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] 2026-05-19 14:57:37 +00:00 • committed by Daniel Han
commit d8ed88e968
2 changed files with 15 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

4
studio/backend/core/inference/llama_cpp.py
View file

@ -3527,9 +3527,7 @@ class LlamaCppBackend:
)
return flags
flags.extend(["--spec-type", effective_mode])
flags.extend(
_build_ngram_map_k_flags(map_caps, variant = effective_mode)
)
flags.extend(_build_ngram_map_k_flags(map_caps, variant = effective_mode))
self._speculative_type = effective_mode
logger.info(f"Spec decoding: {effective_mode}")
return flags

21
studio/backend/routes/inference.py
View file

@ -515,7 +515,8 @@ def _request_matches_loaded_settings(
and request.spec_draft_p_min is not None
and abs(
float(request.spec_draft_p_min) - (llama_backend.spec_draft_p_min or 0.0)
) > 1e-6
)
> 1e-6
):
return False
if (request.chat_template_override or None) != (
@ -657,7 +658,8 @@ async def load_model(
supports_tools = llama_backend.supports_tools,
chat_template = llama_backend.chat_template,
speculative_type = llama_backend.requested_spec_mode,
spec_draft_n_max = llama_backend.spec_draft_n_max, spec_draft_p_min = llama_backend.spec_draft_p_min,
spec_draft_n_max = llama_backend.spec_draft_n_max,
spec_draft_p_min = llama_backend.spec_draft_p_min,
)
else:
if (
@ -788,7 +790,8 @@ async def load_model(
strip_cache = "cache_type_kv" in fields_set,
strip_spec = (
"speculative_type" in fields_set
or "spec_draft_n_max" in fields_set or "spec_draft_p_min" in fields_set
or "spec_draft_n_max" in fields_set
or "spec_draft_p_min" in fields_set
),
strip_template = "chat_template_override" in fields_set,
)
@ -830,7 +833,8 @@ async def load_model(
chat_template_override = request.chat_template_override,
cache_type_kv = request.cache_type_kv,
speculative_type = request.speculative_type,
spec_draft_n_max = request.spec_draft_n_max, spec_draft_p_min = request.spec_draft_p_min,
spec_draft_n_max = request.spec_draft_n_max,
spec_draft_p_min = request.spec_draft_p_min,
n_parallel = _n_parallel,
extra_args = extra_llama_args,
)
@ -854,7 +858,8 @@ async def load_model(
chat_template_override = request.chat_template_override,
cache_type_kv = request.cache_type_kv,
speculative_type = request.speculative_type,
spec_draft_n_max = request.spec_draft_n_max, spec_draft_p_min = request.spec_draft_p_min,
spec_draft_n_max = request.spec_draft_n_max,
spec_draft_p_min = request.spec_draft_p_min,
n_parallel = _n_parallel,
extra_args = extra_llama_args,
)
@ -914,7 +919,8 @@ async def load_model(
cache_type_kv = llama_backend.cache_type_kv,
chat_template = llama_backend.chat_template,
speculative_type = llama_backend.requested_spec_mode,
spec_draft_n_max = llama_backend.spec_draft_n_max, spec_draft_p_min = llama_backend.spec_draft_p_min,
spec_draft_n_max = llama_backend.spec_draft_n_max,
spec_draft_p_min = llama_backend.spec_draft_p_min,
)
# ── Standard path: load via Unsloth/transformers ──────────
@ -1403,7 +1409,8 @@ async def get_status(
cache_type_kv = llama_backend.cache_type_kv,
chat_template_override = llama_backend.chat_template_override,
speculative_type = llama_backend.requested_spec_mode,
spec_draft_n_max = llama_backend.spec_draft_n_max, spec_draft_p_min = llama_backend.spec_draft_p_min,
spec_draft_n_max = llama_backend.spec_draft_n_max,
spec_draft_p_min = llama_backend.spec_draft_p_min,
llama_cpp_supports_mtp = _supports_mtp,
llama_cpp_prebuilt_stale = _stale,
llama_cpp_installed_tag = _installed_tag,