Skip to content

[Model] Support quantization of Qwen2VisionTransformer for Qwen2-VL #9817

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
youkaichao merged 7 commits into from
Oct 31, 2024
Merged

[Model] Support quantization of Qwen2VisionTransformer for Qwen2-VL #9817

Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
259bc5b
Fix prefix strings for quantized VLMs
mgoin Oct 28, 2024
61c49b7
Add prefix to language models
mgoin Oct 28, 2024
a41cff0
Fix qwen2-vl
mgoin Oct 29, 2024
75a4a30
Add optional prefix to model_loader build_model
mgoin Oct 29, 2024
2b7105f
Fix internlm2
mgoin Oct 29, 2024
1e7e236
Support quantization of Qwen2VisionTransformer for Qwen2-VL
mgoin Oct 29, 2024
6e2b706
Merge branch 'main' into qwen2vl-quant-vision
mgoin Oct 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
58 changes: 35 additions & 23 deletions vllm/model_executor/models/qwen2_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,15 +126,18 @@ def __init__(
hidden_features: int = None,
act_layer: Type[nn.Module] = QuickGELU,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
):
super().__init__()
self.fc1 = ColumnParallelLinear(in_features,
hidden_features,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.fc1")
self.act = act_layer()
self.fc2 = RowParallelLinear(hidden_features,
in_features,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.fc2")

def forward(self, x: torch.Tensor) -> torch.Tensor:
x_parallel, _ = self.fc1(x)
Expand Down Expand Up @@ -196,6 +199,7 @@ def __init__(
num_heads: Optional[int] = None,
projection_size: Optional[int] = None,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
super().__init__()
# Per attention head and per partition values.
Expand All @@ -207,10 +211,12 @@ def __init__(

self.qkv = ColumnParallelLinear(input_size=embed_dim,
output_size=3 * projection_size,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.qkv")
self.proj = RowParallelLinear(input_size=projection_size,
output_size=embed_dim,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.proj")

# Detect attention implementation.
self.attn_backend: _Backend = get_vit_attn_backend()
Expand Down Expand Up @@ -310,6 +316,7 @@ def __init__(
act_layer: Type[nn.Module] = QuickGELU,
norm_layer: Type[nn.Module] = None,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
super().__init__()
if norm_layer is None:
Expand All @@ -321,11 +328,13 @@ def __init__(
self.attn = Qwen2VisionAttention(embed_dim=dim,
num_heads=num_heads,
projection_size=dim,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.attn")
self.mlp = Qwen2VisionMLP(dim,
mlp_hidden_dim,
act_layer=act_layer,
quant_config=quant_config)
quant_config=quant_config,
prefix=f"{prefix}.mlp")

def forward(self, x: torch.Tensor, cu_seqlens: torch.Tensor,
rotary_pos_emb: torch.Tensor) -> torch.Tensor:
Expand Down Expand Up @@ -374,6 +383,7 @@ def __init__(
norm_layer: Type[nn.Module] = None,
spatial_merge_size: int = 2,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
super().__init__()
self.hidden_size = context_dim * (spatial_merge_size**2)
Expand All @@ -384,12 +394,14 @@ def __init__(
ColumnParallelLinear(self.hidden_size,
self.hidden_size,
bias=True,
quant_config=quant_config),
quant_config=quant_config,
prefix=f"{prefix}.mlp.0"),
nn.GELU(),
RowParallelLinear(self.hidden_size,
d_model,
bias=True,
quant_config=quant_config),
quant_config=quant_config,
prefix=f"{prefix}.mlp.2"),
])

def forward(self, x: torch.Tensor) -> torch.Tensor:
Expand Down Expand Up @@ -440,6 +452,7 @@ def __init__(
vision_config: Qwen2VLVisionConfig,
norm_eps: float = 1e-6,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
super().__init__()

Expand Down Expand Up @@ -467,28 +480,29 @@ def __init__(
self.rotary_pos_emb = Qwen2VisionRotaryEmbedding(head_dim // 2)

self.blocks = nn.ModuleList([
Qwen2VisionBlock(
dim=embed_dim,
num_heads=num_heads,
mlp_ratio=mlp_ratio,
norm_layer=norm_layer,
quant_config=quant_config,
) for _ in range(depth)
Qwen2VisionBlock(dim=embed_dim,
num_heads=num_heads,
mlp_ratio=mlp_ratio,
norm_layer=norm_layer,
quant_config=quant_config,
prefix=f"{prefix}.blocks.{layer_idx}")
for layer_idx in range(depth)
])
self.merger = Qwen2VisionPatchMerger(
d_model=hidden_size,
context_dim=embed_dim,
norm_layer=norm_layer,
quant_config=quant_config,
prefix=f"{prefix}.merger",
)

@property
def dtype(self) -> torch.dtype:
return self.blocks[0].mlp.fc2.weight.dtype
return self.patch_embed.proj.weight.dtype

@property
def device(self) -> torch.device:
return self.blocks[0].mlp.fc2.weight.device
return self.patch_embed.proj.weight.device

def rot_pos_emb(self, grid_thw: torch.Tensor) -> torch.Tensor:
pos_ids = []
Expand Down Expand Up @@ -932,10 +946,8 @@ def __init__(self,
self.visual = Qwen2VisionTransformer(
config.vision_config,
norm_eps=getattr(config, "rms_norm_eps", 1e-6),

# NOTE: Qwen2-VL vision encoder does not support any
# quantization method now.
quant_config=None,
quant_config=quant_config,
prefix="visual",
)

self.model = Qwen2Model(config,
Expand Down Expand Up @@ -1175,7 +1187,7 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
weight_loader(param, loaded_weight, shard_id)
break
else:
if "visual" in name and "qkv.weight" in name:
if "visual" in name and name.endswith("qkv.weight"):
visual_num_heads = self.config.vision_config.num_heads
visual_embed_dim = self.config.vision_config.embed_dim
head_size = visual_embed_dim // visual_num_heads
Expand All @@ -1184,7 +1196,7 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
visual_embed_dim)
loaded_weight = loaded_weight.transpose(0, 1)
loaded_weight = loaded_weight.reshape(-1, visual_embed_dim)
elif "visual" in name and "qkv.bias" in name:
elif "visual" in name and name.endswith("qkv.bias"):
visual_num_heads = self.config.vision_config.num_heads
visual_embed_dim = self.config.vision_config.embed_dim
head_size = visual_embed_dim // visual_num_heads
Expand Down