[DSV4] Refactor RoPE initialization (#44262)

Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
This commit is contained in:
Woosuk Kwon
2026-06-02 01:26:58 +00:00
committed by GitHub
parent 48c0d13e65
commit 517e74a964
3 changed files with 50 additions and 40 deletions
+7 -20
View File
@@ -30,7 +30,6 @@ from vllm.model_executor.layers.mhc import (
MHCPreOp,
)
from vllm.model_executor.layers.quantization import QuantizationConfig
from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.layers.vocab_parallel_embedding import (
ParallelLMHead,
VocabParallelEmbedding,
@@ -50,6 +49,7 @@ from vllm.models.deepseek_v4.attention import (
DeepseekV4Indexer,
DeepseekV4MLA,
)
from vllm.models.deepseek_v4.common.rope import build_deepseek_v4_rope
from vllm.platforms import current_platform
from vllm.sequence import IntermediateTensors
from vllm.utils.import_utils import has_tilelang
@@ -314,25 +314,12 @@ class DeepseekV4Attention(nn.Module):
self.rope_parameters = config.rope_scaling
# Initialize rotary embedding BEFORE DeepseekV4MLA (which needs it)
rope_parameters = config.rope_parameters
rope_parameters["rope_theta"] = (
config.compress_rope_theta if self.compress_ratio > 1 else config.rope_theta
)
if config.rope_parameters["rope_type"] != "default":
config.rope_parameters["rope_type"] = (
"deepseek_yarn"
if config.rope_parameters.get("apply_yarn_scaling", True)
else "deepseek_llama_scaling"
)
rope_parameters["mscale"] = 0 # Disable mscale
rope_parameters["mscale_all_dim"] = 0 # Disable mscale
rope_parameters["is_deepseek_v4"] = True
rope_parameters["rope_dim"] = self.rope_head_dim
self.rotary_emb = get_rope(
self.head_dim,
max_position=self.max_position_embeddings,
rope_parameters=rope_parameters,
is_neox_style=False,
self.rotary_emb = build_deepseek_v4_rope(
config,
head_dim=self.head_dim,
rope_head_dim=self.rope_head_dim,
max_position_embeddings=self.max_position_embeddings,
compress_ratio=self.compress_ratio,
)
self.indexer = None
+36
View File
@@ -0,0 +1,36 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""DeepseekV4 rotary embedding initialization."""
from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.layers.rotary_embedding.base import RotaryEmbedding
def build_deepseek_v4_rope(
config,
*,
head_dim: int,
rope_head_dim: int,
max_position_embeddings: int,
compress_ratio: int,
) -> RotaryEmbedding:
rope_parameters = config.rope_parameters
rope_parameters["rope_theta"] = (
config.compress_rope_theta if compress_ratio > 1 else config.rope_theta
)
if rope_parameters["rope_type"] != "default":
rope_parameters["rope_type"] = (
"deepseek_yarn"
if rope_parameters.get("apply_yarn_scaling", True)
else "deepseek_llama_scaling"
)
rope_parameters["mscale"] = 0 # Disable mscale
rope_parameters["mscale_all_dim"] = 0 # Disable mscale
rope_parameters["is_deepseek_v4"] = True
rope_parameters["rope_dim"] = rope_head_dim
return get_rope(
head_dim,
max_position=max_position_embeddings,
rope_parameters=rope_parameters,
is_neox_style=False,
)
+7 -20
View File
@@ -35,7 +35,6 @@ from vllm.model_executor.layers.linear import (
)
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.quantization import QuantizationConfig
from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.layers.vocab_parallel_embedding import (
ParallelLMHead,
VocabParallelEmbedding,
@@ -56,6 +55,7 @@ from vllm.models.deepseek_v4.attention import (
DeepseekV4Indexer,
DeepseekV4MLA,
)
from vllm.models.deepseek_v4.common.rope import build_deepseek_v4_rope
from vllm.models.deepseek_v4.nvidia.ops.prepare_megamoe import prepare_megamoe_inputs
from vllm.sequence import IntermediateTensors
@@ -697,25 +697,12 @@ class DeepseekV4Attention(nn.Module):
self.rope_parameters = config.rope_scaling
# Initialize rotary embedding BEFORE DeepseekV4MLA (which needs it)
rope_parameters = config.rope_parameters
rope_parameters["rope_theta"] = (
config.compress_rope_theta if self.compress_ratio > 1 else config.rope_theta
)
if config.rope_parameters["rope_type"] != "default":
config.rope_parameters["rope_type"] = (
"deepseek_yarn"
if config.rope_parameters.get("apply_yarn_scaling", True)
else "deepseek_llama_scaling"
)
rope_parameters["mscale"] = 0 # Disable mscale
rope_parameters["mscale_all_dim"] = 0 # Disable mscale
rope_parameters["is_deepseek_v4"] = True
rope_parameters["rope_dim"] = self.rope_head_dim
self.rotary_emb = get_rope(
self.head_dim,
max_position=self.max_position_embeddings,
rope_parameters=rope_parameters,
is_neox_style=False,
self.rotary_emb = build_deepseek_v4_rope(
config,
head_dim=self.head_dim,
rope_head_dim=self.rope_head_dim,
max_position_embeddings=self.max_position_embeddings,
compress_ratio=self.compress_ratio,
)
self.indexer = None