mirror/OpenLLM
1
0
Fork 0
mirror of https://github.com/bentoml/OpenLLM.git synced 2026-06-12 02:20:32 -04:00
Code Issues Packages Projects Releases Wiki Activity

chore(style): add one blank line

Browse Source 
to conform with Google style

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
This commit is contained in:
aarnphm-ec2-dev
2023-08-26 11:36:57 +00:00
parent 938fd362bb
commit 806a663e4a
111 changed files with 601 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
openllm-python/src/openllm/_quantisation.py
View File

@@ -14,12 +14,15 @@ autogptq, torch, transformers = LazyLoader('autogptq', globals(), 'auto_gptq'),
logger = logging.getLogger(__name__)
QuantiseMode = t.Literal['int8', 'int4', 'gptq']
@overload
def infer_quantisation_config(cls: type[LLM[t.Any, t.Any]], quantise: t.Literal['int8', 'int4'], **attrs: t.Any) -> tuple[transformers.BitsAndBytesConfig, DictStrAny]:
...
@overload
def infer_quantisation_config(cls: type[LLM[t.Any, t.Any]], quantise: t.Literal['gptq'], **attrs: t.Any) -> tuple[autogptq.BaseQuantizeConfig, DictStrAny]:
...
def infer_quantisation_config(cls: type[LLM[t.Any, t.Any]], quantise: QuantiseMode, **attrs: t.Any) -> tuple[transformers.BitsAndBytesConfig | autogptq.BaseQuantizeConfig, DictStrAny]:
# 8 bit configuration
int8_threshold = attrs.pop('llm_int8_threshhold', 6.0)