EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

upgrade to vllm 0.11.2 (#4400)

Browse Source 
Bump vLLM version to v0.11.2

What's broken and changed by vLLM:
1. structured_output is broken by
https://github.com/vllm-project/vllm/pull/26866
2. get_mrope_input_positions is broken by
https://github.com/vllm-project/vllm/pull/28399
3. graph mode is broken by
https://github.com/vllm-project/vllm/pull/25110 we'll upgrade torch to
2.8 to fix the problem later
4. embedding is broken by
https://github.com/vllm-project/vllm/pull/27583
5. `get_attn_backend_cls` and attention backend is broken are broken by
https://github.com/vllm-project/vllm/pull/28534
6. spec decode is broken by
https://github.com/vllm-project/vllm/pull/28771
7. sp feature is broken by
https://github.com/vllm-project/vllm/pull/27126
8. mtp is broken by https://github.com/vllm-project/vllm/pull/27922
9. lora is broken by https://github.com/vllm-project/vllm/pull/21068
10. execute_model is broken by
https://github.com/vllm-project/vllm/pull/26866
11. `VLLM_DISABLE_SHARED_EXPERTS_STREAM` env is broken by
https://github.com/vllm-project/vllm/pull/28159
12. kv cahe is broken by https://github.com/vllm-project/vllm/pull/27753
13. dp is broken by https://github.com/vllm-project/vllm/pull/25110

 
What's broken and changed by ourself:
1. qwen vl is broken by https://github.com/vllm-project/vllm/pull/28455
We'll remove model files in the future to avoid this kind of error
2. Engine core is broken by
https://github.com/vllm-project/vllm/pull/23691 We'll remove the patch
file in the future.
3. Ascend scheduler is broken by
https://github.com/vllm-project/vllm/pull/28733 We'll remove ascend
scheudler later.
4. qwen3-next is broken by
https://github.com/vllm-project/vllm/pull/28083 We'll remove model files
in the future to avoid this kind of error
5. qwen vl is broken by https://github.com/vllm-project/vllm/pull/27764.
We'll remove model files in the future

Known issue:
1. ray doesn't work 
2. the accuracy of qwen3-next is not correct
3. qwen3-vl is broken
4. prefix cache+ ascend scheduler + deepseek v2 lite is broken.

Co-authored-by: MengqingCao <cmq0113@163.com>
Co-authored-by: hfadzxy <starmoon_zhang@163.com>
Co-authored-by: leo-pony <nengjunma@outlook.com>
Co-authored-by: 22dimensions <waitingwind@foxmail.com>
Co-authored-by: shen-shanshan <467638484@qq.com>


- vLLM version: v0.11.2

---------

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Signed-off-by: MengqingCao <cmq0113@163.com>
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
Signed-off-by: leo-pony <nengjunma@outlook.com>
Co-authored-by: MengqingCao <cmq0113@163.com>
Co-authored-by: hfadzxy <starmoon_zhang@163.com>
Co-authored-by: leo-pony <nengjunma@outlook.com>
This commit is contained in:
wangxiyuan
2025-11-26 11:48:58 +08:00
committed by GitHub
parent d5f77f14d0
commit bc69d7cfe1
54 changed files with 744 additions and 437 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
vllm_ascend/models/qwen2_5_vl.py
View File

@@ -40,7 +40,6 @@ from vllm.model_executor.models.qwen2_5_vl import (
Qwen2_5_VLDummyInputsBuilder, Qwen2_5_VLForConditionalGeneration,
Qwen2_5_VLMultiModalProcessor, Qwen2_5_VLProcessingInfo)
from vllm.model_executor.models.utils import maybe_prefix
from vllm.model_executor.models.vision import conv3d_to_linear_weight
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm_ascend.ascend_forward_context import set_ascend_forward_context
@@ -144,8 +143,14 @@ class AscendQwen2_5_VisionBlock(Qwen2_5_VisionBlock):
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
super().__init__(dim, num_heads, mlp_hidden_dim, act_fn, norm_layer,
quant_config, prefix)
super().__init__(dim=dim,
num_heads=num_heads,
mlp_hidden_dim=mlp_hidden_dim,
act_fn=act_fn,
norm_layer=norm_layer,
quant_config=quant_config,
prefix=prefix)
self.attn = AscendQwen2_5_VisionAttention(embed_dim=dim,
num_heads=num_heads,
projection_size=dim,
@@ -161,14 +166,6 @@ class AscendQwen2_5_VisionBlock(Qwen2_5_VisionBlock):
return x
class AscendQwen2_5_VisionPatchEmbed(Qwen2_5_VisionPatchEmbed):
def forward(self, x: torch.Tensor) -> torch.Tensor:
x = x.matmul(
self.proj.weight.data.view(self.hidden_size, -1).transpose(0, 1))
return x
class AscendQwen2_5_VisionRotaryEmbedding(Qwen2_5_VisionRotaryEmbedding):
def __init__(self, dim: int, theta: float = 10000.0) -> None:
@@ -195,7 +192,7 @@ class AscendQwen2_5_VisionTransformer(Qwen2_5_VisionTransformer):
head_dim = self.hidden_size // self.num_heads
self.rotary_pos_emb = AscendQwen2_5_VisionRotaryEmbedding(head_dim //
2)
self.patch_embed = AscendQwen2_5_VisionPatchEmbed(
self.patch_embed = Qwen2_5_VisionPatchEmbed(
patch_size=vision_config.patch_size,
temporal_patch_size=vision_config.temporal_patch_size,
in_channels=vision_config.in_channels,
@@ -357,8 +354,6 @@ class AscendQwen2_5_VisionTransformer(Qwen2_5_VisionTransformer):
params_dict = dict(self.named_parameters(remove_duplicate=False))
loaded_params: Set[str] = set()
for name, loaded_weight in weights:
if name.endswith("patch_embed.proj.weight"):
loaded_weight = conv3d_to_linear_weight(loaded_weight)
for (param_name, weight_name, shard_id) in stacked_params_mapping:
if weight_name not in name:
continue