[Mini-vLLM] SM86-only cut: models, backends, Rust, tests
- Keep only Qwen3.5, Qwen3.5-MTP and Qwen4Exp model paths; trim the model registry to 41 entries - Delete idle attention backends (MLA, FlashInfer, ROCm-only, quant backends) and dead GDN CuTeDSL kernels - Remove inc/quark quant loaders; keep compressed-tensors and Marlin - Drop the Rust frontend, pooling entrypoints, and idle test suites - Rebuild kernels for sm_80;sm_86 and strip the extensions - Verified: full import sweep clean (optional deps only) and a dummy-load Qwen3.5 engine run on RTX 3090 via the Triton GDN path Co-authored-by: Hinata <hinata@idihore.id> Signed-off-by: MOVZX <movzx@yahoo.com>
This commit is contained in:
@@ -41,12 +41,7 @@ from vllm.config.utils import getattr_iter
|
||||
from vllm.distributed import get_pp_group, get_tp_group
|
||||
from vllm.distributed.utils import get_pp_indices
|
||||
from vllm.logger import init_logger
|
||||
from vllm.model_executor.layers.attention import (
|
||||
Attention,
|
||||
EncoderOnlyAttention,
|
||||
MLAAttention,
|
||||
)
|
||||
from vllm.model_executor.layers.attention.mla_attention import get_mla_dims
|
||||
from vllm.model_executor.layers.attention import Attention, EncoderOnlyAttention
|
||||
from vllm.model_executor.layers.attention_layer_base import AttentionLayerBase
|
||||
from vllm.model_executor.layers.fused_moe import MoERunner
|
||||
from vllm.model_executor.models.interfaces import (
|
||||
@@ -91,6 +86,11 @@ if TYPE_CHECKING:
|
||||
|
||||
from vllm.config import VllmConfig
|
||||
|
||||
try:
|
||||
from vllm.model_executor.layers.attention.mla_attention import MLAAttention
|
||||
except ImportError: # MLA support removed from this build
|
||||
MLAAttention = type("AbsentMLAAttention", (), {})
|
||||
|
||||
logger = init_logger(__name__)
|
||||
|
||||
VLLM_ATTN_ATTR = "attn"
|
||||
@@ -657,6 +657,10 @@ class Base(
|
||||
f"Layer {i} ({attn_prefix}) did not fuse as MLA, but the rest "
|
||||
"of the model did, so its latent KV cache cannot be allocated."
|
||||
)
|
||||
from vllm.model_executor.layers.attention.mla_attention import (
|
||||
get_mla_dims,
|
||||
)
|
||||
|
||||
dims = get_mla_dims(self.model_config)
|
||||
kwargs.update(
|
||||
qk_nope_head_dim=dims.qk_nope_head_dim,
|
||||
|
||||
Reference in New Issue
Block a user