[Feat][BugFix]Support the Qwen3-Next-80B-A3B-Instruct quantization model&Fix the NZ issue (#4245)

### What this PR does / why we need it?
Support the Qwen3-Next-80B-A3B-Instruct quantization model and Fix the
NZ issue. Triton kernel doesn't support data format nz, thus we skip
converting weight to nz on layer `conv1d`

- vLLM version: v0.11.0
- vLLM main:
2918c1b49c

---------

Signed-off-by: IncSec <1790766300@qq.com>
This commit is contained in:
InSec
2025-11-21 10:42:56 +08:00
committed by GitHub
parent cbb27feaf2
commit 5a4e8cdeba
10 changed files with 39 additions and 30 deletions

View File

@@ -20,6 +20,12 @@
Run `pytest tests/e2e/multicard/test_qwen3_next.py`.
"""
import os
from unittest.mock import patch
from modelscope import snapshot_download # type: ignore
from tests.e2e.conftest import VllmRunner
@@ -106,3 +112,23 @@ def test_models_distributed_Qwen3_NEXT_MTP_TP4_SIMILARITY():
print(f"spec_output: {spec_output[1]}")
assert matches > int(0.66 * len(ref_outputs))
# TODO: will conduct accuracy verification after the subsequent version becomes stable
@patch.dict(os.environ, {"HCCL_BUFFSIZE": "1024"})
def test_models_distributed_Qwen3_NEXT_W8A8DYNAMIC_WITH_EP():
example_prompts = [
"Hello, my name is",
]
max_tokens = 5
with VllmRunner(
snapshot_download(
"vllm-ascend/Qwen3-Next-80B-A3B-Instruct-W8A8-Pruning"),
max_model_len=4096,
tensor_parallel_size=2,
gpu_memory_utilization=0.4,
max_num_seqs=1,
enable_expert_parallel=True,
quantization="ascend",
) as vllm_model:
vllm_model.generate_greedy(example_prompts, max_tokens)