Files
sglang/python/sglang/srt/models/qwen3_next_mtp.py
T

153 lines
6.1 KiB
Python

# Copyright 2023-2024 SGLang Team
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Inference-only Qwen3Next MTP Speculative Decoding."""
import copy
import logging
from contextlib import ExitStack
from typing import Iterable, Optional, Tuple
import torch
from torch import nn
from transformers import PretrainedConfig
from sglang.srt.distributed import get_pp_group
from sglang.srt.environ import envs
from sglang.srt.eplb.expert_distribution import get_global_expert_distribution_recorder
from sglang.srt.layers.layernorm import GemmaRMSNorm
from sglang.srt.layers.logits_processor import LogitsProcessor
from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.vocab_parallel_embedding import ParallelLMHead
from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.models.qwen3_next import Qwen3NextForCausalLM, Qwen3NextModel
from sglang.srt.runtime_context import get_model, get_parallel, get_spec
from sglang.srt.utils import add_prefix, is_npu
logger = logging.getLogger(__name__)
class Qwen3NextForCausalLMMTP(Qwen3NextForCausalLM):
def __init__(
self,
config: PretrainedConfig,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
) -> None:
nn.Module.__init__(self)
# Deep-copy so MTP mutations below don't leak into the target's config.
config = copy.deepcopy(config)
self.config = config
self.tp_size = get_parallel().tp_size
if is_npu() and get_spec().speculative_draft_model_quantization is None:
quant_config = None
self.quant_config = quant_config
# if not set, model load will be broken in Qwen3NextForCausalLM load_weights()
self.pp_group = get_pp_group()
# currently based on the provided ckpt, we:
# (1) do not use_dedicated_mtp_embeddings provided in ckpt since not provided and directly use the target model embeddings
# (2) hardcode bias=False since not provided
self.fc = nn.Linear(2 * config.hidden_size, config.hidden_size, bias=False)
RMSNorm_cls = GemmaRMSNorm
self.pre_fc_norm_embedding = RMSNorm_cls(
config.hidden_size, config.rms_norm_eps
)
self.pre_fc_norm_hidden = RMSNorm_cls(config.hidden_size, config.rms_norm_eps)
mtp_config = copy.deepcopy(config)
mtp_config.num_hidden_layers = 1
mtp_config.full_attention_interval = 1
self.model = Qwen3NextModel(
mtp_config,
quant_config,
prefix=add_prefix("model", prefix),
is_nextn=True,
)
self.lm_head = ParallelLMHead(
config.vocab_size,
config.hidden_size,
quant_config=quant_config,
prefix=add_prefix("model.shared_head.head", prefix),
use_attn_tp_group=get_parallel().config.enable_dp_lm_head,
)
self.logits_processor = LogitsProcessor(config)
# Mirror Qwen3NextForCausalLM.__init__'s shared-expert fusion setup so
# the inherited load_weights() can find the attribute on the MTP path.
# We compute it from the actual MTP MoE layer (1 layer with is_nextn=True),
# not hardcode it — when the layer's MoE pre-fuses the shared expert,
# load_weights must remap mlp.shared_expert.* into the fused slot.
self.num_fused_shared_experts = self._get_num_fused_shared_experts()
if self.num_fused_shared_experts > 1:
raise ValueError(
"Qwen3-Next MTP shared expert fusion currently supports exactly one "
"shared expert because checkpoint weight remapping maps it into "
"a single fused MoE expert slot."
)
self.enable_shared_expert_fusion = self.num_fused_shared_experts > 0
@torch.no_grad()
def forward(
self,
input_ids: torch.Tensor,
positions: torch.Tensor,
forward_batch: ForwardBatch,
input_embeds: Optional[torch.Tensor] = None,
**kwargs,
):
exit_stack = ExitStack()
if (
is_npu()
and self.quant_config is None
and get_model().quantization is not None
):
# ascend mtp unquant
exit_stack.enter_context(envs.SGLANG_DEEPEP_BF16_DISPATCH.override(True))
exit_stack.enter_context(
envs.DEEP_NORMAL_MODE_USE_INT8_QUANT.override(False)
)
try:
if input_embeds is None:
input_embeds = self.model.embed_tokens(input_ids)
hidden_states = forward_batch.spec_info.hidden_states
# Some idle batch has 0 batch size. GemmaRMSNorm.forward would fail due to bs=0.
if not forward_batch.forward_mode.is_idle():
input_embeds = self.pre_fc_norm_embedding(input_embeds)
hidden_states = self.pre_fc_norm_hidden(hidden_states)
hidden_states = self.fc(torch.cat((input_embeds, hidden_states), dim=-1))
with get_global_expert_distribution_recorder().disable_this_region():
hidden_states = self.model(
input_ids,
positions,
forward_batch,
hidden_states,
)
finally:
exit_stack.close()
return self.logits_processor(
input_ids, hidden_states, self.lm_head, forward_batch
)
def load_weights(
self, weights: Iterable[Tuple[str, torch.Tensor]], is_mtp: bool = False
):
super().load_weights(weights, is_mtp=True)
EntryClass = [Qwen3NextForCausalLMMTP]