The open-source AI voice studio. Clone, dictate, create.
chore(release): Docker on Python 3.12, fix numpy shim _t NameError, changelog line
jamiepine6955272e79e0
diff --git a/CHANGELOG.md b/CHANGELOG.mdindex a4cacaf4..c04b3857 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -63,6 +63,11 @@### Linux+- **The Docker image now runs on Python 3.12.** The `Dockerfile` based both build+ stages on `python:3.11-slim` while the backend declares `>=3.12,<3.13`, so the+ container ran an interpreter the desktop builds and `just setup` never use. Both+ stages now use `python:3.12-slim`, matching the release workflow and the venv+ pin above.- **ROCm setup works on Linux AMD systems.** Docker ROCm builds now keep PyTorchon the ROCm wheel index during dependency installation, so later installs donot replace it with CUDA wheels. The ROCm compose overlay no longer assumesdiff --git a/Dockerfile b/Dockerfileindex adf8c7ea..14d86b9b 100644--- a/Dockerfile+++ b/Dockerfile@@ -32,7 +32,7 @@ RUN cd web && bunx --bun vite build# === Stage 2: Build Python dependencies ===-FROM python:3.11-slim AS backend-builder+FROM python:3.12-slim AS backend-builder# Re-declare ARG inside the stage (Docker scoping requirement).ARG PYTORCH_VARIANT=cpu@@ -68,7 +68,7 @@ RUN pip install --no-cache-dir --prefix=/install \# === Stage 3: Runtime ===-FROM python:3.11-slim+FROM python:3.12-slim# Create non-root user; the entrypoint joins GPU device groups at runtime.RUN groupadd -r voicebox && \diff --git a/backend/pyi_rth_numpy_compat.py b/backend/pyi_rth_numpy_compat.pyindex 683f4dea..d682997b 100644--- a/backend/pyi_rth_numpy_compat.py+++ b/backend/pyi_rth_numpy_compat.py@@ -53,17 +53,17 @@ def _patch_torch_from_numpy():# silently corrupt data (e.g. fp16 tensors from some TTS engines),# so we raise instead.dtype_map = {- "float16": _t.float16,- "float32": _t.float32,- "float64": _t.float64,- "int8": _t.int8,- "int16": _t.int16,- "int32": _t.int32,- "int64": _t.int64,- "uint8": _t.uint8,- "bool": _t.bool,- "complex64": _t.complex64,- "complex128": _t.complex128,+ "float16": torch.float16,+ "float32": torch.float32,+ "float64": torch.float64,+ "int8": torch.int8,+ "int16": torch.int16,+ "int32": torch.int32,+ "int64": torch.int64,+ "uint8": torch.uint8,+ "bool": torch.bool,+ "complex64": torch.complex64,+ "complex128": torch.complex128,}def _safe_from_numpy(