-acceleration.mdx):
+# users whose GPU has no compiled kernels in the bundled PyTorch set this to run
+# on CPU instead of crashing at generation time.
+FORCE_CPU_ENV_VAR = "VOICEBOX_FORCE_CPU"
+FORCE_CPU_ENABLED_VALUE = "1"
+
+
 def get_torch_device(
     *,
     allow_xpu: bool = False,
@@ -92,7 +100,16 @@ def get_torch_device(
         allow_directml: Check for DirectML (Windows) support.
         allow_mps: Allow MPS (Apple Silicon). If False, MPS falls back to CPU.
         force_cpu_on_mac: Force CPU on macOS regardless of GPU availability.
+
+    The VOICEBOX_FORCE_CPU override wins over every other candidate, and is
+    resolved before torch is imported so it still works when the installed
+    build is the reason CPU is wanted.
     """
+    # Stripped: on Windows, where this override matters most, it is usually set
+    # through the GUI environment editor.
+    if os.environ.get(FORCE_CPU_ENV_VAR, "").strip() == FORCE_CPU_ENABLED_VALUE:
+        return "cpu"
+
     if force_cpu_on_mac and platform.system() == "Darwin":
         return "cpu"