ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
unsloth/tests/studio
History
Daniel Han 55c392ff7c
 studio: fix sentence-transformers RAG embedder on Windows ROCm (torchao) (#6608) 
torchao has no working Windows ROCm build. transformers.quantizers imports it,
and it loads torch's c10d distributed backend at module level, which the AMD
Windows wheels omit (no RCCL). The import aborts, transformers can no longer
expose PreTrainedModel, and the sentence-transformers embedder silently falls
back to the llama-server GGUF embedder. Linux ROCm and NVIDIA are unaffected
(the c10d ops are present / torchao is real there).

The training and export workers already install the shared torchao stub before
importing transformers, but the RAG embedder runs in the main backend process,
which never did. Two fixes, both no-ops off Windows ROCm:

- embeddings.py: install_torchao_windows_rocm_stub() before the first
  sentence-transformers import, so an already-installed torchao is neutralized
  (fixes existing venvs).
- install_python_stack.py: stop installing torchao on Windows ROCm; it can only
  crash on import there, so new venvs never ship it.

Add tests covering the embedder stub call and the install skip.
2026-06-23 05:39:02 -07:00
..
install studio: fix sentence-transformers RAG embedder on Windows ROCm (torchao) (#6608) 2026-06-23 05:39:02 -07:00
load_freeze Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
_playwright_robust.py Harden evaluate_fetch against execution-context-destroyed during navigation (#6549) 2026-06-21 21:57:25 -07:00
playwright_chat_ime_i18n.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
playwright_chat_ui.py Studio Playwright: snooze update banner before sending (#6576) 2026-06-22 08:27:18 -07:00
playwright_extra_ui.py Harden evaluate_fetch against execution-context-destroyed during navigation (#6549) 2026-06-21 21:57:25 -07:00
run_real_mlx_smoke.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
studio_api_smoke.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_auth_form_input_count.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_cancel_atomicity.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_cancel_id_wiring.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_chat_preset_builtin_invariants.py Formatting: ruff line-length 100, kwarg-spacing passes, drop blank after short local imports (#6079) 2026-06-08 04:24:13 -07:00
test_cli_repo_variant.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_cli_run_alias.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_cli_studio_defaults.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_cli_studio_stop_windows.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_composer_rtl_bidi_attribute.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_export_output_path_contract.py Formatting: ruff line-length 100, kwarg-spacing passes, drop blank after short local imports (#6079) 2026-06-08 04:24:13 -07:00
test_frontend_dep_removal.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_hardware_dispatch_matrix.py Studio macOS: faster startup, MLX self-heal, drop obsolete prebuilt pins (#6494) 2026-06-22 02:20:08 -07:00
test_is_mlx_dispatch_gate.py Studio macOS: faster startup, MLX self-heal, drop obsolete prebuilt pins (#6494) 2026-06-22 02:20:08 -07:00
test_llama_cpp_wall_clock_cap.py Studio: extend llama.cpp first-token timeout (#5841) 2026-06-12 18:41:38 +02:00
test_mlx_training_worker_behaviors.py MLX training support for Studio on Apple Silicon (#5340) 2026-05-14 05:24:20 -07:00
test_node_decision.ps1 Studio: use an isolated Node.js for the frontend build instead of replacing the system Node/npm (#6533) 2026-06-21 21:17:29 -07:00
test_node_probe_guard.ps1 Studio: use an isolated Node.js for the frontend build instead of replacing the system Node/npm (#6533) 2026-06-21 21:17:29 -07:00
test_resolve_cuda_toolkit.ps1 Windows installer: fix DiskPart UAC mid-install, drive-root cache, and spurious unsloth.exe rename warning (#6296) 2026-06-22 03:09:08 -07:00
test_stream_cancel_registration_timing.py studio: run /generate/stream's sync generator off the event loop to avoid blocking it (#6466) 2026-06-19 14:19:04 +01:00
test_studio_gguf_export_script_pin.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_studio_text_descender_clipping.py Reduce and tighten comments and docstrings across the test suite (#6429) 2026-06-18 01:07:09 -07:00
test_sync_allow_scripts_pins.py Studio: auto-sync allowScripts pins after dependency bumps (#6136) 2026-06-10 02:35:37 -07:00
test_torch_flavor.ps1 Windows installer: fix DiskPart UAC mid-install, drive-root cache, and spurious unsloth.exe rename warning (#6296) 2026-06-22 03:09:08 -07:00
test_uninstall_dual_install_icon.ps1 Windows installer: fix DiskPart UAC mid-install, drive-root cache, and spurious unsloth.exe rename warning (#6296) 2026-06-22 03:09:08 -07:00