[Bugfix] Allow vllm to still work if triton is not installed. (#6786)

Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
This commit is contained in:
Thomas Parnell
2024-07-29 14:51:27 -07:00
committed by GitHub
parent 7f8d612d24
commit 9a7e2d0534
13 changed files with 65 additions and 37 deletions
+1 -1
View File
@@ -5,9 +5,9 @@ from typing import Dict, List, Optional, Tuple
import torch
from vllm.model_executor.layers.ops.sample import get_num_triton_sampler_splits
from vllm.sampling_params import SamplingParams, SamplingType
from vllm.sequence import SequenceData, SequenceGroupMetadata
from vllm.triton_utils.sample import get_num_triton_sampler_splits
from vllm.utils import (async_tensor_h2d, is_pin_memory_available,
make_tensor_with_pad, maybe_expand_dim)