CookieApril/enginex-ascend-910-vllm
1
0
Fork 0
forked from EngineX-Ascend/enginex-ascend-910-vllm
Code Pull Requests Activity

init v0.11.0rc0

Browse Source
This commit is contained in:
luopingyi
2025-10-14 10:38:28 +08:00
parent 67afd0ea78
commit 66dc16f966
278 changed files with 28130 additions and 11708 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
vllm_ascend/torchair/models/torchair_deepseek_mtp.py
View File

@@ -27,14 +27,12 @@ from vllm.config import CacheConfig, ModelConfig, VllmConfig
from vllm.model_executor.layers.layernorm import RMSNorm
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.quantization import QuantizationConfig
from vllm.model_executor.layers.sampler import get_sampler
from vllm.model_executor.layers.vocab_parallel_embedding import (
ParallelLMHead, VocabParallelEmbedding)
from vllm.model_executor.models.deepseek_mtp import (
DeepSeekMTP, DeepSeekMultiTokenPredictor, DeepSeekMultiTokenPredictorLayer,
SharedHead)
from vllm.model_executor.models.utils import maybe_prefix
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import IntermediateTensors
from vllm_ascend.torchair.models.torchair_deepseek_v2 import \
@@ -172,7 +170,7 @@ class TorchairDeepSeekMultiTokenPredictor(DeepSeekMultiTokenPredictor):
def compute_logits(
self,
hidden_states: torch.Tensor,
sampling_metadata: SamplingMetadata,
sampling_metadata=None, # type: ignore
spec_step_idx: int = 0,
) -> torch.Tensor:
current_step_idx = (spec_step_idx % self.num_mtp_layers)
@@ -199,8 +197,6 @@ class TorchairDeepSeekMTP(DeepSeekMTP):
self.model = TorchairDeepSeekMultiTokenPredictor(
vllm_config=vllm_config, prefix=maybe_prefix(prefix, "model"))
self.sampler = get_sampler()
def forward(
self,
input_ids: torch.Tensor,