ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

[pre-commit.ci] auto fixes from pre-commit.com hooks

Browse source 
for more information, see https://pre-commit.ci
This commit is contained in:
pre-commit-ci[bot] 2026-07-26 14:37:36 +00:00
commit 9dee5ac2d0
2 changed files with 6 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

4
studio/backend/routes/inference.py
View file

@ -3747,9 +3747,7 @@ async def _maybe_auto_switch_model(
)
)
saved_gpu_ids = load_kwargs.get("gpu_ids")
if saved_gpu_ids and not _override_gpu_ids_still_resolve(
saved_gpu_ids
):
if saved_gpu_ids and not _override_gpu_ids_still_resolve(saved_gpu_ids):
# A pin saved before a GPU was removed, before a
# visibility-mask change, or on another host. Dropping the
# one dead field beats 400ing the whole load.

20
studio/backend/tests/test_openai_auto_switch.py
View file

@ -4036,9 +4036,7 @@ def test_first_quant_save_keeps_legacy_bare_repo_launch_flags(monkeypatch):
settings.set_model_override("unsloth/B-GGUF", llama_extra_args = ["--flash-attn"])
resp = settings_route.update_openai_auto_switch_override(
settings_route.ModelOverridePayload(
model_id = "unsloth/B-GGUF:Q4_K_M", max_seq_length = 4096
),
settings_route.ModelOverridePayload(model_id = "unsloth/B-GGUF:Q4_K_M", max_seq_length = 4096),
"tester",
)
entry = resp.overrides["unsloth/B-GGUF:Q4_K_M"]
@ -4055,9 +4053,7 @@ def test_bare_repo_carry_over_does_not_split_a_windows_path(monkeypatch):
settings.set_model_override("C", llama_extra_args = ["--flash-attn"])
resp = settings_route.update_openai_auto_switch_override(
settings_route.ModelOverridePayload(
model_id = r"C:\models\x.gguf", max_seq_length = 4096
),
settings_route.ModelOverridePayload(model_id = r"C:\models\x.gguf", max_seq_length = 4096),
"tester",
)
assert "llama_extra_args" not in resp.overrides[r"C:\models\x.gguf"]
@ -4095,9 +4091,7 @@ def test_stale_gpu_ids_are_dropped_not_fatal(monkeypatch):
"get_model_override",
lambda mid: {"gpu_ids": [0, 1], "max_seq_length": 4096},
)
monkeypatch.setattr(
inference_route, "_override_gpu_ids_still_resolve", lambda ids: False
)
monkeypatch.setattr(inference_route, "_override_gpu_ids_still_resolve", lambda ids: False)
_run_hook("unsloth/B-GGUF")
req = rec.calls[0]
@ -4116,12 +4110,8 @@ def test_usable_gpu_ids_are_kept(monkeypatch):
backend = backend,
recorder = rec,
)
monkeypatch.setattr(
settings, "get_model_override", lambda mid: {"gpu_ids": [0, 1]}
)
monkeypatch.setattr(
inference_route, "_override_gpu_ids_still_resolve", lambda ids: True
)
monkeypatch.setattr(settings, "get_model_override", lambda mid: {"gpu_ids": [0, 1]})
monkeypatch.setattr(inference_route, "_override_gpu_ids_still_resolve", lambda ids: True)
_run_hook("unsloth/B-GGUF")
assert rec.calls[0].gpu_ids == [0, 1]