A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 04:30 UTC
[Core] Update to outlines >= 0.1.8 (#10576)
Signed-off-by: Russell Bryant <rbryant@redhat.com>
Russell Bryant committed
Dec 10, 2024 at 15:08 UTC
e7391949267a4eff3d84f02119f442f46b16d163
2 files changed
+2
-2
requirements-common.txt
+1
-1
@@ -18,7 +18,7 @@ prometheus_client >= 0.18.0
18
prometheus-fastapi-instrumentator >= 7.0.0
19
tiktoken >= 0.6.0 # Required for DBRX tokenizer
20
lm-format-enforcer >= 0.10.9, < 0.11
21
-outlines >= 0.0.43, < 0.1
21
+outlines >= 0.1.8
22
xgrammar >= 0.1.6; platform_machine == "x86_64"
23
typing_extensions >= 4.10
24
filelock >= 3.16.1 # need to contain https://github.com/tox-dev/filelock/pull/317
vllm/model_executor/guided_decoding/outlines_logits_processors.py
+1
-1
@@ -99,7 +99,7 @@ class RegexLogitsProcessor(BaseLogitsProcessor):
99
def _get_guide(cls, regex_string: str,
100
tokenizer: PreTrainedTokenizerBase) -> Guide:
101
tokenizer = _adapt_tokenizer(tokenizer)
102
- return RegexGuide(regex_string, tokenizer)
102
+ return RegexGuide.from_regex(regex_string, tokenizer)
103
104
def __init__(self, regex_string: str, tokenizer: PreTrainedTokenizerBase):
105
"""Compile the FSM that drives the regex-structured generation.